Post on 24-Jan-2021
Research ArticleFuzzy Modeling and Control for a Class ofInverted Pendulum System
Liang Zhang,1 Xiaoheng Chang,1 and Hamid Reza Karimi2
1 College of Engineering, Bohai University, Jinzhou 121013, China2Department of Engineering, Faculty of Engineering and Science, The University of Agder, 4898 Grimstad, Norway
Correspondence should be addressed to Liang Zhang; md18638@126.com
Received 14 December 2013; Accepted 24 December 2013; Published 29 January 2014
Academic Editor: Ming Liu
Copyright Β© 2014 Liang Zhang et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Focusing on the issue of nonlinear stability control system about the single-stage inverted pendulum, the T-S fuzzy model isemployed. Firstly, linear approximation method would be applied into fuzzy model for the single-stage inverted pendulum. Atthe same time, for some nonlinear terms which could not be dealt with via linear approximation method, this paper will adopt fanrange method into fuzzy model. After the T-S fuzzy model, the PDC technology is utilized to design the fuzzy controller secondly.Numerical simulation results, obtained byMatlab, demonstrate the well-controlled effectiveness based on the proposedmethod forthe model of T-S fuzzy system and fuzzy controller.
1. Introduction
Traditional control theory has perfect control ability forexplicitly controlled system, however, which is a little weak todescribe too complex or difficult systemaccurately.Therefore,many researchers seek ways to resolve this problem; thoseresearchers have also focused on fuzzy mathematics andapplied it to control problems. Zadeh [1] created fuzzy math-ematics on an uncertainty system of control which is greatcontribution. Since the 70s, some practical controllers appearin succession, so that we have a big step forward in the controlfield. A number of control design approaches using adaptivecontrol [2β4], sliding mode control [5, 6],π»
β[7β9], optimal
control [10β12], control based data driven [10, 13β15], andfuzzy control [16, 17]. The inverted pendulum system iscontrolled by the method of fuzzy control and realizes steadycontrol.The inverted pendulum is a typical automatic controlin the field of controlled object [18], which is multivariableand nonlinear and strong coupling characteristics, and soon. The inverted pendulum system reveals a natural unstableobject, which can accomplish the stability and good perfor-mance by the control methods.
For the stability control of inverted pendulum system,the establishment of the model takes an important role. T-S fuzzy control [19] is the most popular one of the most
promising methods based on modeling of fuzzy controlresearch platform. At present, the T-S fuzzy control is oneof the methods for nonlinear system control research [20],which is very popular. Based on T-S fuzzy model of invertedpendulum system modeling and control have a certainresearch. For inverted pendulum system based on T-S fuzzymode, there are two methods [21]: the first one is the fan ofnonlinear method. Although this method has high precisionin describing the nonlinear system, it obtains many fuzzyrules.Thus it brings to the controller design difficulty, especi-ally for the nonlinear term system. The second one is linearapproximation modeling method, the method at the expenseof the modeling accuracy and less number of rules of T-Sfuzzy model. Since the secondmethod can obtain a simple T-S fuzzy model, so in the inverted pendulum systemmodelingit is widely applied, but there is a very important problem,which is that if for one type of inverted pendulum system itcontains the approximate method to deal with the nonlinearterm, then the fuzzy modeling becomes the key to study.
Based on the above analysis anddiscussion, this thesiswillcarry the fuzzy modeling and control on inverted pendulumsystem of complex nonlinear term. For this point, sectornonlinear and linear approximation method will be adoptedin the T-S fuzzy modeling of some inverted pendulums andthe design of fuzzy controller.The fuzzymodeling and control
Hindawi Publishing CorporationAbstract and Applied AnalysisVolume 2014, Article ID 936868, 6 pageshttp://dx.doi.org/10.1155/2014/936868
2 Abstract and Applied Analysis
method can achieve the stability control of the single invertedpendulum system through the simulation.
