Modelling of fuzzy logic controller for variablestep

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
Volume: 02 Issue: 08 | Aug-2013, Available @ http://www.ijret.org 426
MODELLING OF FUZZY LOGIC CONTROLLER FOR VARIABLE-
STEP MPPT IN PHOTOVOLTAIC SYSTEM
C. P. Roy1
, D. Vijaybhaskar2
, T.Maity3
1, 2, 3
Department of Electrical Engineering,
Indian School of Mines, Dhanbad, Jharkhand, India
prakashroy2507@gmail.com, devaravijay@gmail.com
Abstract
The output power of photovoltaic electrical systems is highly dynamic and non-linear in nature. In order to extract maximum power
from such systems, maximum power point tracking (MPPT) technique is required. MPPT techniques with variable step-size of
perturbation track the maximum power point (MPP) with more efficiency. In this paper, a model of a fuzzy logic controller (FLC) for
determining the step-size of perturbation in duty-cycle of a photovoltaic electrical system to track MPP is presented. The model is
simulated in MATLAB/Simulink®
.
Keywords: Maximum power point tracking, perturb and observe, boost converter, fuzzy logic control, membership
function, crisp universe, centre of area, pulse width modulation.
----------------------------------------------------------------------***-----------------------------------------------------------------------
1. INTRODUCTION
Photovoltaic electrical systems are those which convert the
energy of photons directly into electrical energy. The output
power of such a system is highly sensitive to the environmental
parameters like- insolation and temperature. Hence, the
maximum power that can be extracted from a panel also
changes with change in these parameters. In order to ensure
maximum extraction of power from a PV panel under varying
environmental conditions load should be changed in accordance
with changing environmental parameters, so that the operating
point or quiescent point always lies at the maximum power
point. Since, practically it is not possible to change the load
time to time; there must be some interfacing circuit in between
the PV panel and the load which can change the load (as seen
from the side of PV terminals, i.e. between terminals E&F)
with change in environmental parameters depending upon some
control variables. As the output of PV panel is of DC in nature,
generally switched mode power converters (DC to DC
converters) are used as the interfacing circuit. Fig.1 shows the
block diagram of photovoltaic electrical system.
Fig.1 Block diagram of photovoltaic electrical system
The operating point is changed by varying the duty-cycle of
these power converters [1]. There are so many techniques in
literature according to which the duty-cycle of these converters
can be changed to track the maximum power point. In this
paper, variable-step perturb and observe algorithm has been
used for this purpose [2]. Variable step-size of perturbation
ensures time efficient tracking and at the same time gives better
stability in output power at maximum power point. In order to
determine smooth and quick variation in step-size, fuzzy logic
control has been used in this paper.
The paper is organised as such: in section 2 photovoltaic
electrical system model is presented; maximum power point
algorithm and modelling of fuzzy logic controller is presented
in section 3. Results and discussions are made in section 4 and
finally conclusions are drawn in section 5.

IJRET: International Journal of Research in Engineering and Technology
__________________________________________________________________________________________
Volume: 02 Issue: 08 | Aug-2013, Available @
2. MODELLING OF PHOTOVOLTAIC
ELECTRICAL SYSTEM
2.1 SOLAR PANEL
For analysis purpose, single diode model of solar panel is
considered in this paper as shown in Fig.2.
Fig.2 single diode model of solar cell.
The same model is built in MATLAB/Simulink
Fig.3. The effect of temperature and irradiance on photo
generated current ILis estimated using (1) and (2)
parameter values are tabulated in Table-1.
)(1)(( rTTiKrTLILI −+=








