See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/261524539

Fuzzy logic controlled buck-boost DC-DC converter for solar energy-battery

system

Conference Paper · June 2011

DOI: 10.1109/INISTA.2011.5946099

CITATIONS READS

22 8,421

2 authors:

Mustafa Ergin Şahin Halil Ibrahim Okumuş

Recep Tayyip Erdoğan University Karadeniz Technical University
91 PUBLICATIONS 1,280 CITATIONS 82 PUBLICATIONS 774 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Halil Ibrahim Okumuş on 26 June 2014.

The user has requested enhancement of the downloaded file.

 Fuzzy Logic Controlled Buck-Boost DC-DC
Converter for Solar Energy-Battery System
Mustafa E. Sahin Halil İ. Okumus
Dept. of Physics, University of Rize Dept. of EEE Karadeniz Technical University
53100, Rize, TURKEY 61080, Trabzon, TURKEY
[email protected] [email protected]

Abstract- In this paper, a Fuzzy Logic Controlled (FLC) II. BUCK-BOOST CONVERTER DESIGN
buck-boost DC-DC converter for solar energy-battery
systems is presented. General design of a fuzzy logic
Figure 2 shows the buck-boost converter system
controller (FLC), based on Matlab/Simulink is with the controller. The actual output voltage V0 is
performed. This design compared with Proportional compared to the reference voltage Vref to produce an
Integrated (PI) controller. The complete control system error signal that is used to determine the switching
has been developed, analyzed, and validated by signal duty cycle. The switching signal is applied on
simulation study. Performances have then been the MOSFET used to reduce and enlarge output
evaluated in detail for different study conditions. voltage on the circuit.

Keywords-Fuzzy Logic Control; PI Control; Buck-

The proposed buck-boost converter is designed for
Boost Converter a battery power system that inconstant input voltage
to constant output voltage for different currents. For
I. INTRODUCTION 50Watt power, 20 kHz switching frequency, and 10%
regulation the system is designed.
Fuzzy logic controller (FLC) is one of the most
widely used applications of fuzzy set theory. It can be
used instead of digital control systems using fuzzy
sets. We can compute with words rather than
numbers. Fuzzy sets are described by membership
functions which are the main toll for the fuzzy
operations. The implementation of linguistic fuzzy
rules by human operators is desired for a complex
Figure 2. Fuzzy logic control with buck-boost converter system
and nonlinear systems without the requirements of
mathematical models parameter estimation. The FLC The buck converter system is analyzed for switch
has faster transient responses and is more robust than on time and off time which is called duty cycle ratio
several control method and which are presented [1,2]. D. The circuit equations are written as bellow.
In this paper, firstly a general FLC algorithm 1. The input power Pin is
oriented on MATLAB/Simulink is presented. For the
model system, five membership functions and a rule Pin = Vin .I in = 50Watt (1)
table are described. Mathematical model of the buck-
boost converter is also given [3-4]. 2. Assuming the efficiency n to be %90
Po
The main purpose of this study is presented in n= ⇒ Po = 45Watt (2)
section 5. In this section the proposed converter Pin
system is simulated by using Matlab/Simulink simple 3. The output current is
operational blocks. This simulation can generate two Po 45
different solutions for the control of converter Io = = = 3.75 A (3)
system; one is PI controller and the other one is FLC Vo 12
controller [5-6]. 4. The output resistance is
At the last section, the simulation results for FLC Vout 12
R= = = 3.2Ω (4)
are compared with the simulation results for PI and I out 3.75
Fuzzy + PI controller using battery loads [7-8]. The 5. The inductor minimum value is (for D=0.5)
proposed converter system is basically designed for a
solar energy- battery system as shown in Figure 1. R(1 − D)2 3.2 x0.25
L= = = 20μ H (5)
2f 2 x 20 x103
6. The capacitor minimum value is
ΔV0 DTs 1.2 0.5x10−3
= ⇒ = ⇒ C = 78μF (6)
Figure 1. Solar energy-battery power system
Vo RC. 12 20x3.2xC
978-1-61284-922-5/11/$26.00 ©2011 IEEE
394
 III. FUZZY LOGIC CONTROL ALGORITHM IV. DESIGN OF FUZZY LOGIC CONTOLLER
The system variables and a rule table which A fuzzy logic based controller consists of three
depend on the variables are described for the control sections namely fuzzifier, rule base and defuzzifier as
algorithm. The buck converter output voltage is shown in Figure 4.
controlled by changing the switching duty cycle. The
system error is defined as a difference between the
reference voltage and measured output voltage value
[2]. For the system; r(k) is the reference voltage and
y(k) is the measured output voltage values then the
error voltage is calculated using Equation (7).
e( k ) = r ( k ) − y ( k ) (7)
The change in the error voltage is also calculated as;
de( k ) = e( k ) − e( k − 1) (8)
Figure 4. The basic structure of fuzzy logic controller with the
The membership functions for each of the fuzzy feedback modeling of system
variables are shown in Figure 3(a) and (b).
Two input signals, the error and change in error
signals for each sampling, to the FLC are converted
to generate fuzzy numbers firstly in fuzzifier. Then
they are used in a rule table to determine the fuzzy
number of the compensated output signal. Finally, the
resultant united fuzzy subsets representing the
controller output are converted into the crisp values.
(a) This process represents the products in the
nominator of defuzzification method called centre of
area (coa). The general expression of the center of
area is given by Equation (9)