2. Fuzzy Modeling for the InvertedPendulum System
Assume that the carβs quality is π, the pendulumβs quality isπ, the pendulumβs length is π, the pendulumβs angle is π at an
instant (the angle between the pendulum rod and the verticaldirection), the initial displacement is π₯, π = 9.8 m/s2 is thegravity constant, the level for control is forced acting on thecar is πΉ, π = 1/(π + π), and the inverted pendulumβs statespace is as follows:
(οΏ½οΏ½1(π‘)
οΏ½οΏ½2(π‘)
) = (
π₯2(π‘)
1
(4π/3) β πππcos2 (π₯1(π‘))
[π sin (π₯1(π‘)) β
ππππ₯2
2(π‘) sin2 (π₯
1(π‘))
2β π₯2(π‘) β ππ’ (π‘)]
) , (1)
where π is π₯1(π‘), π is π₯
2(π‘), πΉ = π’(π‘), and π₯
1(π‘) β (0, Β±π/2),
π₯2(π‘) β [βπΌ, πΌ].When π₯
1(π‘) = Β±π/2, the system is uncontrollable, so
we take π₯1(π‘) β [β88
β
, 88β
] as the range. For this invertedpendulum system, T-S fuzzy model can be considered asfollows:
π π
: if π₯1(π‘) is ππ
1, . . . π₯π(π‘) is ππ
π
then οΏ½οΏ½ (π‘) = π΄ππ₯ (π‘) + π΅
ππ’ (π‘) ,
(2)
where π₯(π‘) = [π₯1(π‘) π₯2(π‘) β β β π₯
π(π‘)]π
β π π is the state vari-
ables for the fuzzy system,πππis the fuzzy sets and where π =
1, 2, . . . , π, π = 1, 2, . . . , π, the input vector is π’(π‘) β π π, π΄πβ
π πΓπ, π΅
πβ π πΓπ are coefficient matrix for the system. The
number of fuzzy rules for the system is π.The total fuzzy control system is as follows:
οΏ½οΏ½ (π‘) =βπ
π=1ππ(π₯ (π‘)) (π΄
ππ₯ (π‘) + π΅
ππ’ (π‘))
βπ
π=1ππ(π₯ (π‘))
, (3)
where ππ(π₯(π‘)) = β
π
π=1ππ
π(π₯π(π‘)), and π
π
π(π₯π(π‘)) is denotes
the membership degree, and where π₯π(π‘) for π
π
π. The
βπ(π₯(π‘)) = π
π(π₯(π‘))/β
π
π=1ππ(π₯(π‘)) and (2) will be as follows:
οΏ½οΏ½ (π‘) =
π
β
π=1
βπ(π₯ (π‘)) (π΄
ππ₯ (π‘) + π΅
ππ’ (π‘)) , (4)
where βπ(π₯(π‘)) β₯ 0 and β
π
π=1βπ(π₯(π‘)) = 1.
There is an important nonlinear term in this invertedpendulum system, in other words π₯
2
2(π‘)sin2(π₯
1(π‘)), which
should be paid more attention. The nonlinear term cannotbe conducted through the linear approximation method onthis inverted pendulum system.Thus, the thesis will combinethe linear approximation method with the fan of nonlinearmethod to establish the fuzzymodel.Theprocess is as follows.
(1) If π₯1(π‘) is about 0, through approximate treatment
the systemwith the linear approximationmethod, thefuzzy model of system can be obtained as follows:
then
(οΏ½οΏ½1(π‘)
οΏ½οΏ½2(π‘)
) = (
0 1
π
(4π/3) β πππ
β1
(4π/3) β πππ
)(π₯1(π‘)
π₯2(π‘)
)
+ (
0
βπ
(4π/3) β πππ
)π’ (π‘)
(οΏ½οΏ½1(π‘)
οΏ½οΏ½2(π‘)
) = (
0 1
3π
4π β 3πππ
β3
4π β 3πππ
)(π₯1(π‘)
π₯2(π‘)
)
+ (
0
β3π
4π β 3πππ
)π’ (π‘) .
