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

net/publication/329029015

Study of Fuzzy Logic controller based MPPT and the P&O for the Z-source
inverter integrated in PV system

Conference Paper · November 2018

DOI: 10.1109/CISTEM.2018.8613434

CITATIONS READS

2 193

3 authors:

Ibtissam Chaib El Madjid Berkouk

National Polytechnic School of Algiers National Polytechnic School of Algiers
3 PUBLICATIONS   10 CITATIONS    273 PUBLICATIONS   1,942 CITATIONS   

SEE PROFILE SEE PROFILE

Jean-Paul Gaubert
Université de Poitiers
131 PUBLICATIONS   2,250 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Hybrid renewable energy conversion systems connected to the utility grid and supervisory controller system View project

thesis in power electronics View project

All content following this page was uploaded by Ibtissam Chaib on 06 February 2019.

The user has requested enhancement of the downloaded file.

3rd CISTEM’18 - Algiers, Algeria, October 29-31, 2018

Study of Fuzzy Logic controller based MPPT

and the P&O for the Z-source inverter
integrated in PV system
Chaib ibtissam Berkouk El Madjid Gaubert Jaun Paul
Department of Automatic Department of Automatic Department of Automatic
Ecole National polytechnique Ecole National polytechnique university of poitier
Algiers/Algeria Algiers/Algeria poitier/france
[email protected] [email protected] [email protected]

Abstract—The photovoltaic (PV) generation system becomes software are illustrated in Section IV. The experimental results
important as renewable energy sources, due to its advantages such are presented in Section V. Conclusions are presented in Section
as absence of fuel cost, low maintenance requirement, and VI.
environmental friendliness, the integration of the z-source inverter
in the (PV) generation system has been very popular among II. PHOTOVOLTAIC SYSTEM
researchers, due to its ability to boost the PV voltage with its single-
stage topology. This paper focus on the integration of The Z-source
The photovoltaic (PV) system studied in this work, consists
inverter in a photovoltaic (PV) system generation, and the of a PV module, resistive single phase load R, Z-source inverter
implementation of the Perturb & Observe (P&O) and the fuzzy and MPPT control, as shown in Fig. 1. The MPPT control
logic controller algorithm (FLC), to find the Maximum power algorithm makes a decision about how to move the operating
point (MPPT). At the final we compare the results of the FLC with voltage of the PV module, by adjusting the duty cycle of the Z-
the Perturb & Observe (P&O) algorithm. A Matlab simulink source inverter.
toolbox and M-file have been used for the simulation, in order to
verify the results experimentally. The two algorithm was
Z-source R
implemented in a real time with a dSPACE MicroLabBOX, and
converter Load
the I–V curves was programed with Chroma 62150H-600S/1000S
Solar Array Simulator.
PV Duty cycle D
module Vpv
MPPT
I. INTRODUCTION Ipv Algorithm
The PV system has an unique maximum power point (MPP), Fig. 1 The PV system.
which varies with the variation of climate conditions
represented by the temperature and solar irradiation. This A. EQUIVALENT CIRCUIT OF PV ARRAY
(MPP) can be obtained by several algorithm called MPPT Fig. 2 shows the circuits model for PV cell, with series and
algorithm, this algorithms can regulate the duty cycle of the DC- parallel resistances, Rs and Rsh .
DC as a control parameter [1]. The Perturb and Observe is one
of the conventional techniques, this technique is simple and
easy to implement, but it suffer from oscillations at MPP and its
efficiency is low [2]. Several research are carried out to improve
the performance of these conventional techniques, such as
intelligent control methods, like Fuzzy and Neural. These
methods differ in effectiveness, speed of tracking, sensor
required, complexity, and their cost [1,2]. In this study, we will
Fig. 2 circuits model for PV cell.
implement the fuzzy logic controller based MPPT to regulate
the duty cycle of the z-source inverter, and compare their results The current of the PV model is presented by the following
with the results of the P&O MPPT. The paper is organized as equation derived from Kirchhoff law [1].
follows : Section II describes the PV system a brief description ‫ܫ‬௣௩ ൌ ‫ܫ‬௣௛ െ ‫ܫ‬ௗ െ ‫ܫ‬௦௛
about the PV module using an equivalent circuit and explains ܸ ൅ ‫ܴܫ‬௦ ܸ ൅ ‫ܴܫ‬௦
the Z-source inverter (ZSI). Section III reviews the MPPT ‫ܫ‬௣௩ ൌ ‫ܫ‬௣௛ െ ‫ܫ‬௦ ൤݁‫ ݌ݔ‬൬ ൰ െ ͳ൨ െ ൤ ൨
algorithms and describes the P&O MPPT and the FLC based ்ܸ݊ ܴ௦௛
MPPT. Simulation and comparison results using Matlab where: ‫ܫ‬ௗ : diode current,‫ܫ‬௦௛ : shunt current.

