AN OPTIMIZED DC-DC CONVERTER FOR ELECTRIC VEHICLE … · 2017-09-25 · KY Boost Converter, a...
Transcript of AN OPTIMIZED DC-DC CONVERTER FOR ELECTRIC VEHICLE … · 2017-09-25 · KY Boost Converter, a...
http://www.iaeme.com/IJMET/index.asp 173 [email protected]
International Journal of Mechanical Engineering and Technology (IJMET) Volume 8, Issue 9, September 2017, pp. 173–182, Article ID: IJMET_08_09_018
Available online at http://www.iaeme.com/IJMET/issues.asp?JType=IJMET&VType=8&IType=9
ISSN Print: 0976-6340 and ISSN Online: 0976-6359
© IAEME Publication Scopus Indexed
AN OPTIMIZED DC-DC CONVERTER FOR
ELECTRIC VEHICLE APPLICATION
Prithivi K and Sathyapriya M
Department of Electrical and Electronics Engineering,
Kongu Engineering College, Tamilnadu, India
Ashok Kumar L
Department of Electrical and Electronics Engineering,
PSG College of Technology, Tamilnadu, India
ABSTRACT:
KY Boost Converter, a modern invention in the field of non-isolated DC-DC boost
converter which is identified for minimum voltage ripple. KY Boost Converter is the
combination of KY Converter and traditional Boost Converter. Such a converter has
continuous input and output inductor current, different from the traditional boost
converter. And hence this converter is very suitable for very low-ripple applications.
The open loop controller and Optimization techniques based closed loop controller
are designed in MATLAB/Simulink model and the simulated results show the high
output voltage ripple reduction from open loop controller to closed loop controller
using various optimization techniques.
Keywords: KY Boost converter, Optimization techniques, Output Voltage Ripple (OVR)
reduction
Cite this Article: Prithivi K, Sathyapriya M and Ashok Kumar L, An Optimized DC-
DC Converter for Electric Vehicle Application, International Journal of Mechanical
Engineering and Technology 8(9), 2017, pp. 173–182.
http://www.iaeme.com/IJMET/issues.asp?JType=IJMET&VType=8&IType=9
1. INTRODUCTION
For the function of the power supply using the low voltage battery such as radio-frequency
(RF) amplifier, audio amplifier, often need very high voltage to obtain enough output power.
This is done by boosting the minimum voltage to the required high voltage. For such
applications, the output voltage ripple must be taken into account purposefully. Regarding the
conventional boost converter, their output currents are pulsating; thereby the corresponding
output voltage ripple tends to be large. As generally approved, to overcome this problem, one
way is the usage of low equivalent series resistance (ESR). In [1], the interleaved control
scheme is employed in the dual buck-boost converters. In the literature [2]-[4], the voltage-lift
technique is utilized to boost the output voltage. But these converters [2]-[4] have one right -
Prithivi K, Sathyapriya M and Ashok Kumar L
http://www.iaeme.com/IJMET/index.asp 174 [email protected]
half plane zero in CCM mode. So it is not easy to achieve required boosted output voltage.
And hence the KY boost converter [5] is presented to overcome this problem. In the literature
[6], FUZZY logic controller is used to reduce the output voltage ripple of KY boost converter.
The detailed illustration of the proposed converter is described herein, along with some
experimental results using various optimization techniques to justify the effectiveness of such
a converter and the effectiveness of the optimization techniques also justified.
2. KY BOOST CONVERTER
Fig. 1 shows the proposed positive-voltage KY boost converter composed of the KY
converter combined with the traditional boost converter, which is described as follows. The
KY converter consists of the switches�� an���, the diod���, the energy-transferring
capacito�, the output inductor�� and the output capacito�. The input of the KY
converter is retrieved by the buffer capacitor . On the other hand, the traditional boost
converter consists of the switches�� and��, the input inductor��. The output of the
conventional boost converter is replaced by the buffer capacito . The buffer capacitor
is a buffer between the KY converter and the traditional boost converter.