(5)
(2) If π₯1(π‘) is about Β±π/2, and consider the fan of nonli-
near method, and π§(π‘) = π₯2(π‘)sin2(π₯
1(π‘)),
then
maxπ₯1(π‘), π₯2(π‘)
π§ (π‘) = π₯2(π‘) sin2 (π₯
1(π‘)) β‘ π
1= 0.2595. (6)
Then
minπ₯1(π‘), π₯2(π‘)
π§ (π‘) = π₯2(π‘) sin2 (π₯
1(π‘)) β‘ π
2= β0.2595. (7)
(1) If π§(π‘) is π1, through the linear approximation
method and the fan of nonlinear method toapproximate treatment, the fuzzy model of sys-tem can be obtained as follows:
(οΏ½οΏ½1(π‘)
οΏ½οΏ½2(π‘)
) = (
0 1
π (2/π)
4π/3β(
1
4π/3
ππππ1
2+
1
4π/3))(
π₯1(π‘)
π₯2(π‘)
)
+ (
0
βπ
4π/3
)π’ (π‘) ,
(οΏ½οΏ½1(π‘)
οΏ½οΏ½2(π‘)
)=(
0 1
3π
2ππβ(
3πππ1
8+3
4π))(
π₯1(π‘)
π₯2(π‘)
)+(
0
β3π
4π
) π’ (π‘) .
(8)
Abstract and Applied Analysis 3
Rule 2
00
Rule 1
1
βπ
2
π
2
Figure 1: Membership functions of two-rule model.
(2) If π§(π‘) is π2, through the linear approximation
method and the fan of nonlinear method toapproximate treatment, the fuzzy model of sys-tem can be obtained as follows:
(οΏ½οΏ½1(π‘)
οΏ½οΏ½2(π‘)
)=(
0 1
π (2/π)
4π/3β(
1
4π/3
ππππ2
2+
1
4π/3))(
π₯1(π‘)
π₯2(π‘)
)
+ (
0
βπ
4π/3
)π’ (π‘) ,
(οΏ½οΏ½1(π‘)
οΏ½οΏ½2(π‘)
)=(
0 1
3π
2ππβ(
3πππ2
8+3
4π))(
π₯1(π‘)
π₯2(π‘)
)+(
0
β3π
4π
)π’ (π‘) .
(9)
Here define membership functions. For the part of lin-ear approximation, the membership function is shown asFigure 1.
Rule 1: Consider
π»1(π‘) =
{{
{{
{
2
ππ₯1(π‘) + 1, (β
π
2β€ π₯1(π‘) β€ 0)
β2
ππ₯1(π‘) + 1, (0 < π₯
1(π‘) β€
π
2)
. (10)
Rule 2: Consider
π»2(π‘) = 1 β π»
1(π‘) . (11)
Figure 2 is the membership function for the fan of nonlinearmethod. π§(π‘) can be rewritten as π§(π‘) = β
2
π=1πΈπ(π§(π‘))π
π, where
πΈ1(π§(π‘)) = (π§(π‘)βπ
2)/(π1βπ2) andπΈ
2(π§(π‘)) = (π
1βπ§(π‘))/(π
1βπ2).
The membership functions πΈ1(π§(π‘)) and πΈ
2(π§(π‘)) will meet
the equation πΈ1(π§(π‘)) + πΈ
2(π§(π‘)) = 1.
In conclusion, the finally fuzzy model for the system willbe shown as follows.
Rule 1: if π₯1(π‘) tends to 0, then οΏ½οΏ½(π‘) = π΄
1π₯(π‘)+π΅
1π’(π‘).
Rule 2: if π₯1(π‘) tends to Β±(π/2)(|π₯
1(π‘)| < π/2) and
π§(π‘) takes the maximum value, then οΏ½οΏ½(π‘) = π΄2π₯(π‘) +
π΅2π’(π‘).
E2(z(t)) E1(z(t))
Negative Positive
βa a0
1
z(t)0
Figure 2: Membership function for the fan of nonlinear method.
Table 1: Function parameters.
Parameter Function Valueπ Themass of the cart 1.096 kgπ Themass of the pendulum 0.109 kgπ The length of the pendulum 0.25m
πThe angle of the pendulum fromthe vertical
πΉ The force applied to the cart
Rule 3: if π₯1(π‘) tends to Β±(π/2)(|π₯
1(π‘)| < π/2) and
π§(π‘) takes the minimum value, then οΏ½οΏ½(π‘) = π΄3π₯(π‘) +
π΅3π’(π‘).
The function and value of every parameter are shown inTable 1.