∑ μ ( x(i )) * x(i)
coa = i
(9)
(b) ∑ μ ( x(i ))
i
Figure 3. Membership functions
The crisp values of the fuzzy subsets used in this
(a) Membership function for error
multiplication process are the values that have
(b) Membership function for change in error maximum membership degree of 1.0 in the
corresponding fuzzy subsets. Then the sum of this
If the output space, du, is partitioned into more product is divided by the sum of membership values
than three regions as it has been done for e and de obtained as in Figure 5.
axes, then the positive and negative regions should be
divided into sub-regions such as NB, NS, PS and PB.
At the same time some of the fuzzy rules can be
presented as follows:
1- If e is PB and de is NB then du is ZZ.
2- If e is PS and de is NB then du is NS.
3- If e is ZZ and de is NB then du is NS.
4- If e is NS and de is NB then du is NB.
5- If e is NB and de is NB then du is NB.
These rules can be presented as in Table 1 for
three fuzzy regions. This rule table can be used in the
FLC rule-based controller with Simulink design.

TABLE I. FUZZY RULE DECISION TABLE

Figure 5. Fuzzy reasoning representing the process from

fuzzification to defuzzifacation

395
 V. SIMULATION OF BUCK BOOST VI. SIMULATION RESULTS
CONVERTER WITH MATLAB/SIMULINK The Buck-Boost converter simulation results for
The main steps for developing the FLC are the buck mode references 5V and boost mode
outlined in this part. The main block of converter reference 15V are given in Figure 8. There is no big
with Matlab/Simulink is shown in Figure 6. There is overshoot in the output voltage and no need for any
only a feedback and FLC added for control. The change in parameters to stable the system.
output voltage is compared with the reference voltage 16

and generated and error signal. This error signal is 14

driven by FLC. The output signal of FLC is compared 12

with a saw toot signal to generate a PWM signal 10
which drives the switching MOSFET. 8

Volt
Mosfet 6
Diode
g m
4
D S
Discrete ,
s = 0.0001 2
powergui
12V L C R 0
+
- v
VM -2
0 1 2 3 4 5 6 7 8 9 10
Time(Second)
FLC
Gain Figure 8. Simulation results with FLC.
Relational In e
Operator -K- Out2
In de Scope 1
< Step 1 The simulation FLC results are compared with the
Unit Delay
1 simulation PI results in Figure 9. The PI simulations
Scope 2
Repeating
z response time is less than FLC response time. But the
Scope Sequence oscillation and overshoot with PI controller is not
batter then with FLC.
Figure 6. FLC added Buck-Boost converter simulation
20
The simulation parameters which used in these
simulations are shown in Table 2.
15
TABLE II. SIMULATION PARAMETERS
10
DC Input Voltage 12 Volt
V o lt