Figure 1 KY Boost Converter
Mode 1 operation of KY boost converter
Figure 2 Power flow of Mode 1 operation
In Fig. 2, ��is turned off and�� is turned on. In this case, the negative terminal of � is
pulled to the ground and hence �� is forward biased and turned on. During this mode, is
discharged whereas � is charged. Therefore, the voltage across �� is��, thereby causing �� to
be magnetized, whereas the voltage across �� is �� subtracted from���, thereby causing �� to
be demagnetized. Also, the current flowing through � is equal to ��� minus the current
flowing through��. And hence, the corresponding differential equations are
An Optimized DC-DC Converter for Electric Vehicle Application
http://www.iaeme.com/IJMET/index.asp 175 [email protected]
�� ������ � ��� (1)
�� ����� � ��� � �� (2)
������ � ���� � ��� (3)
� ����� � ��� � �� � (4)
Mode 2 operation of KY boost converter
Figure 3 Power flow of Mode 2 operation
In Fig. 3, �� is turned on and �� is turned off. In this case, �� is in the on-state and hence
�� is reverse biased and turned off. During this mode, is charged whereas � is
discharged. Therefore, the voltage across �� is �� subtracted from��, thereby causing �� to
be demagnetized, whereas the voltage across �� is subtracted from sum of �� and���,
thereby causing �� to be magnetized. Also, the current flowing through � is equal to
���minus the current flowing through���. And hence, the corresponding differential equations
are
�� ������ � �� � �� (5)
�� ������ � 2�� � �� (6)
������ � ��� � ��� (7)
� ����� � ��� � �� � (8)
The average equations which are obtained from equation (1) to equation (8),�� ������ � "��# � "1 � �#"�% # (9)
� ����� � "���# � &�� �' (10)
�� ������ � "2 � �#"�� # � "��# (11)
�"���#�� � "����# ( "1 � �#"���# � �"���# (12)
)*)+ �
,&-.// '"012#3,4&-.// '"012#5016"�47#4"012#3
(13)
Prithivi K, Sathyapriya M and Ashok Kumar L
http://www.iaeme.com/IJMET/index.asp 176 [email protected]
According to the small-ripple approximation and the voltage-second balance, the voltage
ratio can be obtained to be
8�86 �
�47�47 (14)
Where,
d = Duty cycle,
9:= Output voltage in Volts
9; = Input voltage in Volts.
KY Boost Converter Specifications
Table 1 specifications adopted for KY boost converter
Parameter Symbol Value Unit
Input voltage V0 12 V
Rated output voltage V= 36 V
Rated load current i� 10.36 A
Output inductance L= 3.623 mH
Buffer capacitance CA 31.24 mF
Energy transferring
capacitor CB 62.2 mF
Output capacitance C� 470 μF
Switching
frequency fD 195 kHz
Load Separately
excited motor 0.5 hp
Input inductance L0 0.402 mH
3. SIMULINK MODELING OF KY BOOST CONVERTER
The KY boost converter realized in Simulink model is represented by Fig. 4. in which
switches are represented by �� and ��with interrelated body diodes ��and �� . Li represents
input inductor; �� represent output inductor; represent buffer capacitance; � represent
energy transferring capacitor; � is output diode; �� is forward diode. The switching signal is
fed through the connector M and the output voltage taken to Matlab workspace by
connector ��.
3.1 Modeling of fuzzy
Fuzzy controller is formed to diminish the output voltage ripple of the KY boost converter
which is shown in Fig. 5. The input to the Fuzzy controller is error (e) and change in error
(ce), where the error (e) is the deviation of output voltage �� and reference voltage �EFG. The
output of Fuzzy controller is duty cycle (d). To generate the switching signal for KY boost
converter, the duty cycle (d) is given to a PWM generator. Fig. 5 shows the Simulink model
of the proposed Fuzzy controller for KY boost converter which subsists of Fuzzy controller,
PWM generator block with KY boost converter block with reference output voltage of 36 V.
The switching frequency is generated by the PWM generator block is 195 kHz which is fed to
the KY boost converter.
An Optimized DC-DC Converter for Electric Vehicle Application
http://www.iaeme.com/IJMET/index.asp 177 [email protected]
3.2 Modeling of PSO
Particle Swarm Optimization (PSO) comes from the research on the bird and fish flock
progress behavior which is used to find the global minimum of the proposed objective
function [21]. In this method a swarm of particles describe a candidate solution move in the
search space. Each particle position represents the objective function which is depending on
the variables defined by the optimization problem. It is very simple and easy to implement. It
does not require any specific information about the converter model as well as the system
parameters. And the calculation is very simple but the steps for the calculation are high. It is
used in travelling salesman problem, maximum power point tracking in solar panel,
optimization of hydrocarbon field development planning.