All the parameters defined in Table 1 are taken to account,then the system coefficient matrix can be obtained as follow
π΄1= (
0 1
3π
4π β 3πππ
β3
4π β 3πππ
) = (0 1
31.5397 β3.2183) ,
π΅1= (
0
β3π
4π β 3πππ
) = (0
β2.6708) ,
π΄2= (
0 1
3π
2ππβ (
3πππ1
8+
3
4π)) = (
0 1
18.7166 β3.0088) ,
π΅2= (
0
β3π
4π
) = (0
β2.4896) ,
π΄3= (
0 1
π (2/π)
4π/3β(
1
4π/3
ππππ2
2+
1
4π/3))
= (0 1
18.7166 β2.9912) ,
π΅3= (
0
βπ
4π/3
) = (0
β2.4896) .
(12)
4 Abstract and Applied Analysis
0 5 10 15 20β8
β6
β4
β2
0
2
4
Time (s)
x(d
eg)
x1(t)
x2(t)
Figure 3: Simulation result without the controller.
The model rules can be represented as follows:
β1(π‘) = π»
1(π‘) ,
β2(π‘) = π»
2(π‘) Γ πΈ
1(π§ (π‘)) = π»
2(π‘) Γ [
π§ (π‘) β π2
π1β π2
] ,
β3(π‘) = π»
2(π‘) Γ πΈ
2(π§ (π‘)) = π»
2(π‘) Γ [
π1β π§ (π‘)
π1β π2
] .
(13)
For the T-S model of the control object, a parallel distributedcompensation control scheme (PDC) is employed. And theregulations are described as follow:
π π
: if π₯1(π‘) is ππ
1, . . . π₯π(π‘) is ππ
π
then π’ (π‘) = πΎππ₯ (π‘) .
(14)
Here the fuzzy controller and the fuzzy system adopt the samefuzzy rule. The overall model for the fuzzy controller is asfollows:
π’ (π‘) =
π
β
π=1
βπ(π₯ (π‘)) πΎ
ππ₯ (π‘) . (15)
The closed control system can be obtained by combining (2)and (4):
οΏ½οΏ½ =
π
β
π=1
π
β
π=1
βπβπ(π΄π+ π΅ππΎπ) π₯. (16)
3. Based on Linear Matrix Inequalities (LMI)and the Matlab Simulation
Without the controller, the simulation output curves ofthe angular velocity and angular acceleration are shown inFigure 3.
Figure 3 shows that the inverted pendulum system isunstable without the controller.
In the following, by using the linear matrix inequalitiestechnique [23], the fuzzy controller is designed.
Let us define the Lyapunov function as (π₯(π‘)) =
π₯π
(π‘)ππ₯(π‘),π > 0, then the stable criterion of the system for(16) is as follows:π
β
π=1
π
β
π=1
βπβπ(π΄π+ π΅ππΎπ)π
π + π
π
β
π=1
π
β
π=1
βπβπ(π΄π+ π΅ππΎπ) < 0.
(17)
Defineπ = πβ1; we can obtain (18) by multiplyingπ on both
sides contemporary:
π
π
β
π=1
π
β
π=1
βπβπ(π΄π+ π΅ππΎπ)π
+
π
β
π=1
π
β
π=1
βπβπ(π΄π+ π΅ππΎπ)π < 0.
(18)
After defining πΎππ = π
π, the system stability discriminant
conditions can be obtained. And this will guarantee a positivedefinite matrix π and matrix π
πcan be searched, then the
following matrix inequality can be established:
ππ΄π
π+ ππ
ππ΅π
π+ π΄ππ + π΅
πππ< 0,
ππ΄π
π+ππ
ππ΅π
π+π΄ππ+π΅πππ+ππ΄π
π+ππ
ππ΅π
π+π΄ππ+π΅πππ< 0,
π = 1, 2, . . . , π, π < π,
(19)
where the stable controller will be obtained from (20):
πΎπ= πππβ1
. (20)
Through solving (19) by linear matrix inequality with LMI ofMatlab [24], we can obtain
π = [1.1127 β0.4874
β0.4874 1.9975] ,
π1= [14.6725 β 7.9846] ,
π2= [9.7868 β 5.9077] ,
π3= [9.7836 β 5.8945] .
