Final Program

International Conference onSwarm Intelligence

Technical Program & Abstracts

June 12-15, 2010

The Lake View Hotel, Beijing, China

http://www.cil.pku.edu.cn/icsci2010

Welcome Message from General Chair

I am very pleased to welcome you, swarm intelligence researchers and practitioners from

China and all over the world, to attend ICSI 2010 in the beautiful Beijing of China in its best

season. I believe that you will enjoy this important and hard-to-get gathering both from the

perspective of swarm intelligence research and the perspective of the Chinese culture.

Swarm Intelligence is to study the collective behavior in a decentralized system which

is made up by a population of simple individuals interacting locally with one another and

with their environment. Such systems are often be found in nature, including bird flocking,

ant colonies, particles in cloud, fish schooling, bacteria foraging, animal herding, honey bees,

spiders, and sharks to just name a few. Although there is typically no centralized control

harnessing the behavior of the individuals, local interactions among individuals often cause a

global pattern to emerge. Modeling of these natural biological systems and social phenomenon

has become one of the most important methodologies of studying artificial intelligence from the

system point of view. The extension of swarm intelligence methods to optimizing computations,

pattern recognition, the interdisciplinary merging with robotics, control, machine learning,

parallel processing and complex systems will generate an enormous range of research topics

and potential applications in most scientific and engineering fields.

Thanks to the hard work of the Program Committee and the Organization Committee,

this conference ICSI2010 will provide you with a very good program. The excellent plenary

speakers will introduce you to the frontiers of SI research and applications, and will help you

to identify important future research directions.

The venue of ICSI 2010, the Lake View Hotel, next to Peking University, the best university

in China, is in the center of the ”Golden Academic Triangle” of Beijing and located between

Zhong-Guan-Cun and Shang-Di, the Chinese ”Silicon Valley”. Therefore, you not only can

experience the campus atmosphere of the top Chinese university but also see the rapidly

developing Chinese IT industries during your stay at the conference. The Lake View Hotel is

within walking distance from the Park of the Yuan-Min-Yuan Ruins and a short drive from

the Summer Palace and the Olympic Park.

On the other hand, the ICSI 2010 will definitely contribute a lot to the globalization of

research and teaching in China in addition to the enhancement of the research horizons of the

conference attendants. Certainly, the participants of ICSI’2010 can also enjoy Peking operas,

beautiful landscapes in Beijing, and the hospitality of the Chinese people, and a modern China.

2

International Conference on Swarm Intelligence, June 12-15, 2010, Beijing China

I hope ICSI 2010 will be a memorable event for you. In addition, the famous Shanghai

Expo 2010 is also welcoming you, the guests from all over the world, after the conference.

Sincerely yours!

Ying Tan

ICSI2010 General Chair

Peking University, China

3

Welcome Message from Program Committee Chairs

The International Conference on Swarm Intelligence (ICSI 2010) is the first of its series

gatherings in the world for the researchers working on all aspects of swarm intelligence. It

will provide an academic forum for the participants to disseminate their new research findings

and discuss emerging areas of research. It also will create a stimulating environment for the

participants to interact and exchange information on future challenges and opportunities of

swarm intelligence research.

The aim of this important meeting is to exhibit the state-of-the-art research and development

in swarm intelligence from both theoretical and practical aspects.

The ICSI 2010 received totally 394 submissions from about 1241 authors in 22 countries

and regions (Australia, Belgium, Brazil, Canada, China, Cyprus, Hong Kong, Hungary, India,

Islamic Republic of Iran, Japan, Jordan, Republic of Korea, Malaysia, Mexico, Norway, Pakistan,

South Africa, Chinese Taiwan, United Kingdom, United States of America, Vietnam) across

six continents (Asia, Europe, North America, South America, Africa, and Oceania). Each

submission was reviewed by at least 3 reviewers, and on average 3.8 reviewers. Based on

rigorous reviews by the Program Committee members and reviewers, 185 high-quality papers

were selected for publication in the proceedings with the acceptance rate of 46.9%. The papers

are organized in 25 cohesive sections covering all major topics of swarm intelligence research

and development.

In addition to the contributed papers, the ICSI 2010 technical program includes three

plenary speeches by Prof. Benjamin W. Wah (Provost, The Chinese University of Hong Kong,

Hong Kong, China), Prof. Gary G. Yen (President of IEEE Computational Intelligence Society

(CIS), Oklahoma State University, USA), Prof. Nikola Kasabov (President of International

Neural Network Soceity (INNS), Auckland University of Technology, New Zealand). Besides

the regular parallel oral sessions, ICSI 2010 also has three poster sessions focusing on wider

areas.

As Program Co-Chairs of the ICSI 2010, we would like to express our sincere thanks to

Peking University and Xi’an Jiaotong-Liverpool University for their sponsorship, to IEEE

Beijing Section, International Neural Network Society, World Federation on Soft Computing,

Chinese Association for Artificial Intelligence, and National Natural Science Foundation of

China for their technical co-sponsorship, to the National Natural Science Foundation of China

and K. C. Wong Education Foundation, Hong Kong for their financial and logistic supports.

4


We would also like to thank the members of the Advisory Committee for their guidance,

the members of the International Program Committee and all reviewers for reviewing the

papers, and the members of the Publications Committee for checking the accepted papers in

a short period of time. Particularly, we are grateful to the proceedings publisher, Springer,

for publishing the proceedings in the prestigious series of Lecture Notes in Computer Science.

Furthermore, we wish to express our heartfelt appreciation to the plenary speakers, session

chairs, student volunteers, and. colleagues, associates, friends, and supporters who helped us

in immeasurable ways. Last but not the least, we would like to thank all the speakers and

authors and participants for their great contributions that made ICSI 2010 successful and all

the hard work worthwhile.

The technical program offers an outstanding collection of recent research contributions

by top researchers. It should be of interest to both theoreticians and practitioners and is a

must-have resource for researchers interested in swarm intelligence.

We highly appreciate the three plenary speakers for delivering plenary talks. We are greatly

thankful to all the authors for their excellent contributions, to all the invited session organizers

for their joint effort and enthusiasm, and to all the international program committee members

and referees for their time and expertise in the paper review process. Also, special thanks go

to Chao Deng, Huiyun Guo, Yuanchun Zhu, Jun Wang, Pengtao Zhang, Zhongyang Zheng, Xi

Huang, You Zhou and Rui Chao for their time and outstanding work in the organization of

ICSI 2010.

We sincerely hope that all ICSI 2010 participants will enjoy attending conference sessions

and activities, meeting old and new friends, and setting up new research collaborations.

Have a pleasant stay in Beijing and enjoy.

Cheers!

Yuhui Shi

Program Committee Chair

Xi’an Jiaotong-Liverpool University, China

Kay Chen Tan

Program Committee Chair

National University of Singapore, Singapore

5

Contents

Messages 2

Welcome Message from General Chair . . . . . . . . . . . . . . . . . . 2

Welcome Message from Program Committee Chairs . . . . . . . . . . . 4

Venue 7

Sponsors 10

Organizing Committees 11

Programme Committee Members 12

Reviewers 15

Program Overview 16

Conference Excursion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

Talks 18

Plenary Talk I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Plenary Talk II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Plenary Talk III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Technical Program 23

June 13, 2010(Sunday) . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

June 14, 2010(Monday) . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

Posters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

Abstracts 41

Index 93

6


Venue

The International Conference on Swarm Intelligence (ICSI’2010) will be held in The Lake

View Hotel. As the first five-star cultural and business hotel in the Zhongguancun area in

recent years, The Lake View Hotel located in the core area of Zhongguancun, known as the

”Silicon Valley of China”, the hotel boasts some famous neighbors, including Peking University,

the Weiming Lake, the Boya Pagoda and the Olympic Center. A mere 30 minutes drive to

the Capital Airport and 20 minutes to Tiananmen Square, the hotel is close to the Summer

Palace, Yuanmingyuan (Garden of Perfect Splendor) and the Fragrance Hills. Here you can

have a taste of the humanist spirit of the prestigious university with a 100-year history, or

enjoy the noble style of the royal and aristocratic families. Here the Chinese culture with a

long tradition is shining ever brightly. The Lake View Hotel boasts a total construction area

of over 40 thousand m2. The Hotel has 336 luxury guest rooms and suites of all kinds, using

geothermal hot spring 3000 meters from under the ground, a high-end international business

area, complete entertainment and recreational facilities(25m×7.5m Swimming Pool ), and a

6000 m2 Courtyard Garden. With primarily a modern outlook, the hotel has integrated the

profound cultural heritage of Peking University with modern luxury business facilities.

Thanks to its superior geographic location, the Hotel boasts a unique historical and cultural

touch. With a sacred mission for international exchange, it is committed to providing an

important platform of exchange for business elites and building a magnetic field of international

exchange for cultural/business activities. Most of all, forging an international ”Hub for Confucian

Business Elites” is its responsibility and aspiration.

7

The Lake View Hotel

8


Location Map of Conference Rooms on B1 Floor

Useful Local Telephone NumbersCountry code: 86Directory Enquiry: 114Emergency Service: police 110, fire 119 and ambulance 999

Internet AccessWi-fi and fixed line access of the Internet is available at Lake View Hotel

9

Sponsors

Sponsor/Organizer

Peking University

Co-Sponsor

Xi’an Jiaotong-Liverpool University

Financial Co-Sponsors

National Natural Science Foundation of China

K. C. Wong Education Foundation, Hong Kong

Technical Co-Sponsors

IEEE Beijing Section

International Neural Networks Society

World Federation on Soft Computing

Chinese Association for Artificial Intelligence

Publishers

Lecture Notes in Computer Science

Springer

10


Organizing Committees

Honorary Chairs Qidi Wu, Ministry of Education, China

Russell C. Eberhart, Indiana University Purdue University,

USA

General Chair Ying Tan, Peking University, China

Program Committee

Chairs

Yuhui Shi, Xi’an Jiaotong-Liverpool University, China

Kay Chen Tan, National University of Singapore, Singapore

Technical Committee

Chairs

Gary G. Yen, Oklahoma University, USA

Jong-Hwan Kim, KAIST, Republic of Korea

Xiaodong Li, RMIT University, Australia

Xuelong Li, University of London, UK

Frans van den Bergh, South Africa

Advisory Committee

Chairs

Zhenya He, Southeast University, China

Xingui He, Peking University, China

Xin Yao, Birmingham University, UK

Yixin Zhong, Beijing University of Posts and

Telecommunications, China

Plenary Sessions

Chairs

Robert G. Reynolds, Wayne State University, USA

Qingfu Zhang, University of Essex, UK

Special Sessions

Chairs

Martin Middendorf, University of Leipzig, Germany

Jun Zhang, Sun Yat-Sen University, China

Haibo He, Stevens Institute of Technology, USA

Tutorial Chairs Carlos Coello Coello, CINVESTAV-IPN, Mexico

Publications Chairs Zhishun Wang, Columbia University, USA

Publicity Chairs Ponnuthurai N. Suganthan, Nanyang Technological

University, Singapore

Lei Wang, Tongji University, China

Maurice Clerc, Universite de Paris, France

Finance Chair Chao Deng, Peking University, China

Registration Chairs Huiyun Guo, Peking University, China

Yuanchun Zhu, Peking University, China

11

Programme Committee Members

Aimin Zhou, Normal University of East China, China

Amir Hussain, University of Stirling, UK

Andries Engelbrecht, University of Pretoria, South Africa

Arindam Das, University of Washington, USA

Arun Khosla, Dr B R Ambedkar National Institute of Technology, India

Ben Niu, Shenzhen University, China

Bernd Meyer, Monash University, Australia

Bijaya Ketan Panigrahi, IIT Delhi, India

Bing Wang, University of Hull, UK

Bruno Apolloni, University Milano, Italy

Carlos A. Coello Coello, CINVESTAV-IPN, Mexico

Cheng Xiang, National University of Singapore, Singapore

Chris Lokan, University of New South Wales, Australia

Christian Blum, Universitat Politecnica de Catalunya, Spain

Christos Tjortjis, University of Ioannina,University of Western Macedonia, Greece

Colin G. Johnson, University of Kent at Canterbury, UK

Dingli Yu, Liverpool John Moores University, UK

Emilio Corchado, University of Burgos, Spain

Erkki Oja, Helsinki University, FI

Erol Gelenbe, Imperial College London, UK

Fernando Lobo, Universidade do Algarve, Portugal

Francesco Mondada, EPFL C STI C LSRO, Switzerland

Frans van den Bergh, CSIR SAC (Pretoria), South Africa

Franziska Klugl, Orebro University, Germany

G K Venayagamoorthy, Missouri University of Science and Technology, USA

Giovanni Sebastiani, Consiglio Nazionale delle Ricerche, Italy

Guangbin Huang, Nanyang Technological University, Singapore

Guoping Liu, University of Glamorgan, UK

Guoyin Wang, Chongqing University of Posts and Telecommunications, China

Haibin Duan, Beihang University, China

Haibo He, Stevens Institute of Technology, USA

Hongtao Lu, Shanghai Jiaotong University, China

Hongwei Mo, Harbin Engineering University, China

Huosheng Hu, University of Essex, UK

Jie Zhang, University of New Castle, UK

Jinde Cao, Southeast University, China

Jinhua Zheng, Xiangtan University, China

Jinwen Ma, Peking University, China

Jivesh Govil, Cisco Systems Inc., USA

Jose Alfredo Ferreira Costa, Federal University, Brazil

Jose C. Principe, University of Florida, USA

Ju Liu, Shandong University, China

12


Jun Zhang, Sun Yat-Sen University, China

Junqi Zhang, Tongji University, China

Kang Li, Queen’s University Belfast, UK

Kay Chen Tan, National University of Singapore, Singapore

Kusum Deep, Indian Institute of Technology, India

Leandro dos Santos Coelho, Pontifical Catholic University of Parana, Brazil

Lei Wang, Tongji University, China

Liang Chen, University of Northern British Columbia, Canada

Liangpei Zhang, Wuhan University, China

Lifeng Zhang, Renmin University of China, China

Lixiang Li, Beijing University of Post and Telecommunications, China

Ling Wang, Tsinghua University, China

Luca Maria Gambardella, Istituto Dalle Molle di Studi sull’Intelligenza Artificiale, Switzerland

Mahamed G. H. Omran, Gulf University for Science & Technology, Kuwait

Mark Embrechts, Rensselaer Inst., USA

Martin Brown, University of Manchester, UK

Martin Middendorf, University of Leipzig, Germany

Meng Joo Er, Nanyang Technological University, Singapore

Michael Small, Hong Kong Polytechnic University, Hong Kong, China

Mingcong Deng, Okayama University, Japan

Mongguo Gong, Xidian University, China

Nikola Kasabov, Auckland University, New Zealand

Norikazu Takahashi, Kyushu University, Japan

Oscar Cordon, European Centre for Soft Computing, SPAIN

Paul S. Pang, Auckland University of Technology, New Zealand

Payman Arabshahi, University of Washington, USA

Peter Andras, Newcastle University, UK

Peter Erdi, HU & Kalamazoo College, USA

Peter Tino, Birmingham University, UK

Ping Guo, Beijing Normal University, China

Ponnuthurai Nagaratnam Suganthan, Nanyang Technological University, Singapore

Prithviraj (Raj) Dasgupta, University of Nebraska, USA

Qingfu Zhang, University of Essex, UK

Qing-Long Han, Central Queensland University, Australia

Ran Tao, Beijing Institute of Technology, China

Ruhul A. Sarker, Australian Defence Force Academy, Australia

Sabri Arik, Istanbul University, Turkey

Salim Bouzerdoum, University of Wollongong, Australia

Shenli Xie, South-China University of Technology, China

Simon X. Yang, University of Guelph, Canada

Suicheng Gu, Temple University, USA

Thomas E. Potok, Oak Ridge National Laboratory, USA

Wai Keung Fung, University of Manitoba, Canada

13

Wenjian Luo, University of Science and Technology of China, China

Wenlian Lu, Fudan University , China

Xiaohui Cui, Oak Ridge National Laboratory, USA

Xiaohui Hu, Indiana University Purdue University Indianapolis, USA

Xiaoli Li, University of Birmingham, UK

Xiujun Ma, Peking University, China

Xuelong Li, London University, UK

Yangmin Li, University of Macau, Macao, China

Yanqing Zhang, Georgia State University, USA

Yi Shen, Huazhong University of Science and Technology, China

Yi Zhang, Sichuan University, China

Ying Tan, Peking University, China

Yingjie Yang, De Montfort University, UK

Yongsheng Ding, Donghua University, China

Yoshikazu Fukuyama, Honda Inc., Japan

Yuhui Shi, Xi’an Jiaotong-Liverpool University, China

Zhen Ji, Shenzhen University, China

Zhengguang Hou, Institute of Automation, CAS, China

Zheru Chi, Hong Kong Polytechnic University, Hong Kong, China

Zhifeng Hao, South China University of Technology, China

Zhigang Zeng, Hua Zhong University of Technology, China

Zhi-Hua Zhou, Nanjing University, China

14


Reviewers

Ajiboye Saheeb Osunleke

Akira Yanou

Antonin Ponsich

Bingzhao Li

Bo Liu

Carson K. Leung

Changan Jiang

Chen Guici

Ching-Hung Lee

Chonglun Fang

Cong Zheng

Dawei Zhang

Daoqiang Zhang

Dong Li

Fei Ge

Feng Jiang

Gan Huang

Gang Chen

Haibo Bao

Hongyan Wang

Hugo HernAndez

I-Tung Yang

IbaAez Panizo

Jackson Gomes

Janyl Jumadinova

Jin Hu

Jin Xu

Jing Deng

Juan Zhao

Julio Barrera

Jun Guo

Jun Shen

Jun Wang

Ke Cheng

Ke Ding

Kenya Jinno

Liangpei Zhang

Lihua Jiang

Lili Wang

Lin Wang

Liu Lei

Lixiang Li

Lorenzo Valerio

Naoki Ono

Ni Bu

Orlando Coelho

Oscar IbaAez

Pengtao Zhang

Prakash Shelokar

Qiang Lu

Qiang Song

Qiao Cai

Qingshan Liu

Qun Niu

Renato Sassi

Satvir Singh

Sergio P. Santos

Sheng Chen

Shuhui Bi

Simone Bassis

Song Zhu

Spiros Denaxas

Stefano Benedettini

Stelios Timotheou

Takashi Tanizaki

Usman Adeel

Valerio Arnaboldi

Wangli He

Wei Wang

Wen Shengjun

Wenwu Yu

X.M. Zhang

Xi Huang

Xiaolin Li

Xin Geng

Xiwei Liu

Yan Yang

Yanqiao Zhu

Yongqing Yang

Yongsheng Dong

Yulong Wang

Yuan Cao

15

Program Overview

Date Time Events

(At the Lake View Hotel)

June 12, 2010 08:00am-20:00pm Registration Day

(Lobby)

June 13, 2010 08:00am-08:10am Opening Ceremony

(ZhongHua Banquet Hall)

08:20am-12:00am Plenary Talks


12:00am-13:30pm Lunch

(The Greenery Coffee Shop)

13:30pm-17:30pm Poster

(Corridor next to Daxuetang Conference Rooms)

13:30pm-17:30pm Parallel Oral Sessions

(Daxuetang Conference Rooms A&B&C)

18:00pm-20:30pm Banquet


June 14, 2010 08:00pm-12:00am Parallel Oral Sessions


08:00pm-12:00am Poster

(Corridor next to Daxuetang Conference Rooms)

12:00am-13:30pm Lunch

(The Greenery Coffee Shop)

13:30pm-17:50pm Parallel Oral Sessions


June 15, 2010 07:30am-18:30pm City Tour

(Boarding at the main entrance of the Lake View

Hotel before 07:30am)

16


Conference Excursion

One-Day City Tour Scheduling for ICSI 2010 Delegates

7:30 am The sightseeing bus will pick you up at the main entrance of the Lakeview

Hotel (There might be other picking up sites.)

8:00 am Visit the Summer Palace - The largest imperial gardens exists in China.

You will tour around Kunming Lake and Longevity Hill, which are the

main features of the Summer Palace.

Around 12:00

pm

Lunch

13:30 pm Vist the Forbidden City - The Forbidden City was the imperial palaces of

the Ming and Qing dynasties, known as the Palace Museum. The main

tractions you will visit include the three halls of QianChao (Outer Court),

the three palaces of HouQin ( inner court) and YuHuaYuan (imperial

Garden).

Around 4:30 pm Tour around the National Center for the Performing Arts

Around 5:30 pm Tour around and photo taking in Beijing Olympic Park (Bird’s Nest,

Water Cube, etc.)