=
0
)(
G
G
scIrTLI
Where, ILis photo-generated current (A), T is absolute climatic
temperature (K), Ki is short-circuit current temperature
coefficient (A/K), G is insolation (W/m2), GO&TO are
standard insolation and temperature respectively.
Fig.3 Simulink model of solar panel.
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319
__________________________________________________________________________________________
2013, Available @ http://www.ijret.org
OTOVOLTAIC
For analysis purpose, single diode model of solar panel is
single diode model of solar cell.
The same model is built in MATLAB/Simulink® as shown in
. The effect of temperature and irradiance on photo-
is estimated using (1) and (2) [3]. All the
(1)
(2)
generated current (A), T is absolute climatic
circuit current temperature
coefficient (A/K), G is insolation (W/m2), GO&TO are
standard insolation and temperature respectively.
Simulink model of solar panel.
Table-1 Parametric details of solar panel.
Parameters
Vmax
Imax
Pmax
VOC
ISC
Encapsulated solar cell efficiency
Module efficiency
Size of Panel
Cell type
NP
NS
Ki
RS (of a single cell)
RSh(of a single cell)
STC(Standard Test Condition)
2.2 POWER ELECTRONICS
With change in temperature and irradiance, internal resistance
of solar panel also changes. The change in
and P-V curves of a solar panel with varying irradiance and
temperature is shown in Fig
be inferred that with increase in irradiance level, net output
power increases while with increase in ambient temperature,
decreases non-linearly.
0 0.1 0.2
0
0.5
1
1.5
2
2.5
3
3.5
4
Icell(A)/Pcell(W)
I-V
eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
427
Parametric details of solar panel.
Values
17 V
3.5 A
60 W
21 V
3.7 A
Encapsulated solar cell efficiency 15 %
12.1 %
93.75×125 mm2
Mono Crystalline
1
36
1.7×103A/K
0.03 Ω
10000Ω
STC(Standard Test Condition) K,1000 W/m2
ELECTRONICS COMPONENTS
With change in temperature and irradiance, internal resistance
of solar panel also changes. The change in characteristics of I-V
V curves of a solar panel with varying irradiance and
temperature is shown in Fig4. From the curves obtained, it can
be inferred that with increase in irradiance level, net output
power increases while with increase in ambient temperature, it
(a)
0.3 0.4 0.5 0.6 0.7
Vcell (V)
400 W/m
2
600 W/m
2
800 W/m
2
1000 W/m
2
P-V

__________________________________________________________________________________________
(b)
Fig.4 Variation in PV output characteristics with changing (a)
insolation (b) temperature.
To extract maximum output power, the output resistance as
seen from the panel side must be equal to its internal resistance.
Therefore, the interfacing circuit must have an efficient
impedance matching property. Switched mode power regulators
[4] help to incorporate this property in photovoltaic systems.
By varying the duty-cycle of switching element used in the
topology, it is possible to change the resistance as seen from the
input terminals of these regulators as shown in Fig.5[1].
L
C CMosfet
Pulses
D
L
O
A
D
1 2
S P
O A
L N
A E
R L
Rin
Fig.5 Topology of boost converter.
In this paper boost converter [5] is used as the interfacing
circuit as shown in Fig.4. The capacitor C1 has been used to
reduce the fluctuation in voltage at input terminals of the boost
converter. The parametric details of the modelled boost
converter have been provided in Table-2. RL is the equivalent
series resistance (ESR) of inductor; RC1 and RC2 are the ESR
of the capacitors C1 and C2 respectively.
Table-2 Parametric details of boost converter.
Under steady state, the relation between input and output
resistances of a boost converter is governed using (3) [6].
LOADRd
in
R
2
)1( −=
(3)
Where, Rin is the input resistance and RLOAD is the output
load resistance. From (3), it’s quite clear that by varying the
duty-cycle d, the input resistance can easily be changed.
3. MAXIMUM POWER POINT TRACKING
3.1 PERTURB AND OBSERVE ALGORITHM
The switched mode power regulators change the output
resistance of solar panel depending upon the change in its duty-
cycle. But, the manner in which the duty-cycle should be varied
is decided by some artificial intelligence techniques called as
maximum power point tracking (MPPT). In this paper, perturb
and observe (P&O) technique [2] with variable step-size of
perturbation has been used for tracking the maximum power
point (MPP). The complete working of P&O algorithm is
shown as flowchart in Fig.6. The step-size of perturbation is
varied proportional to the distance of quiescent point (i.e.
present operating point) from the MPP using a fuzzy logic
controller. As the quiescent point approaches the MPP, the
step-size is gradually reduced to zero.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
0
0.5
1
1.5
2
2.5
3
3.5
4
Vcell(v)
Icell(A)/Pcell(W)
340K
320K
300K
280K
I-V
P-V
Parameters Values
Rc1 11.23 mΩ
C1 76.8 mF
Rc2 9.44 mΩ
C2 74.8 mF
RL 18.92 mΩ
L 4.2912 mH
Diode threshold voltage 0.7V
Diode ON resistance 190 µΩ
MOSFET ON resistance 20 mΩ
Switching frequency 5 kHz