(21)
Moreover, taking advantage of (20), the fuzzy controller gainwill be obtained and shown as follows:
πΎ1= [12.8038 β 0.8733] ,
πΎ2= [8.3975 β 0.9086] ,
πΎ3= [8.3975 β 0.9020] .
(22)
Put the controller gain into (16); design the simulationprogram in the simulink environment. Here the initial valueπ₯(0) = [β0.01 β0.1] is selected. The results of the fuzzy con-trol simulation of the level single inverted pendulum systemare shown in Figures 4, 5, and 6.
Simulation results show that the system responses con-verge to the equilibriumpoint, which indicates that the designof the controller is stable.
Abstract and Applied Analysis 5
0 5 10 15 20β0.06
β0.04
β0.02
0
0.02
0.04
Time (s)
x1(t)
x1(t)
Figure 4: Simulation result of the angle.
0 5 10 15 20β0.1
β0.05
0
0.05
0.1
Time (s)
x2(t)
x2(t)
Figure 5: Simulation result of the angular velocity.
0 5 10 15 20β0.8
β0.6
β0.4
β0.2
0
0.2
0.4
0.6
Time (s)
u(t)
u(t)
Figure 6: Simulation result of the controller.
4. Conclusion
This thesis takes a class of an inverted pendulum systemas the research object. The system fuzzy model was estab-lished by the methods that combining with the linearizationapproximation processing and fan-shaped interval, and thenthe fuzzy controller was designed. Matlab-Simulink soft-ware toolbox was employed to be on computer simulation.The results show that it achieved a stable control of thesingle-stage inverted pendulum system through fuzzy controlmethod on the basis of this fuzzy model. This model hasthe advantages of less fuzzy rules, high precision, and simplestructure. The research results can provide an effective wayfor the subsequent instability in other nonlinear systemmodeling and fuzzy control.
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper.
Acknowledgments
This work is supported by the National Nature ScienceFoundation of China under Grant (Project no. 61104071),by the Program for Liaoning Excellent Talents in University,China, under Grant (Project no. LJQ2012095), by the OpenProgram of the Key Laboratory of Manufacturing IndustrialIntegrated Automation, Shenyang University, China, underGrant (Project no. 1120211415), and Natural Science Foun-dation of Liaoning, China (Project no. 2013020044). Theauthors highly appreciate the above financial supports.
References
[1] L. A. Zadeh, βFuzzy sets,β Information and Computation, vol. 8,pp. 338β353, 1965.
[2] B. Xiao, Q. Hu, and Y. Zhang, βAdaptive sliding mode faulttolerant attitude tracking control for flexible spacecraft underactuator saturation,β IEEE Transactions on Control SystemsTechnology, vol. 20, no. 6, pp. 1605β1612, 2012.
[3] B. Xiao, Q. Hu, and Y. Zhang, βFault-tolerant attitude controlfor flexible spacecraft without angular velocity magnitudemeasurement,β Journal of Guidance, Control, andDynamics, vol.34, no. 5, pp. 1556β1561, 2011.
[4] Y. Zhao, W. Sun, and H. Gao, βRobust control synthesis for seatsuspension systems with actuator saturation and time-varyinginput delay,β Journal of Sound and Vibration, vol. 329, no. 21, pp.4335β4353, 2010.
[5] L. Wu, X. Su, and P. Shi, βSliding mode control with boundedπΏ2gain performance of Markovian jump singular time-delay
systems,β Automatica, vol. 48, no. 8, pp. 1929β1933, 2012.[6] Q. Hu, βSliding modemaneuvering control and active vibration
damping of three-axis stabilized flexible spacecraft with actua-tor dynamics,β Nonlinear Dynamics, vol. 52, no. 3, pp. 227β248,2008.
[7] G. Pujol, βReliable π»β
control of a class of uncertain inter-connected systems: an LMI approach,β International Journal ofSystems Science, vol. 40, no. 6, pp. 649β657, 2009.
6 Abstract and Applied Analysis
[8] Q. Zheng and F. Wu, βNonlinear π»β
control designs withaxisymmetric spacecraft control,β Journal of Guidance, Control,and Dynamics, vol. 32, no. 3, pp. 850β859, 2009.