6:30 pm The bus will send you back to the Lakeview Hotel

17

Plenary Talk I

Exploring Constraint Partitioning and Their Applications In NaturalComputations

Prof. Benjamin W. WahThe Chinese University of Hong Kong, Hong Kong, China

Abstract

Constraint optimization exists in many natural-computation applications, including neural andevolutionary computations. A key observation on the constraints of many of these applicationproblems is that they are highly structured and involve variables with strong spatial or temporallocality. Based on this observation, large-scale problems in these applications can be partitionedby their constraints into a small number of much simpler subproblems. Because each subproblemhas only a fraction of the original constraints, it is a significant relaxation of the originalproblem and has a dramatically lower complexity. As a result, many problems that cannotbe solved before can now be solved easily. In this talk, we present the application of thisapproach in solving some neural learning and evolutionary computation problems. Based on apartition-and-resolve strategy, we evaluate techniques for resolving violated global constraints.

Biography

Benjamin W. Wah is currently the Provost of the Chinese University ofHong Kong. He has served as the Franklin W. Woeltge Endowed Professorof Electrical and Computer Engineering and Professor of the CoordinatedScience Laboratory of the University of Illinois at Urbana-Champaign,Urbana, IL. In 2009, he served as the Director of the Advanced DigitalSciences Center, a large research center of the University of Illinois locatedin Singapore and funded by Singapores Agency for Science, Technology,and Research (A*STAR). He received his Ph.D. degree in computer sciencefrom the University of California, Berkeley, CA, in 1979. Previously, he

had served on the faculty of Purdue University (1979-85), as a Program Director at the NationalScience Foundation (1988-89), as Fujitsu Visiting Chair Professor of Intelligence Engineering,University of Tokyo (1992), and McKay Visiting Professor of Electrical Engineering andComputer Science, University of California, Berkeley (1994). In 1989, he was awarded aUniversity Scholar of the University of Illinois; in 1998, he received the IEEE ComputerSociety Technical Achievement Award; in 2000, the IEEE Millennium Medal; in 2003, theRaymond T. Yeh Lifetime Achievement Award from the Society for Design and ProcessScience; in 2006, the IEEE Computer Society W. Wallace-McDowell Award and the PanWen-Yuan Outstanding Research Award, in 2007, the IEEE Computer Society Richard E.Merwin Award and the IEEE-CS Technical Committee on Distributed Processing OutstandingAchievement Award, and in 2009, the IEEE-CS Tsutomu Kanai Award. Wah’s current researchinterests are in the areas of nonlinear search and optimization, multimedia signal processing,and computer networks. Wah cofounded the IEEE Transactions on Knowledge and DataEngineering in 1988 and served as its Editor-in-Chief between 1993 and 1996, and is theHonorary Editor-in-Chief of Knowledge and Information Systems. He currently serves on theeditorial boards of Information Sciences, International Journal on Artificial Intelligence Tools,Journal of VLSI Signal Processing, World Wide Web, and Neural Processing Letters. He hadchaired a number of international conferences, including the 2000 IFIP World Congress andthe 2006 IEEE/WIC/ACM International Conferences on Data Mining and Intelligent AgentTechnology. He has served the IEEE Computer Society in various capacities, including VicePresident for Publications (1998 and 1999) and President (2001). He is a Fellow of the AAAS,ACM, and IEEE.

18


Plenary Talk II

Population Control in Multi-Objective Particle Swam Optimization

Prof. Gary G. Yen

Oklahoma State University, U.S.A.

Abstract

Evolutionary computation is the study of biologically motivated computational paradigms

which exert novel ideas and inspiration from natural evolution and adaptation. The application

of population-based heuristics in solving multi-objective Optimization Problems has been

receiving a growing interest from computational intelligence community. To search for a family

of ”acceptable” solutions, a so called Pareto set, by using population-based parallel searching

ability, several multi-objective Particle Swarm Optimization (MOPSOs) have been proposed.

However, most of these designs have difficulty in dealing with the trade-off between uniformly

distributing the computational resources and finding the near-complete and near-optimal Pareto

set. On the other hand, according to the No Free Lunch theorems, no formal assurance of an

algorithms general effectiveness exists if insufficient knowledge of the problem characteristics is

incorporated into the algorithm domain. In this talk, population control is being implemented

into particle swarm optimization, differential evolution and artificial immune system for a

more effective design of multi-objective optimization. We will propose works along this line of

research in dynamically regulating the population as needed in different stage of evolutionary

process, some voluntary while others compulsory, in pursuing a uniformly distributed, near

optimal, and close to complete Pareto front for a given MOP. Through numerical study, we

will show these designs incorporating population control strategy provide very competitive

performances qualitatively and quantitatively compared to some chosen state-of-the-art evolutionary

algorithms.

Biography

Gary G. Yen received the Ph.D. degree in electrical and computer

engineering from the University of Notre Dame, Notre Dame, Indiana in

1992. He is currently a Professor in the School of Electrical and Computer

Engineering, Oklahoma State University (OSU). Before he joined OSU in

1997, he was with the Structure Control Division, U.S. Air Force Research

Laboratory in Albuquerque, NM. His research is supported by the DoD,

DoE, EPA, NASA, NSF, and Process Industry.

19

Dr. Yen was an associate editor of the IEEE Transactions on Neural Networks,

IEEE Control Systems Magazine, IEEE Transactions on Control Systems Technology, IEEE

Transactions on Systems, Man and Cybernetics and IFAC Journal on Automatica and

Mechatronics. He is currently serving as an associate editor for the IEEE Transactions on

Evolutionary Computation. He served as the General Chair for the 2003 IEEE International

Symposium on Intelligent Control held in Houston, TX and 2006 IEEE World Congress on

Computational Intelligence held in Vancouver, Canada. Dr. Yen served as Vice President

for the Technical Activities, IEEE Computational Intelligence Society in 2004-2005 and is the

founding editor-in-chief of the IEEE Computational Intelligence Magazine since 2006. Most

recently, he is elected to serve as President Elect in 2009 and President in 2010-2011 of the

IEEE Computational Intelligence Society. He is a Fellow of IEEE.

20


Plenary Talk III

Quantum Inspired PSO for Integrated Feature and Parameter Optimization of

Evolving Spiking Neural Networks: Methods and Applications

Prof. Nikola Kasabov

Knowledge Engineering and Discovery Research Institute

Auckland University of Technology, New Zealand

Abstract

This talk presents a novel Quantum-inspired Particle Swarm Optimization (QiPSO) method

that further develops the PSO principles and the quantum inspired evolutionary algorithms

(QiEA) . It was proved that the QiEA are multi-model EDA type algorithms, that opens the

application of these algorithms for probabilistic feature and model parameter optimization

across classification and prediction models and applications . The talk reveals the interesting

concept of QiPSO in which information is represented as highly parallel quantum-bit structures

utilizing the principle of state superposition. The proposed QiPSO is applied for an integrated

optimization of features and parameters of Evolving Spiking Neural Networks (ESNN) C

novel and promising connectionist models for classification and pattern recognition . The

mechanism simultaneously optimizes the ESNN parameters and relevant features using the

wrapper approach. A synthetic dataset and application case studies are used to evaluate

the performance of the proposed method. The results show that QiPSO leads to promising

outcomes in obtaining the best combination of ESNN parameters as well as in identifying the

most relevant features in a large dimensional space. The method is also applicable to the newly

introduced probabilistic ESNN. Open questions, challenges and directions for further research

are presented.

Biography

Professor Nikola Kasabov is the Founding Director and the Chief

Scientist of the Knowledge Engineering and Discovery Research Institute

(KEDRI), Auckland (www.kedri.info/). He holds a Chair of Knowledge

Engineering at the School of Computing and Mathematical Sciences at

Auckland University of Technology. He is a Fellow of the Royal Society

of New Zealand and Fellow of the New Zealand Computer Society. He

is the President of the International Neural Network Society (INNS) and a Past President

of the Asia Pacific Neural Network Assembly (APNNA). He is a member of several technical

21

committees of the IEEE Computational Intelligence Society and of the IFIP AI TC12. Kasabov

is a Co-Editor-in-Chief of Evolving Systems J. and Associate Editor of several international

journals, including Neural Networks, IEEE Trans. Fuzzy Systems, Information Science, J.

Theoretical and Computational Nanoscience, Applied Soft Computing. He chairs a series

of int. conferences ANNES/NCEI in New Zealand. Kasabov holds MSc and PhD from

the Technical University of Sofia, Bulgaria. His main research interests are in the areas of

intelligent information systems, soft computing, neuro-computing, bioinformatics, brain study,

novel methods for data mining and knowledge discovery. He has published more than 400

publications that include 15 books, 120 journal papers, 60 book chapters, 28 patents, and

numerous conference papers. He has extensive academic experience at various academic

and research organizations: University of Otago, New Zealand; University of Essex, UK;

University of Trento, Italy; Technical University of Sofia, Bulgaria; University of California

at Berkeley; RIKEN and KIT, Japan; TUniversity Kaiserslautern, Germany, and others.

Kasabov has received numerous awards, among them: The Bayer 2007 Science Innovation

Award; The 2005 APNNA Excellent Service Award; The RSNZ Silver Medal for Science and

Technology; Best paper awards; The UK Leverhulme Trust Scholarship; The Dutch Research

Organisation Scholarship; The NZ FRST and AUT VC Best Postgraduate Supervision Awards;

and others. More information of Prof. Kasabov can be found on the KEDRI web site:

http://www.kedri.info.

22


Technical Program

June 13, 2010(Sunday)

08:00-08:10 Opening Ceremony ZhongHua Banquet Hall

08:20-09:25

Plenary Talk I

Speaker: Prof. Benjamin W. Wah

Chair: Prof. Yuhui Shi

ZhongHua Banquet Hall

09:30-10:35

Plenary Talk II

Speaker: Prof. Gary G. Yen

Chair: Prof. Yuhui Shi


10:35-10:55 Tea/Coffee BreakMain Entrance of ZhongHua

Banquet Hall

10:55-12:00

Plenary Talk III

Speaker: Prof. Nikola Kasabov

Chair: Prof. Ying Tan


12:00-13:30 Lunch The Greenery Coffee Shop

13:30-17:30 Poster Session Corridor

13:30-15:30 Oral Sessions Daxuetang Conference Rooms

Room A Room B Room C

Theoretical Analysis of

Swarm Intelligence

Algorithms

Applications of PSO

Algorithms

PSO Algorithms (I)

15:30-15:50 Tea/Coffea Break Corridor



PSO Algorithms (II) ACO Algorithms Novel Swarm-Based

Optimization Algorithms

(I)

18:00-20:30 Banquet ZhongHua Banquet Hall

23

June 14, 2010(Monday)



Novel Swarm-Based

Optimization Algorithms

(II)

Hybrid Algorithms Artificial Immune System




Multi-Agent Based

Complex Systems

Multi-Robot Systems Classifier Systems

08:00-12:00 Poster Session Corridor

12:00-13:30 Lunch The Greenery Coffee Shop



Evolutionary Computation Data Mining Methods Fuzzy Methods and

Information Processing

System




Signal Processing and

Information Security

Intelligent Control Other Optimization

Algorithms

24


Oral Sessions


Session Theoretical Analysis of Swarm

Intelligence Algorithms

Chair Hong Zhang

Date/Time June 13, 2010(Sunday) 13:30-15:10 Venue Room A

13:30 - 13:50 The Performance Measurement of a Canonical Particle Swarm

Optimizer with Diversive Curiosity

P41

Hong Zhang and Jie Zhang

13:50 - 14:10 On the Farther Analysis of Performance of the Artificial Searching

Swarm Algorithm

P41

Tanggong Chen, Lijie Zhang and Lingling Pang

14:10 - 14:30 Back-Propagation vs Particle Swarm Optimization Algorithm:

which Algorithm is better to adjust the Synaptic Weights of a

Feed-Forward ANN?

P41

Beatriz Aurora Garro Licon, Humberto Sossa Azuela and Roberto

Antonio Vazquez

14:30 - 14:50 Orthogonality and Optimality in Non-Pheromone Mediated

Foraging

P42

Sanza Kazadi, James Yang, James Park and Andrew Park

14:50 - 15:10 An Adaptive Staged PSO Based on Particles’s Search Capabilities P42

Kun Liu and Ying Tan

Session Applications of PSO Algorithms Chair Peng-Yeng Yin

Date/Time June 13, 2010(Sunday) 13:30-15:30 Venue Room B

13:30 - 13:50 Medical Image Registration Based on Generalized Mutual

Information and PSO Algorithm

P43

Jingzhou Zhang, Pengfei Huo, Jionghua Teng, Xue Wang and

Suhuan Wang

25

13:50 - 14:10 Particle Swarm Optimization For Automatic Selection of Relevance

Feedback Heuristics

P43

Peng-Yeng Yin

14:10 - 14:30 A New Particle Swarm Optimization Solution to Nonconvex

Economic Dispatch Problem

P44

Jianhua Zhang, Yingxin Wang, Rui Wang and Guolian Hou

14:30 - 14:50 Optimal Micro-Siting of Wind Farms by Particle Swarm

Optimization

P44

Chunqiu Wan, Jun Wang, Geng Yang and Xing Zhang

14:50 - 15:10 Radial Basis Function Neural Network Based on PSO with Mutation

Operation to Solve Function Approximation Problem

P45

Xiaoyong Liu

15:10 - 15:30 PSO Applied to Table Allocation Problems P45

David A. Braude and Anton van Wyk

Session PSO Algorithms (I) Chair Ben Niu

Date/Time June 13, 2010(Sunday) 13:30-15:10 Venue Room C

13:30 - 13:50 A New Particle Swarm Optimization Algorithm and Its Numerical

Analysis

P45

Yuelin Gao, Fanfan Lei and Miaomiao Wang

13:50 - 14:10 A New PSO Model Mimicking Bio-Parasitic Behavior P46

Quande Qin, Rongjun Li, Ben Niu and Li Li

14:10 - 14:30 Gender-Hierarchy Particle Swarm Optimizer Based on Punishment P46

Jiaquan Gao, Hao Li and Luoke Hu

14:30 - 14:50 An Improved Probability Particle Swarm Optimization Algorithm P47

Qiang Lu and Xuena Qiu

14:50 - 15:10 An Automatic Niching Particle Swarm for Multimodal Function

Optimization

P47

Yu Liu, Zhaofa Yan, Wentao Li, Mingwei Lv and Yuan Yao

26


Session PSO Algorithms (II) Chair Farrukh Khan

Date/Time June 13, 2010(Sunday) 15:50-17:30 Venue Room A

15:50 - 16:10 A Novel Encoding Scheme of PSO for Two-Machine Group

Scheduling

P47

Cheng-Dar Liou and Chun-Hung Liu

16:10 - 16:30 Improved Quantum Particle Swarm Optimization by Bloch Sphere P48

Yu Du, Haibin Duan, Renjie Liao and Xihua Li

16:30 - 16:50 A Discrete Particle Swarm Optimization with Self-adaptive Diversity

Control for the Blocking Permutation Flowshop

P48

Xianpeng Wang and Lixin Tang

16:50 - 17:10 An Improved Particle Swarm Optimization for Permutation

Flowshop Scheduling Problem with Total Flowtime Criterion

P49

Xianpeng Wang and Lixin Tang

17:10 - 17:30 Multi-Objective Particle Swarm Optimization for Optimal Security

of Networks

P49

Hamid Ali, Zulfiqar Ali and Farrukh Khan

Session ACO Algorithms Chair

Co-Chair

Muhammad Sharif

Bo Xing

Date/Time June 13, 2010(Sunday) 15:50-17:30 Venue Room B

15:50 - 16:10 Graph Partitioning using Improved Ant Clustering P50

M. Sami Soliman and Guanzheng Tan

16:10 - 16:30 An Improved Parallel Ant Colony Optimization Based On Message

Passing Interface

P50

Jie Xiong, Xiaohong Meng and Caiyun Liu

16:30 - 16:50 Two-Stage Inter-Cell Layout Design for Cellular Manufacturing by

Using Ant Colony Optimization Algorithms

P51

Bo Xing, Wen-jing Gao, Fulufhelo V. Nelwamondo, Kimberly Battle

and Tshilidzi Marwala

16:50 - 17:10 Ant n-Queen Solver P51

Salabat Khan, Mohsin Bilal, Muhammad Sharif and Rauf Baig

27

17:10 - 17:30 Multi-objective Optimization for Massive Pedestrian Evacuation

Using Ant Colony Algorithm

P52

Xinlu Zong, Shengwu Xiong, Zhixiang Fang and Qiuping Li

Session Novel Swarm-Based Optimization

Algorithms (I)

Chair

Co-Chair

Hongwei Mo

Yuanchun Zhu

Date/Time June 13, 2010(Sunday) 15:50-17:30 Venue Room C

15:50 - 16:10 Fireworks Algorithm for Optimization P52

Ying Tan and Yuanchun Zhu

16:10 - 16:30 A Scatter Search Algorithm for the Slab Stack Shuffling Problem P52

Xu Cheng and Lixin Tang

16:30 - 16:50 Collaboration Algorithm of FSMAS P53

Qingshan Li, Dan Jiang, Haishun Yun and He Liu

16:50 - 17:10 GPU-Based Parallelization Algorithm for 2D Line Integral

Convolution

P53

Bo Qin, Zhanbin Wu, Fang Su and Titi Pang

17:10 - 17:30 Biogeography Migration Algorithm for Traveling Salesman Problem P54

Hongwei Mo and Lifang Xu

28



Session Novel Swarm-Based Optimization

Algorithms (II)

Chair Shanchen Pang

Date/Time June 14, 2010(Monday) 08:00-09:20 Venue Room A

08:00 - 08:20 Novel Bacterial Foraging Optimization Based on Time-varying

Chemotaxis Step

P54

Ben Niu

08:20 - 08:40 RFID Network Scheduling with an Adaptive Bacterial Foraging P54

Hanning Chen, Yunlong Zhu and Kunyuan Hu

08:40 - 09:00 An Approach of Redistricting Based on Simple and Compactness P55

Shanchen Pang, Hua He, Yicheng Li, Tian Zhou and Kangzheng

Xing

09:00 - 09:20 SAR Image Segmentation based on Artificial Bee Colony Algorithm P55

Miao Ma, Jianhui Liang and Hongpeng Tian

Session Hybrid Algorithms Chair Zhaogeng Jiang

Date/Time June 14, 2010(Monday) 08:00-09:40 Venue Room B

08:00 - 08:20 Hybrid Particle Swarm and Conjugate Gradient Optimization

Algorithm

P56

Abdallah Qteish and Mohammad Hamdan

08:20 - 08:40 A Hybrid of Particle Swarm Optimization and Local Search for

Multimodal Functions

P56

Jin Qin, Yixin Yin and Xiaojuan Ban

08:40 - 09:00 A Cooperative Ant Colony System and Genetic Algorithm for TSPs P57

William W Guo and Gaifang Dong

09:00 - 09:20 Adaptive Particle Swarm Optimization Based Artificial Immune

Network Classification Algorithm

P57

Ruochen Liu, Manchun Niu, Lina Tang and Licheng Jiao

09:20 - 09:40 A Hybrid GA-PSO Approach with Gene Clustering for the Inference

of Gene Regulatory Networks

P58

Wei-Po Lee

29

Session Artificial Immune System Chair Ying Tan

Date/Time June 14, 2010(Monday) 08:00-09:40 Venue Room C

08:00 - 08:20 A Quantum Immune Algorithm for Multiobjective Parallel Machine

Scheduling

P58

Zhimin Fang

08:20 - 08:40 Cryptanalysis of Four-Rounded DES using Binary Artificial Immune

System

P59

Syed Hamdani, Sarah Shafiq and Farrukh Khan

08:40 - 09:00 A Immune Concentration Based Virus Detection Approach using

Particle Swarm Optimization

P59

Wei Wang, Pengtao Zhang and Ying Tan

09:00 - 09:20 Immune Algorithm with Memory Coevolution P60

Tao Liu

09:20 - 09:40 An Improved Immune Genetic Algorithm for Multiobjective

Optimization

P60

Guixia He, Jiaquan Gao and Luoke Hu

Session Multi-Agent Based Complex Systems Chair Xiujun Ma


10:00 - 10:20 Impulsive Consensus Seeking in Delayed Networks of Multi-Agents P61