3.3 PSO Controller Simulink Modeling
The aim of the present work is to find out the optimized vector of the switching angles, for
which the ripple components are limited to acceptable low level or even to be eliminated. On
the other side the obtained average value has to be equal or closer to the reference or desired
output voltage. To fulfill these requirements an objective function which has to be minimized
is proposed as follows:
f(d) = e (15)
Where,
e = �EFG - �HIJ (16)
�EFG = 36 (17)
This function contains two terms, the error between the desired average voltage and the
obtained output average voltage and the output voltage ripple which is to be eliminated. For
reducing the output voltage ripple the output voltage should satisfy the following Constrain.
d = 0 to 1 (18)
3.4 Modeling of ACO
Ant Colony Optimization is a technique for solving problems which can be expressed as
finding good paths through graphs, Ant Colony is swarm of ants. Each ant tries to find a route
between its nest and a food source. The first ant search randomly until it finds the food source,
and then it returns to the nest, laying a pheromone trail. Other ants follow one of the paths at
random, also set the position of pheromone trails. Since the ants on the shortest path lay trails
faster, this path gets supplement with more pheromone, making it more appealing to future
ants. The ants become increasing likely to follow the shortened path since it is constantly
reinforced with a larger amount of pheromones. The pheromone trails of the longer paths
vaporize. When one ant finds a good path from the colony to a food source, other ants are
more likely to follow that good path, and positive feedback eventually leads to all the ants
following a single path. It is used in Traveling salesman problem, Vehicle routing (school
buses, deliveries, waste collection), Tracking maximum power in solar panel.
ACO Controller Simulink Modeling
The aim of the work is to find out the switching pulse for the switches for reducing the output
voltage ripple and increasing the input voltage to the required voltage range. To fulfill these
requirements an objective function which has to be minimized is proposed as follows:
f(d) = e (15)
Where,
e = �EFG - �HIJ (16)
Prithivi K, Sathyapriya M and Ashok Kumar L
http://www.iaeme.com/IJMET/index.asp 178 [email protected]
�EFG = 36 (17)
4. SIMULATION RESULTS AND DISCUSSION
The simulated results of the KY boost converter with input voltage 12V and the rated output
voltage of 36V for a separately excited motor load is simulated with Matlab/Simulink R2013a
is shown by the following figures Fig. 8, Fig.9, Fig. 10 and Fig11.
Fig.8 shows the output voltage for 12V input voltage in open loop condition of KY boost
converter. It clearly shows that the output voltage ripple for an output voltage of 32.18V in
open loop condition is about 1.18V.
Fig. 9 shows the output voltage for 12V input voltage in closed loop condition of KY
boost converter using FUZZY controller. It clearly shows that the voltage ripple for an output
voltage of 36V is about 0.54V.
Fig. 10 shows the output voltage for 12V input voltage in closed loop condition of KY
boost converter using PSO based controller. It clearly shows that the voltage ripple for an
output voltage of 36V is about 0.275V.
Fig. 11 shows the output voltage for 12V input voltage in closed loop condition of KY
boost converter using ACO based controller. It clearly shows that the voltage ripple for an
output voltage of 36V is about 0.14V.
Table 2 comparison of existing and proposed technique output
System type
(Open/Closed loop) Output voltage
Output voltage
ripple
Open loop system 32.18V 1.18V
Closed loop using FUZZY 36V 0.54V
Closed loop using PSO 36V 0.275V
Closed loop using ACO 36V 0.14V
Figure 4 KY boost Converter Simulink Model
An Optimized DC-DC Converter for Electric Vehicle Application
http://www.iaeme.com/IJMET/index.asp 179 [email protected]
Figure 5 KY Fuzzy Controller Simulink Model
Figure 6 KY PSO Controller Simulink Model
Figure 7 KY ACO Controller Simulink Model
Prithivi K, Sathyapriya M and Ashok Kumar L
http://www.iaeme.com/IJMET/index.asp 180 [email protected]
Figure 8 Output Voltage Waveform of open loop KY Boost Converter
Figure 9 Output Voltage Waveform of closed loop KY Boost Converter using FUZZY logic
controller
Figure 10 Output Voltage Waveform of closed loop KY Boost Converter using PSO based controller
An Optimized DC-DC Converter for Electric Vehicle Application
http://www.iaeme.com/IJMET/index.asp 181 [email protected]
Figure 11 Output Voltage Waveform of closed loop KY Boost Converter using ACO based controller
5. CONCLUSION
The problem of the KY Boost converter output voltage ripple is presented in this paper.