Reference Output Voltage 5V(Buck ) 15V(Boost)

Switching Frequency 1kHz 5
Filter Inductance 10mH
Filter Capacitance 30mF 0
Output Resistance 5 Ohm
PID Controller Parameters Kp=1, Kı=0.5, Kd=0
-5
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
The FLC controller converter simulated to Time(second)
compare with PI controller. As a new idea for best
simulation results FLC added PI cascade connection. Figure 9. Simulation results with PI controller
The FLC+PI added buck-boost converter circuit is
designed with Matlab/Simulink as shown in Figure 7. The simulation result with FLC+PI controller is
shown in Figure 10. The results of FLC+PI controller
Mosfet
Diode
are shown better than PI and FLC controllers.
g m
D
20
S
Discrete ,
s = 0.0001
powergui 15
12V L C R +v
-
VM 10
V o lt

FLC PID Controller

5
Relational
Gain In e
Operator PID
Scope 1
<
-K- Out2
In de Step 1
0
Unit Delay
1
-5
Scope 2
Repeating
z 0 1 2 3 4 5 6 7 8 9 10
Scope Sequence Time(second)
Figure 10. Simulation results with PI+FLC controller
Figure 7. FLC added Buck-Boost converter simulation

396
 The simulation results with PI+FLC different Matlab/Simulink GUI Environment”, International Journal
situations are shown in Figure 11 and 12. Figure 11 of Information Technology and Intelligent Computing ,
shows first buck mode after than boost mode of 2007, no.4 vol.1.
converter. Figure 12 shows different references and [3] M. E. Sahin, “Designing An Electrolyses System With
output signal of boost mode of converter. Dc/Dc Buck Converter”, M.Sc. Thesis, Gazi University
Institute of Science and Technology, April 2006.
20 [4] Şahin M. E. , Nevruzov V. , “Computer Controlled Solar
House for Measurement Rize City Solar Energy Potential”,
15 NURER, Ankara, 2010.
[5] A. Bakhti and L. Benbaouche, “Simulink-DSP Co- Design of
10 Fuzzy Logic Controller”, IEEE Explore, Jan. 2006,
Volt

pp: 4587-4592.
5 [6] N. Mohan, T. M. Undulant W. P. Robbins, “Power
Electronics”, John Wiley, 1995.
0 [7] B.C.Kuo, “Automatic Control Systems”, New Jersey:
Prentice-Hall, 1995.
-5 [8] Matlab/Simulink, “Modeling A Buck Converter”, Help Files
0 2 4 6 8 10
Mathwork, 2008.
Time(second)
Figure 11. Simulation results with PI+FLC Buck-Boost mode

30

20
Volt

10

-10
0 2 4 6 8 10
Time(second)
Figure 12. Simulation results with PI+FLC for different references

VI. CONCLISIONS
In this paper, a Fuzzy Logic Controlled (FLC)
buck-boost DC-DC converter for solar energy-battery
systems has been presented. The buck-boost
converter circuit has been designed with
Matlab/Simulink and simulation results have been
obtained. The simulation results have been found for
Matlab/Simulink approach. For the controller FLC,
PI, and FLC+PI controllers are used and the
simulation results are compared. It can be seen from
the results that there is no big overshoot in output
voltage and no need for any change in parameters to
stable the system. It can be noted that it is possible to
use a fuzzy logic controller for power electronics
applications and also practical to simulate the
designed circuits in Matlab/Simulink.
For a further study it can be used PV model and
battery model with charge circuit for simulations and
this control algorithms can be tried experimentally.
REFERENCES
[1] M. E. Sahin, H. I Okumus, “Fuzzy Logic Controlled
Synchronous Buck DC-DC Converter for Solar Energy-
Hydrogen Systems”, INISTA 2009 Conference, 2009.
[2] I. H. Altas and A. M. Sharaf, “A Generalized Direct
Approach for Designing Fuzzy Logic Controllers in

397

View publication stats