Quanjun Wu, Lan Xiang and Jin Zhou

10:20 - 10:40 The Application of Multi-Agent Technology on the Level of Repair

Analysis

P61

Xiangkai Liu, Yanfeng Tang, Lin Zheng, Bingfeng Zhu and Jia-ning

Wang

10:40 - 11:00 The Framework of Intelligent Battlefield Damage Assessment System

Based on Multi-agent System

P61

Xiangkai Liu, Huimei Li, Jian Zhang, Jianing Wang and Wenhua

Xing

11:00 - 11:20 Adaptive System of Heterogeneous Multi-agent Investors in an

Articial Evolutionary Double Auction Market

P62

Chi Xu, Xiaoyu Zhao and Zheru Chi

30


11:20 - 11:40 Average Consensus for Directed Networks of Multi-agent with

Time-Varying Delay

P62

Tiecheng Zhang and Hui Yu

11:40 - 12:00 Multi- Cooperative Agent Reinforcement Learning in 3D Virtual

World

P63

Ping Zhang, Xiujun Ma, Zijian Pan, Xiong Li and Kunqing Xie

Session Multi-Robot Systems Chair Hongbo Wang


10:00 - 10:20 Enhanced Mapping of Multi-robot Using Distortion Reducing Filter

Based SIFT

P63

Kyung-Sik Choi, Yoon-Gu Kim, Jinung An and Suk-Gyu Lee

10:20 - 10:40 Development of Image Stabilization System Using Extended Kalman

Filter for a Mobile Robot

P64

Yun Won Choi, Tae Hoon Kang and Suk Gyu Lee

10:40 - 11:00 Swarm robot: from self-assembly and locomotion P64

Hongxing Wei, Yingpeng Cai and Tianmiao Wang

11:00 - 11:20 Diffusing Method for Unknown Environment Exploration in Multi

Robot Systems

P65

Dilshat Saitov, Ki Joon Han and Suk Gyu Lee

11:20 - 11:40 Localization and Full Coverage Path Planning for a Cleaning Robot P65

Hongbo Wang and Zhengwei Hu

11:40 - 12:00 A Novel Spatial Obstructed Distance by Dynamic Piecewise Linear

Chaotic Map and Dynamic Nonlinear PSO

P65

Xueping Zhang, Yawei Liu, Jiayao Wang and Haohua Du

Session Classifier Systems Chair Majid Ahmadi


10:00 - 10:20 SVM Classifier Based Feature Selection Using GA, ACO and PSO

for siRNA Design

P66

Yamuna Prasad, Kanad Biswas and Chakresh Jain

31

10:20 - 10:40 A Discrete-time Recurrent Neural Network for Solving Systems of

Complex-valued Linear Equations

P66

Wudai Liao, Jiangfeng Wang and Junyan Wang

10:40 - 11:00 A Recurrent Neural Network for Solving Complex-valued Quadratic

Programming Problems with Equality Constraints

P67

Wudai Liao, Jiangfeng Wang and Junyan Wang

11:00 - 11:20 Computer-Aided Detection and Classification of Masses in Digitized

Mammograms Using Artificial Neural Network

P67

Mohammed Islam, Majid Ahmadi and Maher A. Sid Ahmed

11:20 - 11:40 Gene Selection and PSO-BP Classifier Encoding a Prior Information P67

Yu Cui, Fei Han and Shiguang Ju

11:40 - 12:00 Object Recognition of a Mobile Robot based on SIFT with

De-speckle Filtering

P68

Zhiguang Xu, Kyung-Sik Choi, Yanyan Dai and Suk-Gyu Lee

Session Evolutionary Computation Chair

Co-Chair

Wen-Jyi Hwang

Marina Yusoff


13:30 - 13:50 Bottom-up Tree Evaluation In Tree-based Genetic Programming P68

Geng Li and XiaoJun Zeng

13:50 - 14:10 Solving Vehicle Assignment Problem Using Evolutionary

Computation

P69

Marina Yusoff, Junaidah Ariffin and Azlinah Mohamed

14:10 - 14:30 An Improved Thermodynamics Evolutionary Algorithm Based on

the Minimal Free Energy

P69

Fahong Yu, Yuanxiang Li and Weiqin Ying

14:30 - 14:50 A Non-dominated Sorting Bit Matrix Genetic Algorithm for P2P

Relay Optimization

P70

Qian He, Junliang Chen, Xiangwu Meng and Yanlei Shang

14:50 - 15:10 Fast Parallel Memetic Algorithm for Vector Quantization Based for

Reconfigurable Hardware and Softcore Processor

P70

Tsung-Yi Yu, Wen-Jyi Hwang and Tsung-Che Chiang

32


15:10 - 15:30 Integrate spatial information into multi-objective genetic algorithm

for spatial optimal location based on GIS

P70

Jinliang Hou, Haiqi Wang and Yujie Liu

Session Data Mining Methods Chair Jin Zhang


13:30 - 13:50 Intelligent Decision Support System for Breast Cancer P71

RR Janghel, Anupam Shukla, Ritu Tiwari and Rahul Kala

13:50 - 14:10 An Automatic Index Validity for Clustering P71

Zizhu Fan, Xiangang Jiang, Baogen Xu and Zhaofeng Jiang

14:10 - 14:30 A Novel Fast Non-negative Matrix Factorization Algorithm and Its

Application in Text Clustering

P72

Fang Li and Qunxiong Zhu

14:30 - 14:50 Application KIII Model to EEG Classification Based on Nonlinear

Dynamic Methods

P72

Jin Zhang

14:50 - 15:10 Brain-Computer Interface System Using Approximate Entropy and

EMD Techniques

P73

Qiwei Shi, Wei Zhou, Jianting Cao, Toshihisa Tanaka and Rubin

Wang

15:10 - 15:30 A Novel Spatial Clustering with Obstacles Constraints Based on

PNPSO and K-Medoids

P73

Xueping Zhang, Haohua Du, Tengfei Yang and Guangcai Zhao

Session Fuzzy Methods and Information

Processing System

Chair Rui Qin


13:30 - 13:50 Modeling Fuzzy Data Envelopment Analysis with Expectation

Criterion

P74

Xiaodong Dai, Ying Liu and Rui Qin

13:50 - 14:10 On Fuzzy Diagnosis Model of Plane’s Revolution Swing Fault and

Simulation Researches

P74

Dongcai Qu, Jihong Cheng, Wanli Dong and Ruizhi Zhang

33

14:10 - 14:30 Asymptotic Equivalent Analysis for LTI Overlapping Large-Scale

Systems and Their Subsystems

P74

Qian Wang and Xuebo Chen

14:30 - 14:50 An Application of LFP Method for Sintering Ore Ratio P75

Xi Cheng, Kailing Pan and Yunfeng Ma

14:50 - 15:10 A Class of Fuzzy Portfolio Optimization Problems: E-S Models P75

Yankui Liu and Xiaoli Wu

15:10 - 15:30 Matrix Estimation Based on Normal Vector of Hyperplane in Sparse

Component Analysis

P76

Feng Gao, Gongxian Sun, Ming Xiao and Jun Lv

Session Signal Processing and Information

Security

Chair Xiaohui Cui


15:50 - 16:10 Pricing and Bidding Strategy in AdWords Auction under

Heterogeneous Products Scenario

P76

E Zhang and Yiqin Zhuo

16:10 - 16:30 FIR Cutoff Frequency Calculating for ECG Signal Noise Removing

Using Artificial Neural Network

P76

Sara Moein

16:30 - 16:50 Force Identification by Using SVM and CPSO Technique P77

Zhichao Fu, Cheng Wei and Yanlong Yang

16:50 - 17:10 On the Strength Evaluation of Lesamnta Against Differential

Cryptanalysis

P77

Yasutaka Igarashi and Toshinobu Kaneko

17:10 - 17:30 Graphics Processing Unit Enhanced Parallel Document Flocking

Clustering

P78

Jesse St. Charles, Xiaohui Cui and Thomas Potok

17:30 - 17:50 Computational Intelligence Algorithms Analysis for Smart Grid

Cyber Security

P78

Yong Wang, Da Ruan, Jianping Xu, Mi Wen and Liwen Deng

34


Session Intelligent Control Chair

Co-Chair

Ding Feng

Viet-Hong Tran


15:50 - 16:10 The Automatic Feed Control Based on OBP Neural Network P79

Ding Feng, Bianyou Tan, Peng Wang, Shouyong Li, Jin Liu, Cheng

Yang, Yongxin Yuan and Guanjun Xu

16:10 - 16:30 GA-Based Integral Sliding Mode Control for AGC P79

Dianwei Qian, Xiangjie Liu, Miaomiao Ma and Chang Xu

16:30 - 16:50 A Distributed Energy-aware Trust Topology Control Algorithm for

Service-oriented Wireless Mesh Networks

P80

Chuanchuan You, Tong Wang, Bingyu Zhou, Hui Dai and Baolin

Sun

16:50 - 17:10 Leader-Follower Formation Control of Multi-Robots by using a

Stable Tracking Control Method

P80

Yanyan Dai, Viet-Hong Tran, Zhiguang Xu and Suk-Gyu Lee

17:10 - 17:30 Stable Swarm Formation Control Using Onboard Sensor Information P80

Viet-Hong Tran and Suk-Gyu Lee

Session Other Optimization Algorithms Chair Xiao Liu


15:50 - 16:10 Research on the Optimization Decision-making Two

Row-sequencing-pairs of Activities with Slacks

P81

Shisen Lv, Jianxun Qi, Xiuhua Zhao and Zhixiong Su

16:10 - 16:30 Sudoku Using Parallel Simulated Annealing P81

Zahra Karimi Dehkordi, Kamran Zamanifar and Ahmad Baraani

Dastjerdi

16:30 - 16:50 The Optimization of Procedure Chain of Three Activities with a

Relax Quantum

P81

Shisen Lv, Jianxun Qi and Xiuhua Zhao

16:50 - 17:10 A Quay Crane Scheduling Model in Container Terminals P82

Qi Tang

35

17:10 - 17:30 A Capacitated Production Planning Problem for Closed-loop Supply

Chain

P82

Jian Zhang and Xiao Liu

36


Posters


Session Poster Seesion 1 Chair Xi Huang

Date/Time June 13, 2010(Sunday)13:30-17:30 Venue Corridor

No.1 Fitness Function of Genetic Algorithm in Structural Constraint

Optimization

P82

Xinchi Yan and Xiaohan Wang

No.2 On the Correlations between Fuzzy Variables P83

Yankui Liu and Xing Zhang

No.3 Design and Implement of a Scheduling Strategy Based on PSO Algorithm P83

Suqin Liu, Jing Wang, Xingsheng Li, Jun Shuo and Huihui Liu

No.4 An Examination on Emergence from Social Behavior: A Case in

Information Retrieval

P84

Daren Li, Muyun Yang, Sheng Li and Tiejun Zhao

No.5 A Novel Fault Diagnosis Method Based on Modified Neural Networks

for Photovoltaic Systems

P84

Kuei-Hsiang Chao, Chao-Ting Chen, Meng-Hui Wang and Chun-Fu Wu

No.6 Wavelet Packet and Generalized Gaussian Density Based Textile Pattern

Classification Using BP Neural Network

P85

Yean Yin, Liang Zhang, Miao Jin and Sunyi Xie

No.7 Air Quality Prediction in Yinchuan by Using Artificial Neural Networks P85

Fengjun Li

No.8 Application of Short-term Load Forecasting Based on Improved

Gray-Markov Residuals Amending of BP Neural Network

P85

Dongxiao Niu, Cong Xu, Jianqing Li and Yanan Wei

No.9 Verifying Election Campaign Optimization Algorithm by Several

Benchmarking Functions

P86

Wenge Lv, Qinghua Xie, Zhiyong Liu, Deyuan Li, Siyuan Cheng,

Shaoming Luo and Xiangwei Zhang

37

No.10 Research on Multi-objective Optimization Design of the UUV Shape

Based on Numerical Simulation

P86

Baowei Song, Qifeng Zhu and Zhanyi Liu

No.11 A Traffic Video Background Extraction Algorithm Based on Image

Content Sensitivity

P87

Bo Qin, Jingjing Wang, Jian Gao, Titi Pang and Fang Su

No.12 A Multimodality Medical Image Fusion Algorithm Based on Wavelet

Transform

P87

Jionghua Teng, Xue Wang and Jingzhou Zhang

No.13 Adjusting the Clustering Results Referencing an External Set P88

Baojia Li, Yongqian Liu and Mingzhu Liu

38



Session Poster Seesion 2 Chair Xi Huang

Date/Time June 14, 2010(Monday)08:00-12:00 Venue Corridor

No.1 Sensitivity Analysis on Single Activity to Network Float in CPM

Network Planning

P88

Zhixiong Su and Jianxun Qi

No.2 Research on Hand Language Video Retrieval P88

Shilin Zhang and Mei Gu

No.3 Research on the Synergy Model between Knowledge Capital and

Regional Economic Development

P89

Cisheng Wu and Meng Song

No.4 A New Algorithm of an Improved Detection of Moving Vehicles P89

Huanglin Zeng and Zhenya Wang

No.5 The Dual Model of a Repairable System P89

Yunfei Guo, Maosheng Lai and Zhe Yin

No.6 A Comprehensive Study of Neutral-point-clamped Voltage Source PWM

Rectifiers

P90

Guojun Tan, Zongbin Ye, Yuan Li, Yaofei Han and Wei Jing

No.7 FPGA-Based Cooling Fan Control System for Automobile Engine P90

Meihua Xu, Fangjie Zhao and Lianzhou Wang

No.8 Fault Diagnosis of Analog Circuits Using Extension Genetic Algorithm P90

Meng-Hui Wang, Kuei-Hsiang Chao and Yu-Kuo Chung

No.9 Research on the Coordination Control of Vehicle Electrical Power

Steering System and ABS

P91

Weihua Qin, Qidong Wang, Wuwei Chen and Shenghui Pan

No.10 Optimization Algorithm of Scheduling Six Parallel Activities to Three

Pairs Order Activities

P91

Xiuhua Zhao, Jianxun Qi, Shisen Lv and Zhixiong Su

39

No.11 Bacterial Foraging Optimization Algorithm with Particle Swarm

Optimization Strategy for Distribution Network Reconfiguration

P92

Tianlei Zang, Zhengyou He and Deyi Ye

No.12 Tracking Control of Uncertain DC Server Motors Using Genetic Fuzzy

System

P92

Wei-Min Hsieh, Yih-Guang Leu, Hao-Cheng Yang and Jian-You Lin

40


Abstracts

Theoretical Analysis of Swarm Intelligence Algorithms

June 13, 2010(Sunday) 13:30-15:10 Room A

The Performance Measurement of a Canonical Particle Swarm

Optimizer with Diversive Curiosity

Hong Zhang1 and Jie Zhang2

1Kyushu Institute of Technology, Japan

2Wuxi Bowen Software Technology Co., Ltd, China

Abstract. For improving the search performance of a canonical particle swarm optimizer(CPSO),

we propose a newly canonical particle swarm optimizer with diversive curiosity (CPSO/DC). A

crucial idea here is to introduce diversive curiosity into the CPSO to comprehensively manage

the trade-off between exploitation and exploration for all eviating stagnation.To demonstrate

the effectiveness of the proposed method,computer experiments on a suite of five-dimensional

benchmark problems are carried out. We investigate the characteristics of the CPSO/DC, and

compare the search performance with other methods. The obtained results indicate that the

search performance of the CPSO/DC is superior to that by EPSO, ECPSO and RGA/E, but

is inferior to that by PSO/DC for the Griewank and Rastrigin problems.

On the Farther Analysis of Performance of the Artificial

Searching Swarm Algorithm

Tanggong Chen1, Lijie Zhang1 and Lingling Pang1

1Hebei University of Technology, China

Abstract. . Artificial Searching Swarm Algorithm (ASSA) is an intelligent optimization

algorithm , and its performance has been analyzed and compared with some famous algorithms.

For farther understanding the running principle of ASSA, this work discusses the function s

of three behavior rules which decide the moves of searching swarm. Some typical functions

are selected to do the simulation tests. The function simulation tests showed that the three

behavior rules are indispensability and endow the ASSA with powerful global optimization

ability together.

Back-Propagation vs Particle Swarm Optimization Algorithm:

which Algorithm is better to adjust the Synaptic Weights of a

Feed-Forward ANN?

Beatriz Aurora Garro Licon1, Humberto Sossa Azuela1 and Roberto Antonio Vazquez1

1National Polytechnic Institute, Mexico

41

Abstract. Bio-inspired algorithms have shown their usefulness in different non-linear optimization

problems. Due to their efficiency and adaptability, these algorithms have been applied to a wide

range of problems. In this paper we compare two ways of training an artificial neural network

(ANN): Particle Swarm Optimization (PSO) algorithms against classical training algorithms

such as: back-propagation and Levenberg Marquardt method. The main contribution of this

paper is to answer the next question: is PSO really better than classical training algorithms in

adjusting the synaptic weights of an ANN? First of all, we explain how the ANN training phase

could be seen as an optimization problem. Then, it is explained how PSO could be applied to

find the best synaptic weights of the ANN. Finally we perform a comparison among an ANN

trained with different classical methods and the PSO approach applied to different non-linear

problems, and also applied to a real object recognition problem.

Orthogonality and Optimality in Non-Pheromone Mediated

Foraging

Sanza Kazadi1, James Yang1, James Park1 and Andrew Park1

1Jisan Research Institute, USA

Abstract. We describe the general foraging task, breaking it into two dierent subtasks:

map-making and collection. Map-making is a task in which a map is constructed which

contains the location(s) of an item or of items in the search area. Collection is the task in

which an item is picked up and carried back to a central known location. We theoretically

examine these tasks, generating minimal conditions for each one to be accomplished. We then

build a swarm made up of two castes to accomplish this, theoretically motivating the design of

the swarm. Finally, we demonstrate that the swarm is optimal in the class of swarms utilizing

line-of-sight communication, and give performance measures for open and closed search spaces

An Adaptive Staged PSO Based on Particles’s Search

Capabilities

Kun Liu1 and Ying Tan1

1Peking University, China

Abstract. This study proposes an adaptive staged particle swarm optimization (ASPSO)

algorithm based on analyses of particles’s search capabilities. First, the search processes of the

standard PSO (SPSO) and the linear decreasing inertia weight PSO (LDWPSO) are analyzed

based on our previous definition of exploitation. Second, three stages of the search process

in PSO are defined. Each stage has its own search preference, which is represented by the

42


exploitation capability of swarm. Third, the mapping between inertia weight, learning factor

(w-c) and the exploitation capability is given. At last, the ASPSO is proposed. By setting

different values of w-c in three stages, one can make swarm search the space with particular

strategy in each stage, and the particles can be directed to find the solution more effectively.

The experimental results show that the proposed ASPSO has better performance than SPSO

and LDWPSO on most of test functions.

Applications of PSO Algorithms

June 13, 2010(Sunday) 13:30-15:30 Room B

Medical Image Registration Based on Generalized Mutual

Information and PSO Algorithm

Jingzhou Zhang1, Pengfei Huo1, Jionghua Teng1, Xue Wang and Suhuan Wang

1Northwestern Polytechnical University, China

Abstract. The medical image registration algorithm uses the mutual information measure

function that has many local extremes. Therefore, we propose our medical image registration

algorithm that combines generalized mutual information with PSO-Powell hybrid algorithm

and uses the objective measure function based on Renyi entropy. The Renyi entropy can

remove the local extremes. We use the particle swarm optimization (PSO) algorithm to locate

the measure function near the local extremes. Then we take the local extremes as initial point

and use the Powell optimization algorithm to search for the global optimal solution. Section

2.2 of the paper presents the six-step procedure of our registration algorithm. We simulate

medical image data with the registration algorithm; the simulation results, given in Table. 2

and 3, show preliminarily that the registration algorithm can eliminate the local extremes of

objective measure function and accelerate the convergence rate, thus obtaining accurate and

better registration results.

Particle Swarm Optimization For Automatic Selection of

Relevance Feedback Heuristics

Peng-Yeng Yin1

1National Chi-Nan University, Taiwan, China

Abstract. Relevance feedback (RF) is an iterative process which refines the retrievals by

utilizing user’s feedback marked on retrieved results. Recent research has focused on the

optimization for RF heuristic selection. In this paper, we propose an automatic RF heuristic

selection framework which automatically chooses the best RF heuristic for the given query.

The proposed method performs two learning tasks: query optimization and heuristic-selection

43

optimization. The particle swarm optimization (PSO) paradigm is applied to assist the learning

tasks. Experimental results tested on a content-based retrieval system with a real-world image

database reveal that the proposed method outperforms several existing RF approaches using

different techniques. The convergence behavior of the proposed method is empirically analyzed.

A New Particle Swarm Optimization Solution to Nonconvex

Economic Dispatch Problem

Jianhua Zhang1, Yingxin Wang1, Rui Wang1 and Guolian Hou1

1North China Electric Power University, China

Abstract. This paper presents an optimal economic dispatch for power plants by using

modified particle swarm optimization (PSO) algorithm. The economic dispatch problem in

power systems is to determine the optimal combination of power outputs for all generating units

in order that the total fuel cost can be minimized, furthermore, all practical constraints can be

satisfied. Several key factors in terms of valve-point effects of coal cost functions, unit operation

constraints and power balance are considered in the computation models. Consequently, a new

adaptive PSO technique is utilized for solving economic dispatch problems. The proposed

algorithm is compared with other PSO algorithms. Simulation results show that the proposed

method is feasible and efficient.