Various optimization techniques are used to solve this problem using an objective function.
This objective function is containing a defined number of switching angles pattern depending
on the degree of ripple reduction. The resulting output values of the closed loop ACO
Controller are compared with the closed loop PSO controller, closed loop FUZZY controller
and open loop system. The reduction of output voltage ripple is proved by the simulation
results.
REFERENCES
[1] Giral R, Arango E, Calvente J & Martinez-Salamero L, In Proceedings IEEE IECON
Conference, 1 (2002) 129-134.
[2] Luo F L & Ye H, IEEE Transactions on Power Electronics, 18 (2003) 1113-1121.
[3] Luo F L & Ye H, IEEE Transactions on Power Electronic, 18 (2003) 105-113.
[4] Luo F L, IEE Proceedings Electric Power Applications, 146 (1999) 415-432.
[5] Hwu K I & Yau Y T, IEEE transactions on Power Electronics, 25 (2010) 2699 – 2703.
[6] Karthikumar S, Mahendran N & Dr. Sriraman S, Emerging trends in Electrical
Engineering and Management, 47 (2012) 181 - 185.
[7] Hwu K I & Jiang W Z, IEEE transaction on industrial electronics, 61 (2014) 1758-1768.
[8] Hwu K I & Yau Y T, IEEE transactions on Power Electronics, 24 (2009) 128-137.
[9] Paolo Mattavelli & Giorgio Spiazzi, IEEE transactions on power electronics, 12 (1997)
79-85
[10] Lijun Hang, Sensen Liu, Gang Yan, Bo Qu & Zheng-yu Lu, IEEE transactions on power
electronics, 26 (2011) 1184 – 1191.
[11] Saetieo S & Torrey D A, IEEE transactions on power electronics, 13 (1998) 419 – 426.
[12] Kouzou A, Abu Rub H,Mahmoudi M.O, Boucherit M & Kennel R, IEEE Transactions
on electronics machines and power electronics, 46 (2011) 758 – 763.
Prithivi K, Sathyapriya M and Ashok Kumar L
http://www.iaeme.com/IJMET/index.asp 182 [email protected]
[13] Kouzou A, Mahmoudi M.O, & Boucherit M.S, IEEE transaction on emerging trends in
electrical engineering and energy management, 46 (2010) 1-6.
[14] Veerachary M & Saxena R, IEEE transactions on industrial electronics, 62 (2015) 1491 –
1502.
[15] Ishaque K, Salam Z, Amjad M & Mekhilef S, IEEE transactions on power electronics, 27
(2012) 3627 – 3638.
[16] Viswanathan K, Oruganti R & Srinivasan D, IEEE transactions on industrial electronics,
52 (2005) 1439 – 1448.
[17] Chia-Feng Juang, Chi-Wei Hung & Chia-Hung Hsu, IEEE transactions on Fuzzy systems,
22 (2014) 723 – 735.
[18] Tianjun Liao, Socha K, Montes de Oca M A, Stutzle T & Dorigo M, IEEE transactions on
evolutionary computation, 18 (2014) 503 – 518.
[19] Birattari M, Pellegrini P & Dorigo M, IEEE transactions on evolutionary computation, 11
(2007) 732 – 742.
[20] Jun Zhang, Chung H S, Lo A W L & Tao Huang, IEEE transactions on power electronics,
24 (2009) 147 – 162.
[21] S Sumathi, Ashok Kumar L, Sureka P Computational Intelligence Paradigms For
Optimization Problems Using Matlab®/Simulink®,–CRC Press, Taylor & Francis Group,
USA, ISBN 9781498743709, 2015 Edition.
[22] J S V Siva Kumar and P Venkata sateesh, design and simulation of dc-dc converter with
high voltage gain for fuel cell based vehicles, International Journal of Mechanical
Engineering and Technology 8(8), 2017,pp. 565–574.
[23] Fernando Santos, Humberto Jorge and Sérgio Cruz, Design of A Multifunctional Flyback
Dc–Dc Converter with Current Control. International Journal of Electrical Engineering &
Technology, 7(3), 2016, pp. 57–72.