[9] W. Sun, H. Gao Sr., and O. Kaynak, βFinite frequencyπ»βcon-
trol for vehicle active suspension systems,β IEEETransactions onControl Systems Technology, vol. 19, no. 2, pp. 416β422, 2011.
[10] S. Yin, H. Luo, and S. Ding, βReal-time implementation of fault-tolerant control systems with performance optimization,β IEEETransactions on Industrial Electronics, vol. 61, no. 5, pp. 2402β2411, 2013.
[11] M. V. Levskii, βOptimal control of reorientation of a spacecraftusing free trajectory method,β Cosmic Research, vol. 49, no. 2,pp. 131β149, 2011.
[12] E. M. Queen and L. Silverberg, βOptimal control of a rigid bodywith dissimilar actuators,β Journal of Guidance, Control, andDynamics, vol. 19, no. 3, pp. 738β740, 1996.
[13] S. Yin, S. X. Ding, A. H. A. Sari, and H. Hao, βData-drivenmonitoring for stochastic systems and its application on batchprocess,β International Journal of Systems Science, vol. 44, no. 7,pp. 1366β1376, 2013.
[14] S. Yin, S. X. Ding, A. Haghani, H. Hao, and P. Zhang, βAcomparison study of basic data-driven fault diagnosis andprocess monitoring methods on the benchmark TennesseeEastman process,β Journal of Process Control, vol. 22, no. 9, pp.1567β1581, 2012.
[15] S. Yin, X. Yang, and H. R. Karimi, βData-driven adaptiveobserver for fault diagnosis,β Mathematical Problems in Engi-neering, vol. 2012, Article ID 832836, 21 pages, 2012.
[16] S.-J. Dong, Y.-Z. Li, J. Wang, and J. Wang, βFuzzy incrementalcontrol algorithm of loop heat pipe cooling system for space-craft applications,βComputers &Mathematics with Applications,vol. 64, no. 5, pp. 877β886, 2012.
[17] Z.Gao, X. Shi, and S. X.Ding, βFuzzy state/disturbance observerdesign for T-S fuzzy systems with application to sensor faultestimation,β IEEE Transactions on Systems, Man, and Cybernet-ics B, vol. 38, no. 3, pp. 875β880, 2008.
[18] S. Udomsin, βInverted pendulum control,β KKU EngineeringJournal, vol. 25, pp. 41β65, 2013.
[19] C.-H. Fang, Y.-S. Liu, S.-W. Kau, L. Hong, and C.-H. Lee, βAnew LMI-based approach to relaxed quadratic stabilization ofT-S fuzzy control systems,β IEEE Transactions on Fuzzy Systems,vol. 14, no. 3, pp. 386β397, 2006.
[20] H. O.Wang, K. Tanaka, andM. F. Griffin, βAn approach to fuzzycontrol of nonlinear systems: stability and design issues,β IEEETransactions on Fuzzy Systems, vol. 4, no. 1, pp. 14β23, 1996.
[21] H. O. Wang, K. Tanaka, and M. Griffin, βParallel distributedcompensation of nonlinear systems by Takagi-Sugeno fuzzymodel,β in Proceedings of the International Joint Conference ofthe 4th IEEE International Conference on Fuzzy Systems and the2nd International Fuzzy Engineering Symposium, vol. 2, pp. 531β538, March 1995.
[22] K. Tanaka, T. Ikeda, and H. O. Wang, βFuzzy regulators andfuzzy observers: relaxed stability conditions and LMI-baseddesigns,β IEEE Transactions on Fuzzy Systems, vol. 6, no. 2, pp.250β265, 1998.
[23] S. Boyd, L. EI Ghaoul, E. Feron, and V. Balakrishnan, LinearMatrix Inequalities in System and ControlTheory, vol. 15 of Stud-ies in Applied Mathematics, Society for Industrial Mathematics,1994.
[24] P. Gahinet, A. Nemirovski, A. J. Laub, and M. Chilali, LMIControl Toolbox, vol. 1 of Userβs Guide Version, Manuales deMATLAB, 1992.
Submit your manuscripts athttp://www.hindawi.com
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttp://www.hindawi.com
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
CombinatoricsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com
Volume 2014 Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Stochastic AnalysisInternational Journal of