Optimal Micro-Siting of Wind Farms by Particle Swarm

Optimization

Chunqiu Wan1, Jun Wang2, Geng Yang1 and Xing Zhang1

1Tsinghua University, China

2Tongji University, China

Abstract. This paper proposes a novel approach to optimal placement of wind turbines in

the continuous space of a wind farm. The control objective is to maximize the power produced

by a farm with a fixed number of turbines while guaranteeing the distance between turbines

no less than the allowed minimal distance for turbine operation safety. The problem of wind

farm micro-siting with space constraints is formulated to a constrained optimization problem

and solved by a particle swarm optimization (PSO) algorithm based on penalty functions.

Simulation results demonstrate that the PSO approach is more suitable and effective for

micro-siting than the classical binary-coded genetic algorithms.

44


Radial Basis Function Neural Network Based on PSO with

Mutation Operation to Solve Function Approximation Problem

Xiaoyong Liu1,2

1Guangdong Polytechnic Normal University, China

2Chinese Academy of Sciences, China

Abstract. This paper presents a novel learning algorithm for training and constructing a

Radial Basis Function Neural Network (RBFNN), called MuPSO-RBFNN algorithm. This

algorithm combines Particle Swarm Optimization algorithm (PSO) with mutation operation

to train RBFNN. PSO with mutation operation and genetic algorithm are respectively used

to train weights and spreads of oRBFNN, which is traditional RBFNN with gradient learning

in this article. Sum Square Error (SSE) function is used to evaluate performance of three

algorithms, oRBFNN, GA-RBFNN and MuPSO-RBFNN algorithms. Several experiments in

function approximation show MuPSO-RBFNN is better than oRBFNN and GA-RBFNN.

PSO Applied to Table Allocation Problems

David A. Braude1 and Anton van Wyk1

1University of the Witwatersrand, South Africa

Abstract. Table allocation is a type of assignment problem. The aim of table allocation is

to assign multiple people to a single table in such a way that it minimizes a cost function.

While particle swarm optimization (PSO) is normally used for continuous variables it has been

adapted to solve this problem. Each particle represents an entire seating arrangement, and

the velocity is the amount of times people swap tables during each iteration. In an example

application PSO shows a signicant improvement in fitness compared to the initial conditions,

and has a low runtime. It also performs better in fitness improvement and runtime compared

to choosing as many random samples as PSO generated. The use of PSO allows for generalized

cost functions, and is simple to implement.

PSO Algorithms (I)

June 13, 2010(Sunday) 13:30-15:10 Room C

A New Particle Swarm Optimization Algorithm and Its

Numerical Analysis

Yuelin Gao1, Fanfan Lei and Miaomiao Wang

1North Ethnic University, China

Abstract. The speed equation of particle swarm optimization is improved by using a convex

45

combination of the current best position of a particle and the current best position which the

whole particle swarm as well as the current position of the particle, so as to enhance global

search capability of basic particle swarm optimization. Thus a new particle swarm optimization

algorithm is proposed. Numerical experiments show that its computing time is short and its

global search capability is powerful as well as its computing accuracy is high in compared with

the basic PSO.

A New PSO Model Mimicking Bio-Parasitic Behavior

Quande Qin1, Rongjun Li1, Ben Niu2 and Li Li2

1South China University of Technology, China

2Shenzhen University, China

Abstract. Based on the analysis of biological symbiotic relationship, the mechanism of

facultative parasitic behaviour is embedded into the particle swarm optimization (PSO) to

construct a two-population PSO model called PSOPB, composed of the host and the parasites

population. In this model, the two populations exchange particles according to the fitness

sorted in a certain number of iterations. In order to embody the law of ”survival of the

fittest” in biological evolution, the poor fitness particles in the host population are eliminated,

replaced by the re-initialization of the particles in order to maintain constant population size.

The results of experiments of a set of 6 benchmark functions show that presented algorithm

model has faster convergence rate and higher search accuracy compared with CPSO, PSOPC

and PSO-LIW.

Gender-Hierarchy Particle Swarm Optimizer Based on

Punishment

Jiaquan Gao1, Hao Li1 and Luoke Hu

1Zhejiang University of Technology, China

Abstract. The paper presents a novel particle swarm optimizer (PSO), called gender-hierarchy

particle swarm optimizer based on punishment (GH-PSO). In the proposed algorithm, the

social part and recognition part of PSO both are modified in order to accelerate the convergence

and improve the accuracy of the optimal solution. Especially, a novel recognition approach,

called general recognition, is presented to furthermore improve the performance of PSO.

Experimental results show that the proposed algorithm shows better behaviors as compared

to the standard PSO, tribes-based PSO and GH-PSO with tribes.

46


An Improved Probability Particle Swarm Optimization

Algorithm

Qiang Lu1 and Xuena Qiu

1Hangzhou Dianzi University, China

Abstract. This paper deals with the problem of unconstrained optimization. An improved

probability particle swarm optimization algorithm is proposed. Firstly, two normal distributions

are used to describe the distributions of particle positions,respectively.One is the normal

distribution with the global best position as mean value and the diffrence between the current

fitness and the global best fitness as standard deviation while another is the distribution with

the previous best position as mean value and the difference between the current fitness and

the previous best fitness as standard deviation. Secondly, a disturbance on the mean values

is introduced into the proposed algorithm. Thirdly,the nal position of particles is determined

by employing a linear weighted method to cope with the sampled information from the two

normal distributions. Finally, benchmark functions are used to illustrate the effectiveness of

the proposed algorithm.

An Automatic Niching Particle Swarm for Multimodal Function

Optimization

Yu Liu1, Zhaofa Yan, Wentao Li1, Mingwei Lv and Yuan Yao

1Dalian University of Technology, China

Abstract. Niching is an important technique for mutlimodal optimization. This paper

proposed an improved niching technique based on particle swarm optimizer to locate multiple

optima. In the proposed algorithm, the algorithm inspired from natural ecosystem form niches

automatically without any prespecified problem dependent parameters. Experiment results

demonstrated that the proposed nich ing method is superior to the classic niching methods

which are with or without niching parameters

PSO Algorithms (II)

June 13, 2010(Sunday) 15:50-17:30 Room A

A Novel Encoding Scheme of PSO for Two-Machine Group

Scheduling

Cheng-Dar Liou1 and Chun-Hung Liu1

1National Formosa University, Taiwan, China

Abstract. This paper investigates the two-machine flow shop group scheduling problem with

47

the transportation times and sequence-dependent setup times considerations. The objective is

to minimize the total completion time. In this paper, a novel encoding scheme of PSO for flow

shop group scheduling is proposed to effectively solve various instances with group numbers

up to 15. Note that the proposed encoding scheme simultaneously determines the sequence of

jobs in each group and the sequence of groups. Three different lower bounds are developed

to evaluate the performance of the proposed PSO algorithm. Limited numerical results show

that the proposed PSO algorithm performs well for all test problems.

Improved Quantum Particle Swarm Optimization by Bloch

Sphere

Yu Du1, Haibin Duan1,2, Renjie Liao1 and Xihua Li1

1Beihang University, China

2Suzhou University, China

Abstract. Quantum Particle Swarm Optimization (QPSO) is a global convergence guaranteed

search method which introduces the Quantum theory into the basic Particle Swarm Optimization

(PSO). QPSO performs better than normal PSO on several benchmark problems. However,

QPSO’s quantum bit(Qubit) is still in Hilbert space’s unit circle with only one variable, so the

quantum properties have been undermined to a large extent. In this paper, the Bloch Sphere

encoding mechanism is adopted into QPSO, which can vividly describe the dynamic behavior

of the quantum. In this way, the diversity of the swarm can be increased, and the local minima

can be effectively avoided. The proposed algorithm, named Bloch QPSO (BQPSO), is tested

with PID controller parameters optimization problem. Experimental results demonstrate that

BQPSO has both stronger global search capability and faster convergence speed, and it is

feasible and effective in solving some complex optimization problems.

A Discrete Particle Swarm Optimization with Self-adaptive

Diversity Control for the Blocking Permutation Flowshop

Xianpeng Wang1 and Lixin Tang1

1Northeastern University, China

Abstract. This paper deals with the m-machine permutation flowshop scheduling problem

with blocking (Fm/blocking/Cmax) to minimize the makespan, which has a strong industrial

background, e.g., many production processes of chemicals and pharmaceuticals in chemical

industry can be reduced to this problem, and proposes a discrete particle swarm optimization

(DPSO) algorithm. To prevent the DPSO from premature, a self-adaptive diversity control

strategy is adopted, i.e. a random perturbation is added to the velocity equation of each

48


particle according to a probability that is controlled by the diversity of the current population,

to diversify the population when necessary. Besides this, when a particle is considered to be

trapped in local optimum, it will be replaced by a new particle generated from a reference set,

which is borrowed from the scatter search (SS). In addition, a local search, called stochastic

variable neighborhood search, is embedded in the DPSO algorithm to further improve the

search intensification. Computational results using benchmark problems show that the proposed

DPSO algorithm is relatively more effective than other approaches proposed for this problem.

An Improved Particle Swarm Optimization for Permutation

Flowshop Scheduling Problem with Total Flowtime Criterion

Xianpeng Wang and Lixin Tang1


Abstract. This paper deals with the m-machine permutation flowshop scheduling problem

to minimize the total flowtime, an NP-complete problem, and proposes an improved particle

swarm optimization (PSO) algorithm. To enhance the exploitation ability of PSO, a stochastic

iterated local search is incorporated. To improve the exploration ability of PSO, a population

update method is applied to replace non-promising particles. In addition, a solution pool

that stores elite solutions found in the search history is adopted, and in the evolution process

each particle learns from this solution pool besides its personal best solution and the global

best solution so as to improve the learning capability of the particles. Experimental results

on benchmark instances show that the proposed PSO algorithm is competitive with other

metaheuristics.

Multi-Objective Particle Swarm Optimization for Optimal

Security of Networks

Hamid Ali1, Zulfiqar Ali1 and Farrukh Khan1

1FAST National University of Computer and Emerging Sciences (NUCES), Pakistan

Abstract. To provide security and make the networking system more reliable, a number

of efforts have been made by researchers for the last several years. Though many successful

security systems have been designed and implemented, a number of issues such as the time

required for designing a secure system, cost, minimizing the damage and maintenance still need

to be resolved. Designing the security system harder and avoiding the unauthorized access with

a low cost simultaneously is a challenging task. Targeting such a multi-objective scenario, a few

approaches have been applied previously to optimize the cost and the damage. In this paper, we

introduce the Multi-Objective Particle Swarm Optimization (MOPSO) technique to optimize

49

the cost and the residual damage for a network security infrastructure. The experiments show

very promising results when the proposed system is compared with the previously proposed

technique based on Non-dominated Sorting Genetic Algorithm II (NSGA-II).

ACO Algorithms

June 13, 2010(Sunday) 15:50-17:30 Room B

Graph Partitioning using Improved Ant Clustering

M. Sami Soliman1 and Guanzheng Tan1

1Central South University, China

Abstract. Parallel computing, network partitioning, and VLSI circuit placement are fundamental

challenges in computer science. These problems can be modeled as graph partitioning problems.

A new Similarity carrying Ant Model (SCAM) is used in the ant-based clustering algorithm to

solve graph partitioning problem. In the proposed model, the ant will be able to collect similar

items while it moves around. The flexible template mechanism had been used integrated with

the proposed model to obtain the partitioning constrains. Random graph has been used to

compare the new model with the original ant model and the model with short-term memory.

The result of the experiments proves the impact of the SCAM compared with other models.

This performance improvement for ant clustering algorithm makes it is feasible to be used in

graph porti oning problem.

An Improved Parallel Ant Colony Optimization Based On

Message Passing Interface

Jie Xiong1, Xiaohong Meng1 and Caiyun Liu2

1China University of Geosciences, China

2Yangtze University, China

Abstract. Ant Colony Optimization (ACO) is recently proposed metaheuristic approach for

solving hard combinatorial optimization problems. Parallel implementation of ACO can reduce

the computational time obviously. An improved parallel ACO algorithm is proposed in this

paper, which use dynamic transition probability to enlarge the search space by stimulating ants

choosing new path at early stage; use polymorphic ant colony to improve convergence speed by

local search and global search; use partially asynchronous parallel implementation, interactive

multi-colony parallel and new information exchange strategy to improve the parallel efficiency.

We implement the algorithm on the Dawn 4000L parallel computer using MPI and C language.

The Numerical result indicates the algorithm proposed in this paper can improve convergence

speed effectively with the fine solution quality.

50


Two-Stage Inter-Cell Layout Design for Cellular Manufacturing

by Using Ant Colony Optimization Algorithms

Bo Xing1, Wen-jing Gao, Fulufhelo V. Nelwamondo, Kimberly Battle and Tshilidzi

Marwala

1University of Johannesburg, South Africa

Abstract. Facility layout planning plays an important role in the manufacturing process and

seriously impacts a company’s profitability. A well-planned layout can significantly reduce the

total material handling cost. The purpose of this paper is to develop a two-stage inter-cell

layout optimization approach by using one of the popular meta-heuristics: the Ant Colony

Optimization algorithm. At the first stage, the cells are formed based on the part-machine

clustering results obtained through the ant system algorithm. In other words, we get the

initial inter-cell layout after this stage. The work at the second stage uses a hybrid ant system

algorithm to improve the solution obtained at previous stage. Different performance measures

are also employed in this paper to evaluate the results.

Ant n-Queen Solver

Salabat Khan1, Mohsin Bilal1, Muhammad Sharif1 and Rauf Baig1

1National University of Computer and Emerging Sciences, Pakistan

Abstract. Swarm intelligence and evolutionary techniques are heavily used by the researchers

to solve combinatorial and NP hard problems. The n-Queen problem is a combinatorial

problem which become intractable for large values of n and thus placed in NP (Non-Deterministic

Polynomial) class problem. In this paper, a solution is proposed for n-Queen problem based

on ACO (Ant Colony Optimization). The n-Queen problem is basically a generalized form

of 8-Queen problem. In 8-Queen problem, the goal is to place eight queens such that no

queen can kill the other using standard chess queen moves. The environment for the ants

is a directed graph which we call search space is constructed for efficiently searching the

valid placement of n-queens such that they do not harm each other. We also develop an

intelligent heuristic function that helps in finding the solution very quickly and effectively.

The paper contains the detail discussion of problem background, problem complexity, Ant

Colony Optimization (Swarm Intelligence), proposed technique design and architecture and a

fair amount of experimental results.

51

Multi-objective Optimization for Massive Pedestrian Evacuation

Using Ant Colony Algorithm

Xinlu Zong1, Shengwu Xiong1, Zhixiang Fang and Qiuping Li

1Wuhan University of Technology, China

Abstract. Evacuation route planning is one of the most crucial tasks for solving massive

evacuation problem. In large public places , pedestrians should be transferred to safe areas

when nature or man-made accidents happen. A multi-objective ant colony algorithm for

massive pedestrian evacuation is presented in this paper. In the algorithm, three objectives,

total evacuation time of all evacuees, total routes risk degree and total crowding degree

are minimized simultaneously. Ants search routes and converge toward the Pareto optimal

solutions in the light of the pheromone. The experimental results show that the approach is

efficient and effective to solve massive evacuation problem with rapid, reasonable and safe plans.

Novel Swarm-Based Optimization Algorithms (I)

June 13, 2010(Sunday) 15:50-17:30 Room C

Fireworks Algorithm for Optimization

Ying Tan1 and Yuanchun Zhu1


Abstract. Inspired by observing fireworks explosion, a novel swarm intelligence algorithm,

called Fireworks Algorithm (FA), is proposed for global optimization of complex functions. In

the proposed FA, two types of explosion (search) processes are employed, and the mechanisms

for keeping diversity of sparks are also well designed. In order to demonstrate the validation of

the FA, a number of experiments were conducted on nine benchmark test functions to compare

the FA with two variants of particle swarm optimization (PSO) algorithms, namely Standard

PSO and Clonal PSO. It turns out from the results that the proposed FA clearly outperforms

the two variants of the PSOs in both convergence speed and global solution accuracy.

A Scatter Search Algorithm for the Slab Stack Shuffling Problem

Xu Cheng1 and Lixin Tang1


Abstract. Slab Stack Shuffling (SSS) problem is a kind of warehousing operations management

problem abstracted from steel industry. SSS problem is to choose appropriate slabs for hot

rolling schedule with the objective of minimizing shuffles during the retrieval process. Different

from previous literatures, the substitute of slabs is a set of slabs which satisfy the given order

52


demand. The problem in this paper considers balancing the shuffles between two sub-yards

and the measurement of one shuffle is also different. The problem is formulated as an integer

programming model by considering above practical requirements. The complexity of the

model motivated us to develop a scatter search algorithm to solve the problem approximately.

Problem-oriented coding scheme and solution combination method are proposed in scatter

search. The computational results tested on real data show that the shuffles are decreased by

36.9% in average compared with the manual schedule.

Collaboration Algorithm of FSMAS

Qingshan Li1, Dan Jiang1, Haishun Yun1 and He Liu1

1Xidian University, China

Abstract. To meet the requirement and solve the problems in system integration field, A

Federation Structure based Multi-Agent System (FSMAS) model is proposed in this paper,

with emphasis on the collaboration algorithm. This paper presents the process of partition

and collaboration of the Agent tasks, the acquaintance first based on CNP algorithm in

collaboration. FSMAS is applied to the development of agent-based system integration platform

and tools . As a test case, a simulation system is developed which verifies the stability and

effici ency of FSMAS in system integration filed.

GPU-Based Parallelization Algorithm for 2D Line Integral

Convolution

Bo Qin1, Zhanbin Wu1, Fang Su1 and Titi Pang1

1Ocean University of China, China

Abstract. GPU (Graphics Processing Unit) technology pro vides an efficient method for

parallel computation. This paper will present a GPU - based Line Integral Convolution (LIC)

parallel algorithm for visualization of discrete vector fields to accelerate LIC algorithm. The

algorithm is implemented with parallel o perations using Compute Unified Device Architecture

(CUDA) programming model in GPU. The method can provide up to about 50× speed - up

without any sacrifice on solution quality, compared to conventional sequential computation.

Experiment results show that it is useful for in - time remote visualization of discrete vector

fields .

53

Biogeography Migration Algorithm for Traveling Salesman

Problem

Hongwei Mo1 and Lifang Xu1

1Harbin Engineering University, China

Abstract. Biogeography-based optimization algorithm(BBO) is a new kind of optimization

algorithm based on Biogeography. It is designed based on the migration strategy of animals

to solve the problem of optimization. In this paper, a new algorithm-Biogeography Migration

Algorithm for Traveling Salesman Problem(TSPBMA) is presented. Migration operator is

designed. It is tested on four classical TSP problems. The comparison results with the

other nature inspired optimization algorithms show that TSPBMA is a very effective for TSP

combination optimization. It provides a new way for this kinds of problem.

Novel Swarm-Based Optimization Algorithms (II)

June 14, 2010(Monday) 08:00-09:20 Room A

Novel Bacterial Foraging Optimization Based on Time-varying

Chemotaxis Step

Ben Niu1,2

1Shenzhen Univeristy, China

2The Univeristy of Hongkong, Hongkong, China

Abstract. Recently, bacterial foraging optimizer (BFO) has emerged as a powerful technique

for optimization problem solving. However, various simulation results obtained from previous

studies suggested that the performance of BFO depends heavily on the chemotaxis step length

in in silico study of the optimization problem. In this paper, two modifications were proposed

to introduce a linear variation and a nonlinear variation of chemotaxis step in order to improve

the speed of convergence as well as fine tune the search in the multidimensional space. To

illustrate the efficiency of the proposed algorithms (BFO-LDC and BFO-NDC), eight different

benchmark functions were selected as testing functions to compare with original BFO and

GA. Analysis of variance (ANOVA) test was also carried out to validate the efficacy of the

proposed algorithms. Results of the comparison indicated that two proposed algorithms

generally outperform classical BFO and GA in all the benchmark functions.

RFID Network Scheduling with an Adaptive Bacterial Foraging

Hanning Chen1, Yunlong Zhu1 and Kunyuan Hu1

1Chinese Academy of Sciences, China

54


Abstract. This paper proposes a novel bacterial colony foraging (BCF) algorithm for complex

optimization problems. The proposed BCF extend original bacterial foraging algorithm to

adaptive and cooperative mode by combining bacterial chemotaxis, cell-to-cell communication,

and a self-adaptive foraging strategy. The cell-to-cell communication enables the historical

search experience sharing among the bacterial colony that can significantly improve convergence.

With the self-adaptive strategy, each bacterium can be characterized by focused and deeper

exploitation of the promising regions and wider exploration of other regions of the search space.

A rigorous performance analysis is given where the proposed algorithm is benchmarked against

four state-of-the-art reference algorithms using both a classical and a composition test function

suites. The individual and collective bacterial foraging behaviors of the proposed algorithmic

model are also studied. Lastly, the proposed algorithm is applied to a real-world application

of dynamic RFID network optimization. Statistical analysis of all these tests highlights the

significant performance improvement due to the beneficial combination and shows that the

proposed algorithm outperforms the reference algorithms.