__________________________________________________________________________________________
Start
Read Vt & It; Pt=Vt*It
Dt+1=Dt+Dm
Pt > Pt-1
Vt > Vt-1
Vt> Vt-1
Dt+1=Dt-Dm
Next Perturbation
Y N
Y N
N
Y
Fig.6 P&O algorithm.
In the flowchart given in Fig.5,t represents the iteration, t-1
represents the previous iteration and t+1 represents the next
future iteration. D is the constant signal which is being
compared with a ramp signal to produce square pulses of
desired duty-cycle and Dm is the step-size of perturbation as
determined by the fuzzy logic controller.
3.2 MODELLING OF FUZZY LOGIC
CONTROLLER
The complete working of Fuzzy Logic Controller is categorised
into three different domains as discussed under [7, 8]:
3.2.1 FUZZIFICATION
Fuzzification is the process of representation of different crisp
variables by the predefined fuzzy subsets. In this paper, slope
of the P-V curve, slope and the perturbation of previous step
i.e. ∆D are chosen as crisp variables and the crisp universe is
partitioned into five different fuzzy subsets giving rise to total
twenty-five subsets in fuzzy output universe. For partition of
crisp universe, triangular membership function has been used as
shown in (4) [9].














−
−
−
−
= 0,
23
3
,
12
1
minmax)(
xx
xx
xx
xx
xµ
(4)
Where, x is the crisp variable and x1, x2&x3 are critical crisp
points corresponding to left end, peak value & right end of the
crisp universe.
The modelling of membership function and the fuzzifier in
MATLAB/Simulink® are shown in Fig.7 (a) and Fig.7 (b)
respectively.
(a)
(b)
Fig.7 Simulink model of (a) membership function (b) crisp
universe partition.
3.2.2 FUZZY RULES
Fuzzy Rules refer to a set of pre-defined instructions which link
the different values of crisp variables with different subsets of
fuzzy output space. The fuzzy based rules used in this paper are
shown in Table-3 and its MATLAB/Simulink® implementation
is presented in Fig.8. These rules are framed based on the logic
that if the quiescent point is far away from MPP, then step-size
of perturbation should be very large and it should be gradually
decreased to zero as the operating point approaches to zero. At
MPP, the slope of P-V curve will be zero; hence the

__________________________________________________________________________________________
perturbation should also become zero so that stability in the
power can be achieved
Table-3 Fuzzy based rules.
Fig.8 Simulink model for fuzzy rules.
3.2.3 DEFUZZIFICATION
Defuzzification is the method of finding final crisp value in
fuzzy output space. The values of membership functions
determine the weightage of different subsets in fuzzy output
universe. To obtain the final crisp output, the centre of area[10]
method on the weightages of these different fuzzy subsets (of
output universe) has been used in this paper. The
MATLAB/Simulink® implementation of defuzzification
process is shown in Fig.9.
Fig.9 Model for defuzzification.
The final overall MATLAB/Simulink® model of MPPT using
FLC for variable step-size estimation is shown in Fig.10. The
critical crisp values for the two inputs X 1st (i.e. ∆D, step-size
of previous perturbation) and X 2nd (i.e. slope) and the output
(i.e. dD) have been fixed as per the choice and requirement
after iterative simulations of the model. The crisp values are
same as provided in Table-3.
X 2nd
X
1
S
t
SLOPE PB PS ZZ NS NB
dD 5.0 0.1 0.0 -0.1 -5.0
PB 0.10 ZZ PS ZZ PS PB
PS 0.01 PB ZZ ZZ PS PB
ZZ 0.00 PB PS ZZ PS PB
PS 0.01 PB PS ZZ ZZ PB
PB 0.10 PB PS ZZ PS ZZ