An Approach of Redistricting Based on Simple and Compactness

Shanchen Pang1, Hua He1, Yicheng Li2, Tian Zhou1 and Kangzheng Xing3

1Shandong University of Science and Technology, China

2State University of New York, USA

3College of Sciences, Shanghai University, China

Abstract. Redistricting is a Clustering Problem in optimization. The optimum redistricting

is a convincing argument to voters that this solution is fair. In this paper, we set up a kind of

model basing on the multi-factor model of clustering of the population pots by adopting the

theory of optimization and the tools of stochastic simulation. Through this method, we solve

the problem of how to realize the redistribution and the judging problem. Using the statistical

data and practical model, we can get the districts of the state of New York satisfied with the

rules of all principles.

SAR Image Segmentation based on Artificial Bee Colony

Algorithm

Miao Ma1, Jianhui Liang1 and Hongpeng Tian

1Shaanxi Normal University, China

Abstract. Technology of parallel searching threshold is an effective method to fasten the

speed of image segmentation. This paper proposes a fast Synthetic Aperture Radar (SAR)

image segmentation method, which makes full use of wavelet transform, grey entropy model

55

and Artificial Bee Colony (ABC) algorithm. In the method, a filtered image is produced

by performing a noise reduction to the approximation image reconstructed by low-frequency

coefficients in wavelet domain. At the same time, a gradient image is produced by reconstructing

high-frequency coefficients in wavelet domain. Hence, a co-occurrence matrix based on the

filtered image and the gradient image is constructed. And then, a grey entropy model is

improved to act as the fitness function of ABC algorithm. Finally, swarm intelligence of

employed bees, onlookers and scouts is used to locate the optimal threshold quickly. Experimental

results indicate that the method is superior to Genetic Algorithm (GA) or Artificial Fish Swarm

(AFS) algorithm based methods.

Hybrid Algorithms

June 14, 2010(Monday) 08:00-09:40 Room B

Hybrid Particle Swarm and Conjugate Gradient Optimization

Algorithm

Abdallah Qteish1 and Mohammad Hamdan1

1Yarmouk University, Jordan

Abstract. In this work we propose a different particle swarm optimization (PSO) algorithm

that employs two key features of the conjugate gradient (CG) method. Namely, adaptive

weight factor for each particle and iteration number (calculated as in the CG approach), and

periodic restart. Experimental results for four well known test problems have showed the

superiority of the new PSO-CG approach, compared with the classical PSO algorithm, in

terms of convergence speed and quality of obtained solutions

A Hybrid of Particle Swarm Optimization and Local Search for

Multimodal Functions

Jin Qin1,2, Yixin Yin1 and Xiaojuan Ban1

1University of Science & Technology, China

2Guizhou University, China

Abstract. The standard PSO has problems with consistently converging to good solutions,

especially for multimodal functions. The reason for PSO failing to find (global) optima is

premature convergence. Also, it has been shown in many empirical studies that PSO algorithms

lack exploitation abilities. In this paper, we propose a hybrid of particle swarm optimization

and local search, in which a standard PSO algorithm incorporates a local search algorithm.

The standard PSO algorithm and the local search algorithm are devoted to exploration and

exploitation of solution space, respectively. Particles current position is updated using update

56


equation of standard PSO and then is refined by local search algorithm. The introduction

of a local search improves the capability of exploitation of local region of standard PSO and

prevents from premature convergence. The hybrid algorithm can locate multiple solutions

without use of specific niching techniques. The hybrid algorithm showed superior performance

on a set of multimodal functions.

A Cooperative Ant Colony System and Genetic Algorithm for

TSPs

William W Guo1 and Gaifang Dong

1Central Queensland University, Australia

Abstract. The travelling salesman problem (TSP) is a classic problem of com-binatorial

optimization and is unlikely to find an efficient algorithm for solving TSPs directly. In the

last two decades, ant colony optimization (ACO) has been successfully used to solve TSPs and

their associated applicable problems. De-spite the success, ACO algorithms have been facing

constantly challenges for improving the slow convergence and avoiding stagnation at the local

optima. In this paper, we propose a new hybrid algorithm, cooperative ant colony system and

genetic algorithm (CoACSGA) to deal with these problems. Unlike the previous studies that

regarded GA as a sequential part of the whole searching process and only used the result from

GA as the input to the subsequent ACO iteration, this new approach combines both GA and

ACS together in a coopera-tive and concurrent fashion to improve the performance of ACO

for solving TSPs. The mutual information exchange between ACS and GA at the end of each

iteration ensures the selection of the best solution for the next round, which accelerates the

convergence. The cooperative approach also creates a bet-ter chance for reaching the global

optimal solution because the independent running of GA will maintain a high level of diversity

in producing next genera-tion of solutions. Compared with the results of other algorithms, our

simulation demonstrates that CoACSGA is superior to other ACO related algorithms in terms

of convergence, quality of solution, and consistency of achieving the global optimal solution,

particularly for small-size TSPs.

Adaptive Particle Swarm Optimization Based Artificial Immune

Network Classification Algorithm

Ruochen Liu1, Manchun Niu1, Lina Tang1 and Licheng Jiao1

1Xidian University, China

Abstract. Artificial immune network algorithm (AIN) is a new computational intelligence

method. The mutation operators in most of existed artificial immune network algorithms

57

for classifier are random mutation, which leads to a just passable search ability and low

classification accuracy. In order to overcome such problem and guide the B-cells to evolve

in optimal direction, an adaptive Particle Swarm Optimization (PSO) is introduced into AIN

as a new mutation operation and a new classification algorithm - Adaptive PSO based Artificial

Immune Network Classification algorithm (APAINC) is proposed. The proposed algorithm has

been extensively compared with Artificial Immune Network Classification algorithm based on

random mutation (AINC) and Artificial Immune Network Classification Algorithm based on

PSO (PSOAINC) over four UCI data sets with large size and two artificial texture images and

three SAR images. The result of experiment indicates the superiority of APAINC over AINC

and PSOAINC on classification accuracy.

A Hybrid GA-PSO Approach with Gene Clustering for the

Inference of Gene Regulatory Networks

Wei-Po Lee1

1National Sun Yat-sen University, Taiwan, China

Abstract. The construction of gene regulatory networks from expression data is one of the

most important issues in systems biology research. However, building such networks is a

tedious task, especially when both the number of genes and the complexity of gene regulation

increase. In this work, we adopt the S-system model to represent the gene network and

establish a methodology to infer the model. Our work mainly includes an adaptive GA-PSO

hybrid method to infer appropriate network parameters, and a gene clustering method to

decompose a large network into several smaller networks for dimension reduction. To validate

the proposed methods, different series of experiments have been conducted and the results show

that the proposed methods can be used to infer S-system models of gene networks efficiently

and successfully.

Artificial Immune System

June 14, 2010(Monday) 08:00-09:40 Room C

A Quantum Immune Algorithm for Multiobjective Parallel

Machine Scheduling

Zhimin Fang1


Abstract. The study presents a novel quantum immune algorithm (QIA) for solving the

parallel machine scheduling in the textile manufacturing industry. In this proposed algorithm,

there are distinct characteristics as follows. First, the encoding method is based on Q-bit

58


representation. Second, a novel mutation operator with a chaos-based rotation gate is proposed.

Most importantly, two diversity schemes, suppression algorithm and similarity-based truncation

algorithm, are employed to preserve the diversity of the population, and a new selection scheme

is proposed to create the new population. Simulation results show that QIA is better than two

quantum-inspired evolutionary algorithms.

Cryptanalysis of Four-Rounded DES using Binary Artificial

Immune System

Syed Hamdani1, Sarah Shafiq1 and Farrukh Khan1

1FAST National University of Computer and Emerging Sciences (NUCES), Pakistan

Abstract. In this paper, we present a new approach for the cryptanalysis of four-rounded Data

Encryption Standard (DES) based on Artificial Immune System (AIS). The proposed algorithm

is a combination of exploitation and exploration of fitness landscape where it performs local

as well as global search. The algorithm has the property of automatically determining the

population size and maintaining the local solutions in generations to generate results close to

the global results. It is actually a known plaintext attack that aims at deducing optimum

keys depending upon their fitness values. The set of deduced or optimum keys is scanned to

extract the valuable bits out by counting all bits from the deduced key set. These valuable

extracted bits produce a major divergence from other observed bits. This results in a 56-bit

key deduction without probing the whole search space. To the best of our knowledge, the

proposed algorithm is the first attempt to perform cryptanalysis of four-rounded DES using

Artificial Immune System.

A Immune Concentration Based Virus Detection Approach using

Particle Swarm Optimization

Wei Wang1, Pengtao Zhang1 and Ying Tan1


Abstract. This paper proposes an immune concentration based virus detection approach

which utilizes a two-element concentration vector to construct the feature. In this approach, self

and nonself concentrations are extracted through self and nonself detector libraries, respectively,

to form a vector with two elements of concentrations for characterizing the program efficiently

and fast. Several classifiers including k-nearest neighbor (KNN), RBF neural network and

support vector machine (SVM) with this vector as input are then employed to classify the

programs. The selection of detector library determinant and parameters associated with a

certain classifier is here considered as an optimization problem aiming at maximizing the

59

accuracy of classification. A clonal particle swarm optimization (CPSO) algorithm is used for

this purpose. Experimental results demonstrate that the proposed approach not only has a

very much fast speed but also gives around 98% of accuracy under optimum conditions.

Immune Algorithm with Memory Coevolution

Tao Liu1

1Suzhou Vocational University, China

Abstract. Although various immune algorithms have been proposed by researchers to be put

to use in engineering practice, these immune algorithms fail to take into account the impact

of the complex relationship between the environment and the individual on the evolution of

the individual. Therefore, the convergence rate of such algorithm can be slow in practical

applications. The Memory Coevolution Immune Algorithm (MCIA) is proposed to overcome

the above defect. The strategy of memory coevolution was proposed; the distance concentration

and affinity function were defined. Based on the evaluation of the antibodies and utilizing the

synergic evolution philosophy, the antibodies in the memory library are selected and crossed

with the cloned antibodies according to the affinity value. As a result, the excellent antibody

genes are spread among different antibodies. Meanwhile, on the basis of the full mutation, the

other antibodies in the memory library are selected, and are crossed with the cloned antibodies.

As a result the antibody genes with poor quality have been contained. The experiment shows

that the adoption of memory coevolution mechanism in MCIA enhanced the algorithm’s search

capabilities.

An Improved Immune Genetic Algorithm for Multiobjective

Optimization

Guixia He1, Jiaquan Gao and Luoke Hu


Abstract. The study presents a novel weight-based multiobjective immune genetic algorithm

(WBMOIGA), which is an improvement of its first version. In this proposed algorithm, there

are distinct characteristics as follows. First, a randomly weighted sum of multiple objectives is

used as a fitness function, and a local search procedure is utilized to facilitate the exploitation of

the search space. Second, a new mate selection scheme, called tournament selection algorithm

with similar individuals (TSASI), and a new environmental selection scheme, named truncation

algorithm with similar individuals (TASI), are presented. Third, we also suggest a new

selection scheme to create the new population based on TASI. Simulation results on three

standard problems (ZDT3, VNT, and BNH) show WBMOIGA can find much better spread

60


of solutions and better convergence near the true Pareto-optimal front compared to the elitist

non-dominated sorting genetic algorithm (NSGA-II).

Multi-Agent Based Complex Systems


Impulsive Consensus Seeking in Delayed Networks of

Multi-Agents

Quanjun Wu1, Lan Xiang1 and Jin Zhou1

1Shanghai University, China

Abstract. This present paper addresses impulsive consensus problem in directed networks of

dynamic agents having communication delays. Based on impulsive control theory on delayed

dynamical systems, a simple impulsive consensus protocol for such networks is proposed, and a

generic criterion for solving average consensus problem is analytically derived. It is shown that

global average consensus of a directed delayed networked multi-agent systems can be achieved

by a suitable design of the impulsive gain and impulsive interval. Simulations are presented

that are consistent with the theoretical results.

The Application of Multi-Agent Technology on the Level of

Repair Analysis

Xiangkai Liu1, Yanfeng Tang1, Lin Zheng2, Bingfeng Zhu2 and Jia-ning Wang2

1Academy of Military Transportation, China

2Academy of Logistic Commanding, China

Abstract. The basic theory of level of repair analysis (LORA) has been discussed. The

route to apply multi-agent technology to accomplish the computer aided analysis for LORA

has been investigated. A LORA system based on multi-agent system has been presented, the

structure, the non-economic analysis agent has been researched. It can overcome the problem

of information sharing and the cooperation between analysts effectively, and provide a new

route to accomplish the computer aided analysis for LORA.

The Framework of Intelligent Battlefield Damage Assessment

System Based on Multi-agent System

Xiangkai Liu1, Huimei Li1, Jian Zhang, Jianing Wang1 and Wenhua Xing

1Academy of Military Transportation, China

Abstract. The basic content and procedure of Battlefield Damage Assessment(BDA) has

been discussed and researched. The structure, the disposal strategy, the cooperation between

61

agents, and the data flow of an Intelligent Battlefield Damage Assessment System(IBDAS)

based on multi-agent system(MAS) has been studied. This system can solve the difficulty

of BDA under the complicated and changing battlefield environment, and lay the theoretical

foundation for the realization of a practical IBDAS based on multi-agent system.

Adaptive System of Heterogeneous Multi-agent Investors in an

Articial Evolutionary Double Auction Market

Chi Xu1, Xiaoyu Zhao and Zheru Chi1

1The HK Polytechnic University, Hongkong, China

Abstract. In this paper, an adaptive system is proposed which attempts to combine together

the approaches of studies of historical data and re-searches of multi-agent artificial market by

evolving a double auction market model with diversity of different traders. The purpose of

this re-search is to construct an artificial market which is more close to realistic one and more

practical for future researches. The model with heterogeneous agents and the environment with

which agents and market interact is complicated but controllable by data mining the optimal

proportion of the different agents at the input to the market that generates an output which

can fit historical data curve. The simulation results suggest that the system performance is

close to the expecting values in the testing with adequate training in advance.

Average Consensus for Directed Networks of Multi-agent with

Time-Varying Delay

Tiecheng Zhang1 and Hui Yu1

1China Three Gorges University, China

Abstract. The average consensus in directed network of multi-agent with both switching

topology and time-varying delay is studied. An orthogonal matrix is introduced to change

the initial system into a reduced dimensional system. Based on linear matrix inequalities

(LMIs) technique, a sufficient condition about average consensus problem is proposed. A

novel form in terms of LMIs is obtained via taking the relationship between the terms in the

Newton-Leibniz formula into account. Because some free weighted matrices are employed in

the analysis processing and are selected through solving LMIs appropriately, our method is

less conservative and more general. Finally, simulation examples are given to demonstrate the

effectiveness of the theoretical results.

62


Multi- Cooperative Agent Reinforcement Learning in 3D Virtual

World

Ping Zhang1, Xiujun Ma1, Zijian Pan1, Xiong Li1 and Kunqing Xie


Abstract. A virtual world is an online community in the form of a computer-based simulated

environment, through which users can interact with one another and use and create objects.

The non-player characters (NPC) in virtual world are following a fixed set of pre-programmed

behaviors and lack the ability to adapt with the changing surrounding. Reinforcement learning

agent is a way to deal with this problem. However, in a cooperative social environment, NPC

should learn not only by trial and error, but also through cooperation by sharing information.

The key investigation of this paper is: modeling the NPCs as multi-agent, and enable them to

conduct cooperative learning, then speeding up the learning process. By using a fire fighting

scenario in Robocup Rescue, our research shows that sharing information between cooperative

agents will outperform independent agents who do not communicate during learning. The

further work and some important issues of multi-agent reinforcement learning in virtual world

will also be discussed in this paper.

Multi-Robot Systems


Enhanced Mapping of Multi-robot Using Distortion Reducing

Filter Based SIFT

Kyung-Sik Choi1, Yoon-Gu Kim2, Jinung An2 and Suk-Gyu Lee1

1Yeungnam University, Republic of Korea

2Daegu Gyeongbuk Institute of Science and Technology, Republic of Korea

Abstract. This paper proposes an enhanced mapping of multi-robot using a DSIFT to reduce

the mapping calculation time. In this approach, the master robot transmits each robot’s

mapping information in SLAM by DSIFT, which incorporates an additional step on the SIFT.

The DSIFT uses a keypoint to reduce the distortional information throughout the Gaussian

filter after the step of the image descriptor. The master robot calculates the slave robot’s pose

using picture images, and serves the results to all the robots. Simulation results are presented

based on DSIFT showing better performance than using the SIFT in multi-robot mapping

situations.

63

Development of Image Stabilization System Using Extended

Kalman Filter for a Mobile Robot

Yun Won Choi1, Tae Hoon Kang2 and Suk Gyu Lee1


2Pohang Institute of Intelligent Robotics, Republic of Korea

Abstract. This paper proposes a robust image stabilization system for a mobile robot using

Extended Kalman Filter (EKF). Though image information is one of the most efficient data

for robot navigation, it is subject to noise which results from internal vibration as well as

external factors such as uneven terrain, stairs, or marshy surface . The vibration of camera

deteriorates the definition of image by destroying image sharpness, which seriously prevents

mobile robots from recognizing their environment for navigation. In this paper, inclinometer

was used to measure the vibration angle of the camera system mounted on the robot to obtain

a reliable image by compensating for the angle of the camera shake caused by vibration. In

addition angle prediction by using the EKF enhances responsibility of image analysis for real

time performance. The Experimental results show effectiveness of the proposed system to

compensate for the blurring of the images.

Swarm robot: from self-assembly and locomotion

Hongxing Wei1, Yingpeng Cai1 and Tianmiao Wang1


Abstract. Inspired by the swarm behaviors of social insects, the research into the self-assembly

of the swarm robots has become an attractive issue in the robotic community. Unfortunately,

there are very few platforms with self-assembly and locomotion in field of the swarm robotics.

The Sambot is a novel self-assembly modular robot that shares characteristics with the swarm

robots and the self-reconfigurable robots. Each Sambot can move autonomously and connect

with the other. This paper discusses the concept combining self-assembly and locomotion for

swarm robots. The distributed control algorithms of the self-assembly and the locomotion

is proposed. Using 5 physical Sambots, the experiments of the autonomous docking and

the self-assembly and the locomotion have been implemented. Our control algorithm of

self-assembly can also be used to realize the autonomous construction and self-repair of robotic

structure consisting of a large number of Sambots.

64


Diffusing Method for Unknown Environment Exploration in

Multi Robot Systems

Dilshat Saitov1, Ki Joon Han and Suk Gyu Lee1


Abstract. This paper proposes an algorithm for an efficient navigation and building a precise

map in multi-robot systems. One of the fundamental problems in mobile robotics is an effective

investigation of unknown environments. The basis of navigation algorithm in this paper is

Extented Wave Algorithm, which is in our point of view, appropriate in getting accurate.

Secondly, particle filter, which proved its reliability, was considered as localization algorithm.

Finally, overlapping algorithm is responsible for mapping. The technique has been tested

extensively in simulation runs. The results given in this paper demonstrate that our algorithm

significantly reduces the exploration time compared to previous approaches.

Localization and Full Coverage Path Planning for a Cleaning

Robot

Hongbo Wang1 and Zhengwei Hu1

1Yanshan University, China

Abstract. This paper describes a novel method of using multi-sensors to realize indoor and

outdoor localization and proposes an improved full coverage path planning algorithm for a

cleaning robot. First, the motion model of the robot is presents and the computer simulation

shows that the localization method has a high accuracy. Second, the improved path planning

algorithm based on biologically inspired neural network is introduced. Then, a control system

for the cleaning robot prototype is designed. Finally, experiments are conducted and the results

show that the proposed localization method and the path planning algorithm are feasible, and

the control system is effective.

A Novel Spatial Obstructed Distance by Dynamic Piecewise

Linear Chaotic Map and Dynamic Nonlinear PSO

Xueping Zhang1, Yawei Liu, Jiayao Wang and Haohua Du

1Henan University of Technology, China

Abstract. Spatial Clustering with Obstacles Constraints (SCOC) has been a new topic in

Spatial Data Mining (SDM). Spatial Obstructed Distance (SOD) is the key to SCOC. The

obstacles constraint is generally ignored in computing distance between two points, and it

leads to the clustering result which is of no value, so obstructed distance has a great effect

65

upon clustering result. In this paper, we propose a novel Spatial Obstructed Distance using

Dynamic Piecewise Linear Chaotic Map and Dynamic Nonlinear Particle Swarm Optimization

(PNPSO) based on Grid model to obtain obstructed distance, which is named PNPGSOD, it

is not only simple and easy to actualize, but also convergent rapidly, the experimental results

are provided to verify the effectiveness and practicability.