__________________________________________________________________________________________
(a)
(b)
Fig.10 Complete Simulink model of (a) fuzzy logic controller,
(b) MPPT algorithm using FLC
The panel output current and voltage are fed as input to both
the P&O algorithm (for determining the direction of
perturbation) and FL based controller (for determining the step-
size of perturbation). The ultimate value of D as obtained from
this block is used to generate pulses for switching device of
boost converter. For switching pulse generation, ramp PWM
technique has been used.
4. RESULTS AND DISCUSSIONS
The Simulink model of the PV system is shown in Fig.11, it
was simulated in MATLABfor a resistive load of 30 ohm to
judge the performance of variable-step MPPT for PV system
under changing insolation condition.
Fig.11 Simulink model of overall PV system
The tracking of MPP is presented on power(W) vs time(s)
curve as obtained from simulation in Fig.12. The simulation
was started at t=0; and after reaching the MPP at 900 W/m2,
the irradiance was changed to 1000 W/m2using a manual
switch as shown in Fig.11.The complete model as shown in
Fig.10 is represented by a single block in Fig.11. From the
result shown in Fig.12 it’s quite clear that the system works
efficiently in tracking the MPP under varyingenvironmental
condition. Here, it has been tested for change in irradiance
level.
Fig.12 MPP tracking using FLC based variable-step P&O
algorithm
CONCLUSIONS
In this paper, a fuzzy logic controller is modelled in
MATLAB/SIMULINK® platform for determining the step-size
of perturbation of variable step-size P&O algorithm for
maximum power point tracking of a photovoltaic electrical
0 5 10 15 20 25 30
0
10
20
30
40
50
60
time (s)
power(W)

__________________________________________________________________________________________
system. The result shows that when the irradiance level is
increased from 900 W/m2 to 1000 W/m2, the algorithm directs
the operating point towards MPP to extract maximum power.
REFERENCES
[1] C.H and C.S , ”Control of DC/DC converters for solar
energy system with maximum power tracking” , IECON 97.
23rd , 1997, vol. 2, pp. 827-832.
[2] W. T. Chee, T. C. Green, and A. H.-A. Carlos, “Analysis of
perturb and observe maximum power point tracking algorithm
for photovoltaic applications”, IEEE 2nd Int. Pow. and Energy
Conf., Johor Bahru, Malaysia, 2008.
[3] Mohsen Taherbaneh, Mohammad B.Menhaj, “A fuzzy-
based maximum power point tracker for body mounted solar
panels in LEO satellites”, In Proc: IEEE Ind & Comm Pow.
Syst Tech., May 2007, pp. 1-6.
[4] Solodovnik E.V., Shengyi Liu, Dougal R.A., “ Power
controller design for maximum power tracking in solar
installations”, J Trans Power Electron, 2004, Vol. 19(5), pp.
1295-1304.
[5] H. S. Kim, J. M. Park, K. B. Jo, H. W. L, “A Study on the
Boost Converter for MPPT Using Micro-controller in PV
System ”, KSES CNF , 2003, pp. 51-60.
[6] H. Chung, Y. K. Mok, “Development of a switched-
capacitor dc–dc boost converter with continuous input current
waveform”, IEEE Trans. Circuits Syst. I, Fundam. Theory
Appl., Jun. 1999, vol. 46(6), pp. 756–759.
[7] Won C.-Y., Kim D.-H., Kim S.-C., Kim W.-S., Kim H.-S.,
“A new maximum power point tracker of photovoltaic arrays
using fuzzy controller”. In Proc: 25th Annual IEEE PESC, 20–
25 Jun. 1994, vol. 1, pp. 396–403.
[8] Simoes M.G., Franceschettin N.N.: ‘Fuzzy optimization
based control of a solar array’, In Proc. IEE Electr. Power
Appl., 1999, 146, (5), pp. 552–558.
[9] Sung-Jun Kang, Jae-Sub Ko, Jung-Sik Choi, Mi-Geum
Jang, Ju-Hui Mun, Jin-Gook Lee, Dong-Hwa Chung, “A Novel
MPPT Control of photovoltaic system using FLC algorithm”,
11th Int. Conf. on Cont. Auto. and Syst., Oct. 26-29, 2011, pp.
434-439.
[10] K. Passino and S. Yurkovich, “Fuzzy control”, Reading,
MA, USA: Addison-Wesley, 1998.

Modelling of fuzzy logic controller for variablestep

Recommended

More Related Content

What's hot (18)

Viewers also liked (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Modelling of fuzzy logic controller for variablestep