Classifier Systems


SVM Classifier Based Feature Selection Using GA, ACO and

PSO for siRNA Design

Yamuna Prasad1, Kanad Biswas1 and Chakresh Jain

1Indian Institute of Technology, India

Abstract. Recently there has been considerable interest in applying evolutionary and natural

computing techniques for analyzing large datasets with large number of features. In particular,

efficacy prediction of siRNA has attracted a lot of researchers, because of large number of

features involved. In the present work, we have applied the SVM based classifier along with

PSO, ACO and GA on Huesken dataset of siRNA features as well as on two other wine and

wdbc breast cancer gene benchmark dataset and achieved considerably high accuracy and

the results have been presented. We have also highlighted the necessary data size for better

accuracy in SVM for selected kernel. Both groups of features (sequential and thermodynamic)

are important in the efficacy prediction of siRNA. The results of our study have been compared

with other results available in the literature.

A Discrete-time Recurrent Neural Network for Solving Systems

of Complex-valued Linear Equations

Wudai Liao1, Jiangfeng Wang1 and Junyan Wang1

1Zhongyuan University of Technology, China

Abstract. A discrete-time recurrent neural network is presented in this paper for solving

systems of complex-valued linear equations. The network shown in this paper is simple in

structure and can converge to the solutions of complex-valued linear equations. The condition

for the neural network to globally converge to the complex-valued linear equations is given.

An illustrative example is presented to illustrate its performance.

66


A Recurrent Neural Network for Solving Complex-valued

Quadratic Programming Problems with Equality Constraints

Wudai Liao1, Jiangfeng Wang1 and Junyan Wang1

1Zhongyuan University of Technology, China

Abstract. A recurrent neural network is presented for solving systems of quadratic programming

problems with equality constraints involving complex-valued coefficients. The proposed recurrent

neural network is asymptotically stable and able to generate optimal solutions to quadratic

programs with equality constraints. An opamp based analogue circuit realization of the

recurrent neural network is described. An illustrative example is also discussed to demonstrate

the performance and characteristics of the analogue neural network.

Computer-Aided Detection and Classification of Masses in

Digitized Mammograms Using Artificial Neural Network

Mohammed Islam1, Majid Ahmadi1 and Maher A. Sid Ahmed1

1University of Windsor, Canada

Abstract. In this paper we present a computer aided diagnosis (CAD) system for mass

detection and classification in digitized mammograms, which performs mass detection on

regions of interest (ROI) followed by the benign-malignant classification on detected masses.

In order to detect mass effectively, a sequence of preprocessing steps are proposed to enhance

the contrast of the image, remove the noise effects, remove the x-ray label and pectoral muscle

and locate the suspicious masses using Haralick texture features generated from the spatial

gray level dependence (SGLD) matrix. The main aim of the CAD system is to increase the

effectiveness and efficiency of the diagnosis and classification process in an objective manner

to reduce the numbers of false-positive of malignancies. Artificial neural network (ANN) is

proposed for classifying the marked regions into benign and malignant and 83.87% correct

classification for benign and 90.91% for malignant is achieved.

Gene Selection and PSO-BP Classifier Encoding a Prior

Information

Yu Cui1, Fei Han1 and Shiguang Ju1

1Jiangsu University, China

Abstract. . Selecting a relevant and discriminative combination of genes for cancer classification

and building high - performing classifier are common and critical tasks in cancer classification

problems. In this paper, a new approach is proposed to address the two issues at the same

67

time. In details, BP neural network is employed to construct a classifier, and PSO algorithm is

used to select a discriminative combination of genes and optimize the BP classifier accordingly.

Besides, sample’s prior information is encoded into PSO algorithm for better performance.

The proposed approach is validated on the leukemia data set . The experimental results show

that our novel method selects fewer discriminative genes while has comparable performance to

the traditional classification approaches.

Object Recognition of a Mobile Robot based on SIFT with

De-speckle Filtering

Zhiguang Xu1, Kyung-Sik Choi1, Yanyan Dai1 and Suk-Gyu Lee1


Abstract. T his paper present s a novel object recognition method, of a mobile robot, by

combining scale invariant feature transform (SIFT) and de-speckle filtering to enhance the

recognition capability. The main idea of the proposed algorithm is to use SIFT programming

to identify other robot s after de-speckle filtering process to remove outside noise. Since

a number of features are much larger than needed, SIFT method requires a long time to

extract and match the features. The proposed method shows a faster and more efficient

performance, which enhances localization accuracy of the slave robots. From the simulation

results, the method using de-speckle filtering based SIFT shows that the number of features in

the extraction process, and that the points in matching process are reduced.

Evolutionary Computation


Bottom-up Tree Evaluation In Tree-based Genetic Programming

Geng Li1 and XiaoJun Zeng1

1University of Manchester, UK

Abstract. In tree-based genetic programming (GP) performance optimization, the primary

optimization target is the process of fitness evaluation. This is because fitness evaluation takes

most of execution time in GP. Standard fitness evaluation uses the top-down tree evaluation

algorithm. Top-down tree evaluation evaluates program tree from the root to the leaf of the

tree. The algorithm reflects the nature of computer program execution and hence it is the most

widely used tree evaluation algorithm. In this paper, we identify a cenario in tree evaluation

where top-down evaluation is costly and less effective. We then propose a new tree evaluation

algorithm called bottom-up tree evaluation explicitly addressing the problem identified. Both

theoretical analysis and practical experiments are performed to compare the performance of

68


bottom-up tree evaluation and top-down tree evaluation. It is found that bottom-up tree

evaluation algorithm outperforms standard top-down tree evaluation when the program tree

depth is small.

Solving Vehicle Assignment Problem Using Evolutionary

Computation

Marina Yusoff1, Junaidah Ariffin1 and Azlinah Mohamed1

1Universiti Teknologi MARA, Malaysia

Abstract. This paper examines the use of evolutionary computation (EC) to find optimal

solution in vehicle assignment problem (VAP). The VAP refers to the allocation of the expected

number of people in a potentially flooded area to various types of available vehicles in evacuation

process. A novel discrete particle swarm optimization (DPSO) algorithm and genetic algorithm

(GA) are presented to solve this problem. Both of these algorithms employed a discrete solution

representation and incorporated a min-max approach for a random initialization of discrete

particle position. A min-max approach is based on minimum capacity and maximum capacity

of vehicles. We analyzed the performance of the algorithms using evacuation datasets. The

quality of solutions were measured based on the objective function which is to find a maximum

number of assigned people to vehicles in the potentially flooded areas and central processing

unit (CPU) processing time of the algorithms. Overall, DPSO provides an optimal solutions

and successfully achieved the objective function whereas GA gives sub optimal solution for the

VAP.

An Improved Thermodynamics Evolutionary Algorithm Based

on the Minimal Free Energy

Fahong Yu1, Yuanxiang Li1 and Weiqin Ying1

1Wuhan University, China

Abstract. In this paper, an improved thermodynamics evolutionary algorithm (ITEA) is

proposed. The purpose of the new algorithm is to systematically harmonize the conflict between

selective pressure and population diversity while searching for the optimal solutions. ITEA

conforms to the principle of minimal free energy in simulating the competitive mechanism

between energy and entropy in annealing process, in which population diversity is measured

by similarity entropy and the minimum free energy is simulated with an efficient and effective

competition by free energy component. Through solving some typical numerical optimization

problems, satisfactory results were achieved, which showed that ITEA was a preferable algorithm

to avoid the premature convergence effectively and reduce the cost in search to some extent.

69

A Non-dominated Sorting Bit Matrix Genetic Algorithm for P2P

Relay Optimization

Qian He1,2, Junliang Chen1, Xiangwu Meng1 and Yanlei Shang1

1Beijing University of Posts and Telecommunications, China

2Guilin University of Electronic Technology, China

Abstract. Cooperative caching and relaying content in ISPs can decrease the bandwidth

costs and distribution time. The relay resources installed at ISP are limited and the upload

rates of relay servers are various. After formulating the optimization problem, we design a

Nondominated Sorting Bit matrix Genetic Algorithm (NSBGA) to solve it. Constraint-satisfied

population is initialized according to resource ratio dynamically; improved alone point crossover

and symmetric mutation is designed; population is non-dominated sorted. The experiments

show that NSBGA is better than NSGAII and it can support P2P relay optimization very well.

The relations between performances and parameters as the numbers of ISPs, source channels

and relay servers are analyzed.As a general optimization algorithm, NSBGA also can be used

in other application fields.

Fast Parallel Memetic Algorithm for Vector Quantization Based

for Reconfigurable Hardware and Softcore Processor

Tsung-Yi Yu1, Wen-Jyi Hwang1 and Tsung-Che Chiang1

1National Taiwan Normal University, Taiwan, China

Abstract. A novel parallel memetic algorithm (MA) architecture for the design of vector

quantizers is presented in this paper. The architecture contains a number of modules operating

memetic optimization concurrently. Each module uses steady-state genetic algorithm (GA) for

global search, and K-means algorithm for local refinement. A shift register based circuit for

accelerating mutation and crossover operations for steady state GA operations is adopted in

the design. A pipeline architecture for the hardware implementation of K-means algorithm is

also used. The proposed architecture is embedded in a softcore CPU, and implemented on a

field programmable logic array (FPGA) device for physical performance measurement.

Integrate spatial information into multi-objective genetic

algorithm for spatial optimal location based on GIS

Jinliang Hou1, Haiqi Wang1 and Yujie Liu1

1China University of Petroleum, China

Abstract. This paper demonstrates the method integrate spatial information into multi-objective

70


genetic algorithm to solve spatial optimal location problem based on GIS. Firstly, we have

a brief introduction of Modified Non-dominated Sorting Genetic Algorithm. Secondly, we

elaborate on the way of how the spatial information introduced into NSGA-II and combined

with GIS technology. Finally, we will verify this method by a case of selecting the optimal

location of disease surveillance and control sites in Shandong Province, China. It is concluded

that our method can converge to the Pareto-optimal set and is a feasible way of solving

multi-objective spatial optimal location problem.

Data Mining Methods


Intelligent Decision Support System for Breast Cancer

RR Janghel1, Anupam Shukla, Ritu Tiwari and Rahul Kala

1Indian Institute of Information Technology and Management, India

Abstract. Breast cancer is the second leading cause of cancer deaths in women worldwide

and occurs in nearly one out of eight women. Currently there are three techniques to diagnose

breast cancer: mammography, FNA (Fine Needle Aspirate) and surgical biopsy. In this paper

we develop an integrated expert system for diagnosis, prognosis and prediction for breast

cancer using soft computing techniques. The basic aim is to compare the various neural

network models from the recent literature. Breast cancer database used for this purpose is

from the University of Wisconsin (UCI) Machine Learning Repository. Three different data

sets have been used, each employing different diagnostic technique. It can use diagnosis,

prognosis and survivability prediction of breast cancer patient in one intelligent system. We

implement six models of neural networks namely Back Propagation Algorithm, Radial Basis

Function Networks, Learning vector Quantization, Probabilistic Neural Networks, Recurrent

Neural Network, and Competitive Neural network. Experimental Results show that different

models give optimal performance for different types of data sets. However, all the models are

able to solve the problem to a reasonable extent.

An Automatic Index Validity for Clustering

Zizhu Fan1,2, Xiangang Jiang2, Baogen Xu2 and Zhaofeng Jiang2

1Harbin Institute of Technology, China

2East China Jiaotong University, China

Abstract. Many validity index algorithms have been proposed to determine the number of

clusters. These methods usually employ the Euclidean distance as the measurement. However,

it is difficult for the Euclidean distance metric to evaluate the compactness of data when

71

non-linear relationship exists between different components of data. Moreover, most current

algorithms can not estimate well the scope of the number of clusters. To address these problems,

in this paper, we adopt the kernel-induced distance to measure the relationship among data

points. We first estimate the upper bound of the number of clusters to effectively reduce

iteration time of validity index algorithm. Then, to determine the number of clusters, we

design a kernelized validity index algorithm to automatically determine the optimal number of

clusters. Experiments show that the proposed approach can obtain promising results.

A Novel Fast Non-negative Matrix Factorization Algorithm and

Its Application in Text Clustering

Fang Li1 and Qunxiong Zhu1

1Beijing University of Chemical Technology, China

Abstract. In non-negative matrix factorization, it is difficult to find the optimal non-negative

factor matrix in each iterative update. However, with the help of transformation matrix,

it is able to derive the optimal non-negative factor matrix for the transformed cost function.

Transformation matrix based nonnegative matrix factorization method is proposed and analyzed.

It shows that this new method, with comparable complexity as the priori schemes, is efficient

in enhancing nonnegative matrix factorization and achieves better performance in NMF based

text clustering.

Application KIII Model to EEG Classification Based on

Nonlinear Dynamic Methods

Jin Zhang1

1Beijing Normal University, China

Abstract. Based on the biological olfactory systems, a chaotic neural network, KIII model,

was proposed by Prof. Walter J. Freeman. KIII model not only can simulate the output

waveforms in electroencephalogram (EEG), but also has the capability of pattern recognition.

Based on nonlinear dynamic methods, two nonlinear dynamics indexes, ApEn and Lyapunov

exponents, are used to extract EEG feature. KIII model is used to recognize hypoxia EEG.

Experimental results show that (1) ApEn and Lyapunov exponents can denote the characteristics

effectively; (2) KIII model has good performance to recognize the nonlinear signals.

72


Brain-Computer Interface System Using Approximate Entropy

and EMD Techniques

Qiwei Shi1, Wei Zhou1, Jianting Cao1, Toshihisa Tanaka2 and Rubin Wang3

1Saitama Institute of Technology, Japan

2Tokyo University of Agriculture and Technology, Japan

3East China University of Science and Technology, China

Abstract. Brain-computer interface (BCI) is a technology which would enable us to communicate

with external world via brain activities. The electroencephalography (EEG) now is one of the

non-invasive approaches and has been widely studied for the brain computer interface. In this

paper, we present a motor imaginary based BCI system. The subject’s EEG data recorded

during left and right wrist motor imagery is used as the input signal of BCI system. It is known

that motor imagery attenuates EEG u and B rhythms over contralateral sensorimotor cortices.

Through offline analysis of the collected data, a approximate entropy (ApEn) based complexity

measure is first applied to analyze the complexity between two channels located in different

hemispheres. Then, empirical mode decomposition (EMD) is used to extract informative brain

activity features to discriminate left and right wrist motor imagery tasks. The satisfactory

results we obtained suggest that the proposed method has the potential for the classification

of mental tasks in brain-computer interface system.

A Novel Spatial Clustering with Obstacles Constraints Based on

PNPSO and K-Medoids

Xueping Zhang1, Haohua Du, Tengfei Yang and Guangcai Zhao

1Henan University of Technology, China

Abstract. In this paper, we propose a novel Spatial Clustering with Obstacles Constraints

(SCOC) based on Dynamic Piecewise Linear Chaotic Map and Dynamic Nonlinear Particle

Swarm Optimization (PNPSO) and K-Medoids, which is called PNPKSCOC. The contrastive

experiments show that PNPKSCOC is effective and has better practicalities, and it performs

better than PSO K-Medoids SCOC in terms of quantization error and has higher constringency

speed than Genetic K-Medoids SCOC.

Fuzzy Methods and Information Processing System


73

Modeling Fuzzy Data Envelopment Analysis with Expectation

Criterion

Xiaodong Dai1, Ying Liu1 and Rui Qin1

1Hebei University, China

Abstract. This paper presents a new class of fuzzy expectation data envelopment analysis

(FEDEA) models with credibility constraints. Since the proposed model contains the credibility

of fuzzy events in the constraints and the expected value of a fuzzy variable in the objective,

the solution process is very complex. Thus, in the case when the inputs and outputs are

mutually independent trapezoidal fuzzy variables, we discuss the equivalent nonlinear forms of

the programming model, which can be solved by standard optimization software. At the end of

this paper, one numerical example is also provided to illustrate the efficiency of decision-making

unites (DMUs) in the proposed model.

On Fuzzy Diagnosis Model of Plane’s Revolution Swing Fault

and Simulation Researches

Dongcai Qu1, Jihong Cheng, Wanli Dong1 and Ruizhi Zhang

1Naval Aeronautical and Astronautical University, China

Abstract. Considering the fact that traditional fault diagnosis can’t absorb human’s experiences

well, this paper simulated the procedure of expert’s interference with fuzzy interference to build

a fault diagnosis model, and use fuzzy network to improve the model. The result of simulation

proved that this model can absorb the experiences of human and make accurate judgments;

the trained fuzzy network has the same function and can reach the self-learning demand.

Asymptotic Equivalent Analysis for LTI Overlapping Large-Scale

Systems and Their Subsystems

Qian Wang1 and Xuebo Chen1

1Liaoning University of Science and Technology, China

Abstract. According to the matrix exponential function and the matrix stability, a criterion

of asymptotic equivalence is proposed in this paper. The criterion is presented for linear

time-invariant (LTI) overlapping large-scale systems and their pair-wise subsystems which

are decomposed by the inclusion principle. The study of asymptotic equivalence offers the

convenience for the stable analysis, furthermore, offers rationale for the asymptotic equivalent

analysis for the other large-scale systems and their isolated subsystems. Simultaneously, an

example has been given to illustrate the feasibility and the validity of this method. Keywords:

74


LTI overlapping large-scale systems, Pair-wise subsystems, Asymptotic equivalence, Matrix

exponential function

An Application of LFP Method for Sintering Ore Ratio

Xi Cheng1, Kailing Pan1 and Yunfeng Ma1

1Wuhan University of Science and Technology, China

Abstract. The proper ratio decision of sintering burden is a significant section for both of

decreasing sintering costs and increasing quality of iron. At present most company in China

take the Fixed - ratio method and linear programming (LP) model to calculate the proper ratio

for sintering. The former is the performance appraisal method for production cost management

of iron . The latter is to use maths method to improve the computation process. This paper

brings up a linear fractional programming (LFP) model combining the advantages of both

methods to compute the proper ratio to minimize the iron cost per ton for sintering. Next b

ased on the production data from some steel company this paper takes use of MATLAB to

solve the problem. Then comparing the solutions with the original method, the traditional LP

model and LFP model the conclus ions are revealed in the end .

A Class of Fuzzy Portfolio Optimization Problems: E-S Models

Yankui Liu1 and Xiaoli Wu1


Abstract. This paper adopts the spread of fuzzy variable as a new criteria in practical

risk management problems, and develops a novel fuzzy expectation-spread (E-S) model for

portfolio optimization problem. Since the spread is defined by Lebesgue-Stieltjes (L-S) integral,

its computation for general fuzzy variables is a challenge issue for research, and usually

depends on approximation scheme and soft computing. But for frequently used trapezoidal

and triangular fuzzy variables, the spread can be represented as quadratic functions with

respect to fuzzy parameters. These new representations facilitate us to turn the proposed E-S

model into its equivalent parametric programming problem. As a consequence, given the fuzzy

parameters, the E-S model becomes a quadratic programming problem that can be solved by

general purpose software or conventional optimization algorithms. Finally, we demonstrate the

developed modeling idea via two numerical examples.

75

Matrix Estimation Based on Normal Vector of Hyperplane in

Sparse Component Analysis

Feng Gao1, Gongxian Sun1, Ming Xiao2 and Jun Lv1

1South China University of Technology, China

2Maoming University, China

Abstract. This paper discusses the matrix estimation for sparse component analysis under

the k-SCA condition. Here, to estimate the mixing matrix using hyperplane clustering, we

propose a new algorithm based on normal vector for hyperplane. Compared with the Hough

SCA algorithm, we give a method to calculate normal vector for hyperplane, and the algorithm

has lower complexity and higher precision. Two examples demonstrates its performance.

Signal Processing and Information Security


Pricing and Bidding Strategy in AdWords Auction under

Heterogeneous Products Scenario

E Zhang1 and Yiqin Zhuo1

1Shanghai University of Finance and Economics, China

Abstract. This paper focus on biding and pricing strategies in a scenario two heterogeneous

products manufacturers selling through on-line channel. The firms competes customers in

quality to price ratio. The value of prominent AdWords advertising position and the resulting

price dispersion patterns are studied. We found that prominent position of an Ad words is

not always favorite to all firms according to the analysis based on game theory. For the firm

which produced high-quality products, the revenue gained from listed on the prominent place is

always higher than in the second place; However, for the low-quality product firm the revenue

gained from its advertisement listed on the prominent place might less than on the second place.

Meanwhile the attractiveness of the prominent Ad place depends on the market structure in

terms of consumer preference and consumer search behavior. The more consumers purchase

from the firm listed in the prominent Ad places or the more consumers prefer high-quality

product the more strict area in which the low-quality product manufacture has positive profit.

FIR Cutoff Frequency Calculating for ECG Signal Noise

Removing Using Artificial Neural Network

Sara Moein1

1Azad University of Najafabad Branch, Iran

76


Abstract. In this paper, an automated approach for electrocardiogram (ECG) signal noise

removing using artificial neural network is investigated. First, 150 of noisy heart signal are

collected form MIT-BIH database. Then signals are transformed to frequency domain and

cutoff frequency is calculated. Since heart signals are lowpass frequency, a Finite Impulse

Response (FIR) filter is adequate to remove the noise. In the next step, a dataset is configured

for a multilayer perceptron (MLP) training with feedforward algorithm. Finally, the MLP is

trained and results of cutoff frequency calculation are shown.

Force Identification by Using SVM and CPSO Technique

Zhichao Fu1, Cheng Wei1 and Yanlong Yang1


Abstract. A novel method is presented to determine the external dynamic forces applied

on structures from measured structural responses in this paper. The method utilizes a new

SVM-CPSO model that hybridized the chaos particle swarm optimization (CPSO) technique

and support vector machines (SVM) to tackle the problem of force identification. Both

numerical simulations and experimental study are performed to demonstrate the effectiveness,

robustness and applicability of the proposed method. It is potential that the proposed method

is practical to the real-life application.

On the Strength Evaluation of Lesamnta Against Differential

Cryptanalysis

Yasutaka Igarashi1 and Toshinobu Kaneko1

1Tokyo University of Science, Japan

Abstract. We focus on the cryptographic hash algorithm Lesamnta-256. Lesamnta-256

consists of the Merkle-Damgard iteration of a compression function and an output function.

The compression function consists of a mixing function and a key scheduling function. The

mixing function consists of 32 rounds of four-way generalized Feistel structure. On each round

there is a nonlinear function F with 64-bit input/output, which consists of the 4 steps of

AES type of SPN (Substitution Permutation Network) structure. A subkey is XORed only at

the first step of the SPN. The designers analyzed its security by assuming that the subkey is

XORed at every step of the SPN. Such an independent subkey assumption is also applied to

the analysis of other SHA-3 candidates, e.g. Grøstl, LANE, Luffa. However we analyze the

security of these components of Lesamnta as is. We show that the 2 steps of SPN referred to as

XS have the maximum differential probability 2−11.415. This probability is greater than both

of the differential characteristic probability 2−18 and the differential probability 2−12 derived

77

under the independent subkey assumption. On the strength of whole compression function, we

show that there are at least 15 active F functions in the mixing function on 64-bit truncated

analysis. As the input bit length of the mixing function is 256, we can say that it is secure

against differential attack if the maximum differential probability of F function is less than

2−256/15 ≈ 2−17.067. We also show that the key scheduling function is secure against differential

cryptanalysis.

Graphics Processing Unit Enhanced Parallel Document Flocking

Clustering

Jesse St. Charles1, Xiaohui Cui2 and Thomas Potok2

1Carnegie Mellon University, USA

2Oak Ridge National Laboratory, USA

Abstract. Analyzing and clustering documents is a complex problem. One explored method

of solving this problem borrows from nature, imitating the flocking behavior of birds. One

limitation of this method of document clustering is its complexity O(n2). As the number of

documents grows, it becomes increasingly difficult to generate results in a reasonable amount

of time. In the last few years, the graphics processing unit (GPU) has received attention for

its ability to solve highly-parallel and semi-parallel problems much faster than the traditional

sequential processor. In this paper, we have conducted research to exploit this architecture

and apply its strengths to the flocking based document clustering problem. Using the CUDA

platform from NVIDIA, we developed a Multiple Species Flocking implementation to be run on

the NVIDIA GPU. Performance gains ranged from thirty-six to nearly sixty times improvement

of the GPU over the CPU implementation.

Computational Intelligence Algorithms Analysis for Smart Grid

Cyber Security

Yong Wang1, Da Ruan2, Jianping Xu1, Mi Wen1 and Liwen Deng3

1Shanghai University of Electric Power, China

2Belgian Nuclear Research Centre, Belgium

3Shanghai Changjiang Computer Group Corporation, China

Abstract. The cyber attack risks are threatening the smart grid security. Malicious worm

could spread from meter to meter to take out power in a simulated attack . The North American

Electric Reliability Corporation (NERC) has thus developed several iterations of cyber security

standards. According to the NERC cyber standards CIP-002-2 requirements, in this paper ,

we present cyber security risk analysis using computational intelligence methods and review on

78


core methods, such as in risk assessment HHM, IIM, RFRM algorithms, fault analysis FTA,

ETA, FMEA, FMECA algorithms, fuzzy sets, intrusion detection systems, artificial neural

networks and artificial immune systems. Through the analysis of the core computational

intelligence algorithms used in the smart grid cyber security in power system network security

lab, we clearly defined existing smart grid research challenges.

Intelligent Control


The Automatic Feed Control Based on OBP Neural Network

Ding Feng1, Bianyou Tan1, Peng Wang1, Shouyong Li1, Jin Liu1, Cheng Yang1,

Yongxin Yuan1 and Guanjun Xu1

1Yangtze University, China

Abstract. It is the important technology to take the optimum control of automatic drilling

in the course of oilfield drilling in accordance with actual situation. Due to the complexity of

drilling process and the non-linear relationship between input and output of drilling system;

it’s difficult to acquire satisfied results to adopt general control method. This article presents

a new control method which based on the OBP neural network. The OBP algorithm and the

design of control system are elaborated in details in this paper. The automatic feed control

method based on OBP neural network has applied successfully in Liaohe and Xinjiang oilfield.

The result indicated that the control system is efficient and response, stability of the system,

the control precision is improved. All the characters index arrive the control required.

GA-Based Integral Sliding Mode Control for AGC

Dianwei Qian1, Xiangjie Liu1, Miaomiao Ma and Chang Xu2

1North China Electric Power University, China

2Hohai University, China

Abstract. This paper addresses an integral sliding mode control approach for automatic

generation control (AGC) of a single area power system. Genetic algorithm (GA) is employed

to search the parameters of the sliding surface. The proposed design is investigated for AGC of

a single area power system, made up of reheated thermal and gas power generations. Compared

with the GA-based proportion-integral (PI) control, simulation results show the feasibility of

the presented method.

79

A Distributed Energy-aware Trust Topology Control Algorithm

for Service-oriented Wireless Mesh Networks

Chuanchuan You, Tong Wang1, Bingyu Zhou, Hui Dai and Baolin Sun

1Wuhan University, China

Abstract. In this paper, we introduce the Energy-aware Trust Topology Control algorithm

based on Ant colony approach (ETTC) that adapts the biological metaphor of Swarm Intelligence

to control topology of wireless mesh networks. As trust is important to consider while forwarding

packets, this paper propose a novel model that integrated the energy consumption and trust

evaluation. The simulations of ETTC show the joint energy-aware and trust effect on the

performance metrics such as network connectivity, node failure rate, etc.

Leader-Follower Formation Control of Multi-Robots by using a

Stable Tracking Control Method

Yanyan Dai1, Viet-Hong Tran1, Zhiguang Xu1 and Suk-Gyu Lee1


Abstract. In this paper, the leader-waypoint-follower robot formation is constructed based

on the relative motion states to form and maintain the formation of multi-robots by stable

tracking control method. The main idea of this method is to find areas onable target velocity

and angular velocity to change the robot’s current state. The propose d Lyapunov functions

prove that robots change current velocities to target velocities which we propose, in globally

asymptotically stable mode . The simulation results based on the proposed approach show

better performance in accuracy and efficiency comparing with EKF based approach which is

applied in multiple robots system in common.

Stable Swarm Formation Control Using Onboard Sensor

Information

Viet-Hong Tran1 and Suk-Gyu Lee1


Abstract. In this paper, a stable leader-following formation control for multiple mobile robot

systems with limited sensor information is studied. The proposed algorithm is to control a

robot (follower) to follow another robot (leader), and easily extended to form any complex

formation. The control algorithm requires information available from onboard sensors only,

and utilizes estimation of leader’s acceleration in a simple form to reduce measurement of

indirect information. There is also a rule to tune parameters of control in application.

80


Other Optimization Algorithms


Research on the Optimization Decision-making Two

Row-sequencing-pairs of Activities with Slacks

Shisen Lv1, Jianxun Qi1, Xiuhua Zhao1 and Zhixiong Su1

1North China Electicity Power University, China

Abstract. In Operation research, how to schedule parallel activities to sequential ones

is typical project scheduling problem with restrained resources, and it is also complicated

optimization problem. In this paper, on the basis of characteristic of CPM (Critical Path

Method) network, theories of Deficient Values of sequencing-pair and standard row-sequencing-pair

are deduced. Based on these theories, an optimization method on selecting 4 activities from

N parallel activities to constitute 2 row-sequencing-pairs of activities is designed. By proof,

using the method could get optimal solution.

Sudoku Using Parallel Simulated Annealing

Zahra Karimi Dehkordi1, Kamran Zamanifar1 and Ahmad Baraani Dastjerdi1

1University of Isfahan, Iran

Abstract. Parallel Simulated Annealing was applied to solving Sudoku puzzle. Simulated

annealing is a stochastic search strategy which is best known for not getting trapped at the local

optimums. Although SA suffers from low efficiency, it has been recognized as one of successful

solutions in Sudoku. Sudoku puzzles in this study were formulated as an optimization problem

in multi-agent environments. Variants of parallel SA could successfully solve this optimization

problem. In this paper we implemented 3 different parallel SA in JADE and compared them.

The results show that parallel search with periodic jumps gets better efficiency and success rate.

The Optimization of Procedure Chain of Three Activities with a

Relax Quantum

Shisen Lv1, Jianxun Qi1 and Xiuhua Zhao1


Abstract. In order to solve the problem of four parallel activities being adjusted to a procedure

chain of three activities and a parallel activity in deterministic activity-on-arc networks of the

CPM type, a new theory and the comparative method of special procedure chains of three

activities are proposed. Based on the theory and method, a branch and bound of the decision

tree is described.

81

A Quay Crane Scheduling Model in Container Terminals

Qi Tang1

1Tianjin Polytechnic University, China

Abstract. This paper discusses the problem of scheduling quay cranes, the most important

equipment in port terminals. A simulation model is developed for evaluating time of quay

cranes. Then a dynamic scheduling model using objective programming for quay cranes is

developed based on genetic algorithm approach. Finally, numerical experiments on a specific

container terminal are made for propose approach. Computational results suggest that the

proposed method is able to solve the problem efficiently.

A Capacitated Production Planning Problem for Closed-loop

Supply Chain

Jian Zhang and Xiao Liu1

1Shanghai Jiaotong University, China

Abstract. This paper addresses a dynamic Capacitated Production Planning (CPP)probleminsteel

enterpriseemployingaclosed-loop supply chain strategy, in which a remanufacturing process is

adopted. We develop a model in which all demands are met byproduction or remanufacturing

without backlogs, under the context that both the production and remanufacturing setup cost

functions are arbitrary and time-varying. We also develop a corresponding genetic algorithm

(GA) heuristic approach, and run a numerical simulation to test our algorithm’s efficiency by

comparing with Branch and Bound method. The simulation results illustrate our algorithm’s

accuracy and efficiency in large scale problems.

Poster Seesion 1

June 14, 2010(Monday)13:30-17:30 Corridor

Fitness Function of Genetic Algorithm in Structural Constraint

Optimization

Xinchi Yan1 and Xiaohan Wang1

1Jiangnan University, China

Abstract. The mathematics models of Reliability-based Structural Optimization (RBSO)

were presented in this paper, then how to handle the constraint become sixty-four-dollar

question of establishing the fitness function. Based on exterior penalty function method,

penalty gene is made adaptively according to population’s evolution, then the fitness function

is established, which is mapping formula of objective function and constraints. Subsequently

82


laxity variable is introduced in primary mathematic model, based on Lagrange multiplier

method, a new fitness function mapping formula is made, this method can avoid penalty

function morbidity by means of adding a Lagrange multiplier, and has a more quick and stable

convergence. Then, using GA successfully solved a numerical constrained optimization issue

by this two mapping functions. The calculation shows that the two equations are reasonable

and efficient, and Lagrange multiplier method has better global optimal capability.

On the Correlations between Fuzzy Variables

Yankui Liu1 and Xing Zhang1


Abstract. The expected value and variance of a fuzzy variable have been well studied in

the literature, and they provide important characterizations of the possibility distribution for

the fuzzy variable. In this paper, we seek a similar characterization of the joint possibility

distribution for a pair of fuzzy variables. In view of the success of introducing the expected

value and variance as fuzzy integrals of appropriate functions of single fuzzy variable, it is

natural to look to fuzzy integrals of appropriate functions of a pair of fuzzy variables. We

consider one such function to obtain the covariance of the pair fuzzy variables and focus on its

computation for common possibility distributions. Under mild assumptions, we derive several

useful covariance formulas for triangular and trapezoidal fuzzy variables, which have potential

applications in quantitative finance problems when we consider the correlations among fuzzy

returns.

Design and Implement of a Scheduling Strategy Based on PSO

Algorithm

Suqin Liu1, Jing Wang1, Xingsheng Li1, Jun Shuo and Huihui Liu1

1China University of Petroleum, China

Abstract. The job scheduling technology is an effective way to achieve resource sharing and

to improve computational efficiency. Scheduling problem has been proved to be NP-complete

problems; Particle Swarm Optimization (PSO) algorithm has demonstrated outstanding performance

in solving such issues. In cognizance of the characteristics of cluster scheduling problem, a

schedule strategy based on PSO was designed and implemented. Comparing with backfilling

algorithm, PSO algorithm can improve the fairness of jobs better. It can avoid the problem

that bigger jobs cant be executed quickly. The speed and accuracy of strategy generation

are improved significantly. The experiment results show that the scheduling strategy based

on PSO algorithm can increase the utilization of the CPU and reduce average response time

83

significantly.

An Examination on Emergence from Social Behavior: A Case in

Information Retrieval

Daren Li1, Muyun Yang1, Sheng Li1 and Tiejun Zhao1


Abstract. The swarm intelligence has been applied to enhancing web search. But few

researches investigate the emergence from the behaviors of users when they forage information

through web search engine. In this paper we study the emergence in users click behaviors

in AOL log and examine its reliability as the key to queries. We introduce kappa statistic

to characterize the emergence through the consistency of users clicks on the same query. By

analyzing the kappa distribution, we reveal that emergence only occurs to the query issued by

a large number of users; and for the queries issued by a fewer users, the clicks are not very

reliable as an emergence. We further infer that the occurrence of emergence in users click

behaviors is somewhat related to the scale of users. It may be unreliable to apply techniques

of swarm intelligence to enhancing web search for all the queries through considering all users

as agents.

A Novel Fault Diagnosis Method Based on Modified Neural

Networks for Photovoltaic Systems

Kuei-Hsiang Chao1, Chao-Ting Chen1, Meng-Hui Wang1 and Chun-Fu Wu2

1National Chin-Yi University of Technology, Taiwan, China

2Industrial Technology Research Institute, Taiwan, China

Abstract. The main purpose of this paper is to propose an intelligent fault diagnostic method

for photovoltaic (PV) systems. First, Solar Pro software package was used to simulate a

photovoltaic system for gathering power generation data of photovoltaic modules during normal

operations and malfunctions. Then, the collected power generation data was used to construct

matter-element models based on extension theory for PV systems. The matter-element model

combines with the neural networks to form an intelligent fault diagnosis system for PV systems.

The proposed fault diagnosis method was adopted to identify the faulty types of a 3.15kW

PV system. The simulation results indicate that the proposed fault diagnosis method can

detect the malfunction types of PV system rapidly and accurately with less time and memory

consumption.

84


Wavelet Packet and Generalized Gaussian Density Based Textile

Pattern Classification Using BP Neural Network

Yean Yin1, Liang Zhang, Miao Jin and Sunyi Xie

1Wuhan University of Science and Engineering, China

Abstract. This paper presents a combined approach to classify the textile patterns based on

wavelet packet decomposition and a BP neural network classifier. On the accurate modeling

of the marginal distribution of wavelet packet coefficients using generalized Gaussian density

(GGD), two parameters are calculated for every level wavelet packet sub-band by moment

matching estimation (MME) or by maximum likelihood estimation (MLE). The parameter

vectors then are taken as the pattern matrix to a BP neural network for recognition. The

proposed method was verified by experiments that using 16 classes of textile patterns, in

which the correct recognition rate is as high as 95.3%.

Air Quality Prediction in Yinchuan by Using Artificial Neural

Networks

Fengjun Li1

1Ningxia University, China

Abstract. A field study was carried out in Yinchuan to gather and evaluate information

about the real environment. O3 (Ozone), PM10 (particle 10 um in diameter and smaller) and

SO2 (sulphur monoxide) constitute the major concern for air quality of Yinchuan. This paper

addresses the problem of the predictions of such three pollutants by using the ANN. Because

ANNs are non-linear mapping structure based on the function of the human brain. They have

been shown to be universal and highly flexible function approximation for any date. These

make powerful tools for models, especially when the underlying data relationship is unknown.

Application of Short-term Load Forecasting Based on Improved

Gray-Markov Residuals Amending of BP Neural Network

Dongxiao Niu1, Cong Xu1, Jianqing Li1 and Yanan Wei

1Norch China Electric Power University, China

Abstract. For the characteristics of short-term load forecasting, we established load forecasting

model based on BP neural network, combined the advantages of gray prediction and Markov

forecasting, and make an amendment for the prediction residual, this has greatly improved

the precision of prediction. Research has shown that neural network and gray-Markov residual

error correction model has the value of popularization and application.

85

Verifying Election Campaign Optimization Algorithm by Several

Benchmarking Functions

Wenge Lv1, Qinghua Xie1, Zhiyong Liu1, Deyuan Li1, Siyuan Cheng, Shaoming Luo

and Xiangwei Zhang

1Guangdong University of Technology, China

Abstract. Election Campaign Optimization (ECO) algorithm is a new heuristic algorithm,

it works by simulating the behavior that the election candidates pursue the highest support

in campaign. The candidates can influence the voters round them. The higher prestige a

candidate comports, the larger effect he has. Voters have to allot their support proportionally

according to the effects imposed by the candidates. Global and local survey-sample to voters

are done to investigate the support of candidates. The proportion of the support to a candidate

from a voter to the sum of the support of the candidate from all voters is the contribution

of a voter to the candidate. The sum of location coordinates of every voters powered by its

contribution is a new location coordinates, it is the next position of the candidate. Such cycle

is done continually until a candidate finds the position of the highest support. In this paper,

several benchmarking functions are used to verify ECO algorithm.

Research on Multi-objective Optimization Design of the UUV

Shape Based on Numerical Simulation

Baowei Song1, Qifeng Zhu1 and Zhanyi Liu1


Abstract. The numerical simulation research of the Underwater Unmanned Vehicle(UUV)

shell shape is carried out by applying the modern fluid dynamics numerical simulation technology.

In the numerical simulation process, while focusing on the characteristics of the UUV shape,

this treatise work properly with its computational model, computational domain, computational

grid and boundary conditions. Based on numerical simulation techniques, the UUV shape is

multi-objectively optimized by integrated fluid dynamics simulation software-fluent on iSIGHT

optimization design platform. Multi-island genetic algorithm is adopted as the optimization

algorithm, and Reynolds-averaged Navier-Stokes equation and turbulence model are as well

applied in the optimization design process. The results show that by reducing the drag

coefficient and the flow noise of the UUV shape, the optimized design has made a significant

improvement in its comprehensive performance, which provides a new way for today’s optimization

design of UUV shape.

86


A Traffic Video Background Extraction Algorithm Based on

Image Content Sensitivity

Bo Qin1, Jingjing Wang1, Jian Gao2, Titi Pang1 and Fang Su1

1Ocean University of China, China

2Qingdao University of Science & Technology, China

Abstract. A Traffic Video Background Extraction Algorithm based on Image Content Sensitivity

(CSBE) is presented in this paper. Different image has different Entropy Energy (EE), the

algorithm analyzes the images content according to it. Firstly, obtain the initial background

image that has the least EE in the moving region through mixture Gaussian background

modeling algorithm. Then, weight factor is selected dynamically by EE and the mixture

Gaussian model (GMM) of every pixel in the current image is updated. Finally, every pixels

value in the background image is updated by weighted average. Experiments show that the

method is simple, robust and well delays the occurrence time of the stationary vehicles in some

degree. Especially, the processing effect is better for the condition that a number of vehicles

into or out of the scene quickly.

A Multimodality Medical Image Fusion Algorithm Based on

Wavelet Transform

Jionghua Teng1, Xue Wang1 and Jingzhou Zhang1


Abstract. According to the characteristics of a medical image, this paper presents a multimodality

medical image fusion algorithm based on wavelet transform. For the low-frequency coefficients

of the medical image, the fusion algorithm adopts the fusion rule of pixel absolute value

maximization; for the high-frequency coefficients, the fusion algorithm uses the fusion rule

that combines the regional information entropy contrast degree selection with the weighted

averaging method. Then the fusion algorithm obtains the fused medical image with inverse

wavelet transform. We select two groups of CT/MRI images and PET/ MRI images to

simulate our fusion algorithm and compare its simulation results with the commonly-used

wavelet transform fusion algorithm. The simulation results show that our fusion algorithm

cannot only preserve more information on a source medical image but also greatly enhance the

characteristic and brightness information of a fused medical image, thus being an effective and

feasible medical image fusion algorithm.

87

Adjusting the Clustering Results Referencing an External Set

Baojia Li1, Yongqian Liu1 and Mingzhu Liu1


Abstract. With the improvement of the information enriching and sharing, it is possible

and valuable to increase the information content of the clustering results referencing external

information. Two concepts, internal set and external set, are put forward in this paper. The

definition of adjusted distance is also given. Based on these, we introduce a method which

adjusts the clustering results of data set referencing the information of an external set. The

effectiveness of the method is illustrated by the results of numeric experiments.

Poster Seesion 2

June 14, 2010(Monday)08:00-12:00 Corridor

Sensitivity Analysis on Single Activity to Network Float in CPM

Network Planning

Zhixiong Su1 and Jianxun Qi1


Abstract. In CPM network planning, one activity consumes float may affect other activities

floats, so that making the gross of networks floats change. The objective of sensitivity analysis

on single activity to network floats is to measure the effect. Aiming at the matter, firstly,

conception of eigenvalue parameter and method of computing the parameter are given; secondly,

sensitivity of single activity to network total float, network safe float, network free float and

network node float is analyzed by using the parameter; thirdly, functions of relation between

quantity consumed by single activity and degree of network floats affected by the activity are

designed, and correctness of the functions are proven; and finally, the feasibility is validated

though example.

Research on Hand Language Video Retrieval

Shilin Zhang1 and Mei Gu

1North China University of Technology, China

Abstract. In order to provide for Chinese disabled person an efficient content-based hand

language video retrieval system, this paper presents a system called DCMR. Content-based

video retrieval is a challenging field, and most research focus on the low level features such as

color histogram, texture and etc. In this paper, we solve the searching problem by high level

features used by hand language recognition. Experiment results on a large of hand language

88


videos show that our searching system performs much better than existing methods on hand

language video searching systems. Compared with the traditional methods, our system reduces

the average searching time by half and the searching precision has doubled.

Research on the Synergy Model between Knowledge Capital and

Regional Economic Development

Cisheng Wu1 and Meng Song1

1Hefei University of Technology, China

Abstract. Regional knowledge capital has become one of the key factors of economic development

of a nation and region, and enhanced the regional competitiveness greatly. In this paper, firstly,

the connotation of the regional knowledge capital is introduced, and then the synergy model of

the system between knowledge capital and regional economic growth is constructed. Secondly,

the synergy of the model is analyzed, and the correctness and operability of the model is

verified through a case which indicate that in the process of regional economic development,

the synergy between knowledge capital and regional economic development has great influence

on the sustainable development of the regional economy.

A New Algorithm of an Improved Detection of Moving Vehicles

Huanglin Zeng1 and Zhenya Wang

1Sichuan University of Science and Engineering, China

Abstract. A new algorithm of vehicle detection on improved background difference method

is presented in this article. An input image frames with fewer moving targets is chosen

as the background initialization frames in the background model initialization phase. A

new background or a foreground image of moving vehicles is distinguished based on the

results of the difference and the value of threshold set. Updating of background is associated

with changes of actual background using dynamic weighting factor adaptively. Extraction

and detection of moving vehicles are completed on the basis of comparing the segmentation

threshold determined. It is shown that the improved background difference model proposed

here is of a higher precision and efficiency in extraction and detection of moving vehicles by

simulation testing.

The Dual Model of a Repairable System

Yunfei Guo1, Maosheng Lai2 and Zhe Yin1

1Yanji University, China


89

Abstract. This paper builds a new class of repairable system model from the dual perspective,

that is, the dual model of the original one, and points out the practical meaning of this model;

at the same time, we give the adjoint operator (A+E)∗ of the operator (A+E), which is the

infinitesimal generator corresponding the Cauchy problem of the former model. We, at last,

give the stable solutions of the new dual model.

A Comprehensive Study of Neutral-point-clamped Voltage

Source PWM Rectifiers

Guojun Tan1, Zongbin Ye1, Yuan Li1, Yaofei Han and Wei Jing

1China University of Mining and Technology, China

Abstract. In this paper, the neutral-point-clamped voltage source PWM rectifier is studied.

The neutral-point balancing problem is analyzed in detail. A new neutral-point balancing

method which is useful for four-quadrant operation of rectifiers is proposed. Besides, a new

simplified three-level SVPWM is presented. The neutral-point potential balance can be easily

realized using this method. Finally, the strategy is verified by experiments.

FPGA-Based Cooling Fan Control System for Automobile Engine

Meihua Xu1, Fangjie Zhao1 and Lianzhou Wang1

1Shanghai University, China

Abstract. In this paper FPGA is used as the master chip to design the new four-phase

sensorless BLDC (brushless direct current) motor control system of automobile engine cooling

fan. The design uses BEMF zero-crossing detection algorithm and the hardware modular

design approach, while the open-loop circuit adopted to raise the frequency and voltage and

the digital phase-shift circuit based on FIPS algorithm are introduced, which makes the whole

control system can work in a variety of adverse environment stably and accurately. Experiments

show that the system has the good error tolerance for the outside interference and meets the

requirements of reliable operation of the motor controller in vehicle engine cooling fan.

Fault Diagnosis of Analog Circuits Using Extension Genetic

Algorithm

Meng-Hui Wang1, Kuei-Hsiang Chao1 and Yu-Kuo Chung1

1National Chin-Yi University of Technology, Taiwan, China

Abstract. This paper proposed a new fault diagnosis method based on the extension genetic

algorithm (EGA) for analog circuits. Analog circuits were difference at some node with the

normal and failure conditions. However, the identification of the faulted location was not

90


easily task due to the variability of circuit components. So this paper presented a novel EGA

method for fault diagnosis of analog circuits, EGA is a combination of extension theory (ET)

and genetic algorithm (GA). In the past, ET had to depend on experiences to set the classical

domain and weight, but setting classical domain and weight were tedious and complicated steps

in classified process. In order to improve this defect, this paper proposes an EGA to find the

best parameter of classical domain and increase accuracy of the classification. The proposed

method has been tested on a practical analog circuit, and compared with other classified

method. The application of this new method to some testing cases has given promising results.

Research on the Coordination Control of Vehicle Electrical

Power Steering System and ABS

Weihua Qin1, Qidong Wang1, Wuwei Chen1 and Shenghui Pan

1Hefei University of Technology, China

Abstract. Based on the multibody dynamics model of Electric Power Steering System(EPS)

and the lateral and longitudinal dynamics vehicle model, the EPS and Anti-lock Braking

System (ABS) sub-controllers and the coordination controller are respectively designed according

to the motion coupling relation between the steering system and the braking system. The

coordination controller supervises and coordinates each sub-controller as the upper controller.

The simulation under Matlab and the vehicle test with hardware-in-the-loop(HIL) based on

LabVIEW have evaluated, tested and verified the vehicle maneuverability and braking performance

under coordination control. The results show that the coordination control effectively improved

the comprehensive performance of vehicle, and the application of multibody dynamics model

and hardware-in-the-loop test were convenient and feasible in the coordination control research

of vehicle chassis.

Optimization Algorithm of Scheduling Six Parallel Activities to

Three Pairs Order Activities

Xiuhua Zhao1, Jianxun Qi1, Shisen Lv1 and Zhixiong Su1


Abstract. It is a special resource allocation problem to adjust 2N paralleling activities into N

activity pairs within resource limits in a CPM (Critical Path Method) network planning; also it

is a hot topic in the field of project scheduling. So far, no simple and effective method has been

designed to solve this problem. In this paper, an optimized algorithm is developed to adjust

6 paralleling activities into 3 activity pairs. Firstly, an algorithm is designed to calculated the

tardiness which can be applied in any circumstance; then, the standard activity pair theory

91

and normalized activity pair theory are created; finally, an optimum method is developed on

the basis of the theories and algorithms as mentioned above.

Bacterial Foraging Optimization Algorithm with Particle Swarm

Optimization Strategy for Distribution Network Reconfiguration

Tianlei Zang, Zhengyou He1 and Deyi Ye1

1Southwest Jiaotong University, China

Abstract. Distribution network reconfiguration for loss minimization is a complex, large-scale

combinatorial optimization problem. In this paper, a novel method called bacterial foraging

optimization algorithm with particle swarm optimization strategy (BF-PSO) algorithm is

applied to solve this problem. To verify the effectiveness of the proposed method, the optimization

calculations of IEEE 69-bus testing system by the presented method are conducted and the

calculation results are compared with pertinent literatures. Simulation results show that the

proposed algorithm possesses fast convergence speed while the quality of solution and stability

is ensured.

Tracking Control of Uncertain DC Server Motors Using Genetic

Fuzzy System

Wei-Min Hsieh, Yih-Guang Leu1, Hao-Cheng Yang and Jian-You Lin

1National Taiwan Normal University, Taiwan, China

Abstract. A controller of uncertain DC server motor is presented by using the fuzzy system

with a real-time genetic algorithm. The parameters of the fuzzy system are online adjusted

by the real-time genetic algorithm in order to generate appropriate control input. For the

purpose of on-line evaluating the stability of the closed-loop system, an energy fitness function

derived from backstepping technique is involved in the genetic algorithm. According to the

experimental results, the genetic fuzzy control scheme performs on-line tracking successfully.

92

Index (c=chair cc=cochair)

Ahmadi, Majid, 31c, 32

Ahmed, Maher A. Sid, 32

Ali, Hamid, 27

Ali, Zulfiqar, 27

An, Jinung, 31

Ariffin, Junaidah, 32

Azuela, Humberto Sossa, 25

Baig, Rauf, 27

Ban, Xiaojuan, 29

Battle, Kimberly, 27

Bilal, Mohsin, 27

Biswas, Kanad, 31

Braude, David A., 26

Cai, Yingpeng, 31

Cao, Jianting, 33

Chao, Kuei-Hsiang, 37, 39

Charles, Jesse St., 34

Chen, Chao-Ting, 37

Chen, Hanning, 29

Chen, Junliang, 32

Chen, Tanggong, 25

Chen, Wuwei, 39

Chen, Xuebo, 34

Cheng, Jihong, 33

Cheng, Siyuan, 37

Cheng, Xi, 34

Cheng, Xu, 28

Chi, Zheru, 30

Chiang, Tsung-Che, 32

Choi, Kyung-Sik, 31, 32

Choi, Yun Won, 31

Chung, Yu-Kuo, 39

Cui, Xiaohui, 34, 34c

Cui, Yu, 32

Dai, Hui, 35

Dai, Xiaodong, 33

Dai, Yanyan, 32, 35

Dastjerdi, Ahmad Baraani, 35

Dehkordi, Zahra Karimi, 35

Deng, Liwen, 34

Dong, Gaifang, 29

Dong, Wanli, 33

Du, Haohua, 31, 33

Du, Yu, 27

Duan, Haibin, 27

Fan, Zizhu, 33

Fang, Zhimin, 30

Fang, Zhixiang, 28

Feng, Ding, 35, 35c

Fu, Zhichao, 34

Gao, Feng, 34

Gao, Jian, 38

Gao, Jiaquan, 26, 30

Gao, Wen-jing, 27

Gao, Yuelin, 26

93

Gu, Mei, 39

Guo, William W, 29

Guo, Yunfei, 39

Hamdan, Mohammad, 29

Hamdani, Syed, 30

Han, Fei, 32

Han, Ki Joon, 31

Han, Yaofei, 39

He, Guixia, 30

He, Hua, 29

He, Qian, 32

He, Zhengyou, 40

Hou, Guolian, 26

Hou, Jinliang, 33

Hsieh, Wei-Min, 40

Hu, Kunyuan, 29

Hu, Luoke, 26, 30

Hu, Zhengwei, 31

Huang, Xi, 37c, 39c

Huo, Pengfei, 25

Hwang, Wen-Jyi, 32, 32c

Igarashi, Yasutaka, 34

Islam, Mohammed, 32

Jain, Chakresh, 31

Janghel, RR, 33

Jiang, Dan, 28

Jiang, Xiangang, 33

Jiang, Zhaofeng, 33

Jiang, Zhaogeng, 29c

Jiao, Licheng, 29

Jin, Miao, 37

Jing, Wei, 39

Ju, Shiguang, 32

Kala, Rahul, 33

Kaneko, Toshinobu, 34

Kang, Tae Hoon, 31

Kazadi, Sanza, 25

Khan, Farrukh, 27, 27c, 30

Khan, Salabat, 27

Kim, Yoon-Gu, 31

Lai, Maosheng, 39

Lee, Suk Gyu, 31

Lee, Suk-Gyu, 31, 32, 35

Lee, Wei-Po, 29

Lei, Fanfan, 26

Leu, Yih-Guang, 40

Li, Baojia, 38

Li, Daren, 37

Li, Deyuan, 37

Li, Fang, 33

Li, Fengjun, 37

Li, Geng, 32

Li, Hao, 26

Li, Huimei, 30

Li, Jianqing, 37

Li, Li, 26

Li, Qingshan, 28

Li, Qiuping, 28

Li, Rongjun, 26

Li, Sheng, 37

Li, Shouyong, 35

Li, Wentao, 26

Li, Xihua, 27

Li, Xingsheng, 37

Li, Xiong, 31

Li, Yicheng, 29

Li, Yuan, 39

94


Li, Yuanxiang, 32

Liang, Jianhui, 29

Liao, Renjie, 27

Liao, Wudai, 32

Licon, Beatriz Aurora Garro, 25

Lin, Jian-You, 40

Liou, Cheng-Dar, 27

Liu, Caiyun, 27

Liu, Chun-Hung, 27

Liu, He, 28

Liu, Huihui, 37

Liu, Jin, 35

Liu, Kun, 25

Liu, Mingzhu, 38

Liu, Ruochen, 29

Liu, Suqin, 37

Liu, Tao, 30

Liu, Xiangjie, 35

Liu, Xiangkai, 30

Liu, Xiao, 35c, 36

Liu, Xiaoyong, 26

Liu, Yankui, 34, 37

Liu, Yawei, 31

Liu, Ying, 33

Liu, Yongqian, 38

Liu, Yu, 26

Liu, Yujie, 33

Liu, Zhanyi, 38

Liu, Zhiyong, 37

Lu, Qiang, 26

Luo, Shaoming, 37

Lv, Jun, 34

Lv, Mingwei, 26

Lv, Shisen, 35, 39

Lv, Wenge, 37

Ma, Miao, 29

Ma, Miaomiao, 35

Ma, Xiujun, 30c, 31

Ma, Yunfeng, 34

Marwala, Tshilidzi, 27

Meng, Xiangwu, 32

Meng, Xiaohong, 27

Mo, Hongwei, 28, 28c

Moein, Sara, 34

Mohamed, Azlinah, 32

Nelwamondo, Fulufhelo V., 27

Niu, Ben, 26, 26c, 29

Niu, Dongxiao, 37

Niu, Manchun, 29

Pan, Kailing, 34

Pan, Shenghui, 39

Pan, Zijian, 31

Pang, Lingling, 25

Pang, Shanchen, 29, 29c

Pang, Titi, 28, 38

Park, Andrew, 25

Park, James, 25

Potok, Thomas, 34

Prasad, Yamuna, 31

Qi, Jianxun, 35, 39

Qian, Dianwei, 35

Qin, Bo, 28, 38

Qin, Jin, 29

Qin, Quande, 26

Qin, Rui, 33, 33c

Qin, Weihua, 39

Qiu, Xuena, 26

Qteish, Abdallah, 29

95

Qu, Dongcai, 33

Ruan, Da, 34

Saitov, Dilshat, 31

Shafiq, Sarah, 30

Shang, Yanlei, 32

Sharif, Muhammad, 27, 27c

Shi, Qiwei, 33

Shukla, Anupam, 33

Shuo, Jun, 37

Soliman, M. Sami, 27

Song, Baowei, 38

Song, Meng, 39

Su, Fang, 28, 38

Su, Zhixiong, 35, 39

Sun, Baolin, 35

Sun, Gongxian, 34

Tan, Bianyou, 35

Tan, Guanzheng, 27

Tan, Guojun, 39

Tan, Ying, 25, 28, 30, 30c

Tanaka, Toshihisa, 33

Tang, Lina, 29

Tang, Lixin, 27, 28

Tang, Qi, 35

Tang, Yanfeng, 30

Teng, Jionghua, 25, 38

Tian, Hongpeng, 29

Tiwari, Ritu, 33

Tran, Viet-Hong, 35, 35cc

Vazquez, Roberto Antonio, 25

Wan, Chunqiu, 26

Wang, Haiqi, 33

Wang, Hongbo, 31, 31c

Wang, Jia-ning, 30

Wang, Jiangfeng, 32

Wang, Jianing, 30

Wang, Jiayao, 31

Wang, Jing, 37

Wang, Jingjing, 38

Wang, Jun, 26

Wang, Junyan, 32

Wang, Lianzhou, 39

Wang, Meng-Hui, 37, 39

Wang, Miaomiao, 26

Wang, Peng, 35

Wang, Qian, 34

Wang, Qidong, 39

Wang, Rubin, 33

Wang, Rui, 26

Wang, Suhuan, 25

Wang, Tianmiao, 31

Wang, Tong, 35

Wang, Wei, 30

Wang, Xianpeng, 27

Wang, Xiaohan, 37

Wang, Xue, 25, 38

Wang, Yingxin, 26

Wang, Yong, 34

Wang, Zhenya, 39

Wei, Cheng, 34

Wei, Hongxing, 31

Wei, Yanan, 37

Wen, Mi, 34

Wu, Chun-Fu, 37

Wu, Cisheng, 39

Wu, Quanjun, 30

Wu, Xiaoli, 34

96


Wu, Zhanbin, 28

Wyk, Anton van, 26

Xiang, Lan, 30

Xiao, Ming, 34

Xie, Kunqing, 31

Xie, Qinghua, 37

Xie, Sunyi, 37

Xing, Bo, 27, 27cc

Xing, Kangzheng, 29

Xing, Wenhua, 30

Xiong, Jie, 27

Xiong, Shengwu, 28

Xu, Baogen, 33

Xu, Chang, 35

Xu, Chi, 30

Xu, Cong, 37

Xu, Guanjun, 35

Xu, Jianping, 34

Xu, Lifang, 28

Xu, Meihua, 39

Xu, Zhiguang, 32, 35

Yan, Xinchi, 37

Yan, Zhaofa, 26

Yang, Cheng, 35

Yang, Geng, 26

Yang, Hao-Cheng, 40

Yang, James, 25

Yang, Muyun, 37

Yang, Tengfei, 33

Yang, Yanlong, 34

Yao, Yuan, 26

Ye, Deyi, 40

Ye, Zongbin, 39

Yin, Peng-Yeng, 25c, 26

Yin, Yean, 37

Yin, Yixin, 29

Yin, Zhe, 39

Ying, Weiqin, 32

You, Chuanchuan, 35

Yu, Fahong, 32

Yu, Hui, 31

Yu, Tsung-Yi, 32

Yuan, Yongxin, 35

Yun, Haishun, 28

Yusoff, Marina, 32, 32cc

Zamanifar, Kamran, 35

Zang, Tianlei, 40

Zeng, Huanglin, 39

Zeng, XiaoJun, 32

Zhang, E, 34

Zhang, Hong, 25, 25c

Zhang, Jian, 30, 36

Zhang, Jianhua, 26

Zhang, Jie, 25

Zhang, Jin, 33, 33c

Zhang, Jingzhou, 25, 38

Zhang, Liang, 37

Zhang, Lijie, 25

Zhang, Pengtao, 30

Zhang, Ping, 31

Zhang, Ruizhi, 33

Zhang, Shilin, 39

Zhang, Tiecheng, 31

Zhang, Xiangwei, 37

Zhang, Xing, 26, 37

Zhang, Xueping, 31, 33

Zhao, Fangjie, 39

Zhao, Guangcai, 33

97

Zhao, Tiejun, 37

Zhao, Xiaoyu, 30

Zhao, Xiuhua, 35, 39

Zheng, Lin, 30

Zhou, Bingyu, 35

Zhou, Jin, 30

Zhou, Tian, 29

Zhou, Wei, 33

Zhu, Bingfeng, 30

Zhu, Qifeng, 38

Zhu, Qunxiong, 33

Zhu, Yuanchun, 28, 28cc

Zhu, Yunlong, 29

Zhuo, Yiqin, 34

Zong, Xinlu, 28

98

Final Program

Documents

Transcript of Final Program