978-1-5386-4988-6/18/$31.00 ©2018 IEEE

3rd CISTEM’18 - Algiers, Algeria, October 29-31, 2018

The photocurrent mainly depends on the solar insolation and Simulation of PV array was done by doing step variations of G.
the cell temperature, which is described as [3]: The I-V and P-V characteristics through simulation are shown
‫ܩ‬ in Fig.3 and Fig.4.
‫ܫ‬௣௛ ൌ ൣ‫ܫ‬௦௖௥ ൅ ‫ܭ‬ூ ൫ܶ௖ െ ܶ௥௘௙ ൯൧
‫ܩ‬௥௘௙
where: ‫ܫ‬௦௖௥ : solar cell short-circuit current; ‫ܩ‬௥௘௙ : reference B. The Z-source inverter
solar insolation in ܹ Τ݉ଶ ; ܶ௥௘௙ : cell reference As described in [5] The Z-source inverter employs a unique
temperature; ‫ܭ‬ூ : cell short-circuit current temperature impedance network to couple the inverter main circuit to the dc
power supply. This two-port impedance network consist of a
coefficient; ‫ ܩ‬: solar insolation in ܹ Τ݉ଶ. On the other hand,
split-inductor L1 and L2 and capacitors C1 and C2 connected
the cell saturation current varies with the cell temperature,
in X shape. The ZSI has six active states and two zero states
which is described as:
ଷ which are the same as conventional inverter and an addition
ܶ௖ ‫ܧݍ‬௚ ͳ ͳ shoot-through state (it is forbidden in conventional inverters)
‫ܫ‬௦ ൌ ‫ܫ‬ோௌ ቆ ቇ ݁‫ ݌ݔ‬ቈ ቆ െ ቇ቉
ܶ௥௘௙ ݊݇ ܶ௥௘௙ ܶ௖ which is utilized advantageously to boost the dc-bus voltage [8].
where : ‫ܫ‬ோௌ : cell reverse saturation current at standard condition To generate the shoot-through state, we use the modified PWM
(T=25°C, G=1000ܹ Τ݉ଶ ); ‫ܧ‬௚ : band-gap energy of the Si solar (as the simple boost control, the maximum boost control...)
cell, where ‫ܧ‬௚ = 1.10 eV, and n : is the series cell number and [5][8], where the shoot-through state is located only into the
it dependents on PV technology [3]. zero state , without affecting the active state.
The reverse saturation current at reference temperature can be The z-source inverter and the two basic operation mode of
approximately obtained as: ZSI are illustrated in Fig. 5.
‫ܫ‬௦௖
‫ܫ‬ோௌ ൌ
‫ܸݍ‬
݁‫ ݌ݔ‬ቀ ௢௖ ቁ െ ͳ
݊݇ܶ௖
In this work, the PV module Conergy PowerPlus 214P solar
was used. The module delivers a maximum power output equal
to 214W at standard temperature condition (STC). At STC, ܶ ൌ
ܶ௥௘௙ ൌ ʹͷι‫ ܥ‬and solar irradiation ‫ܩ‬௥௘௙ =1000 ܹ Τ݉ଶ . The
electrical parameters of this module are listed in Table I [4]. (a)
TABLE I
ELECTRICAL PARAMETERS OF Conergy PowerPlus 214P
Parameter value
Maximum power (ܲ݉ܽ‫ ݔ‬ሻ 214 W
Voltage at ܲ݉ܽ‫ ݌ܸ݉( ݔ‬ሻ 28.55 V
Current at ܲ݉ܽ‫ ݌݉ܫ( ݔ‬ሻ 7.63 A
(b)
Short-circuit voltage (ܸ‫) ܿ݋‬ 35.54 V
Short-circuit current (‫) ܿ݋ܫ‬ 8.11 A
Temperature Co-efficient of ܸ‫ܿ݋‬ -0.117 V/°C
Temperature Co-efficient of ‫ܿݏܫ‬ 4.73 mA/°C

(c)
Fig 5. a) The z-source inverter
b) Equivalent circuits of ZSI Shoot-through mode
c) Equivalent circuits of ZSI Non-shoot through mode

As verified in detail in the basic principal, relationship for ZSI

is:
ͳെ ଴
ୡభ ൌ ୡమ ൌ ୡ ൌ ǡ ൌ
ͳ െ ʹ ୥
Where
Fig. 3 I-V characteristics of PV module ͳ
ൌ ൌ ୥ ‘•Š‘‘– െ –Š”‘—‰Š•–ƒ–‡
300
൝ ୢୡ ͳ െ ʹ ୥
ୢୡ ൌ Ͳ•Š‘‘– െ –Š”‘—‰Š•–ƒ–‡
250 1 kW/m 2
Power (W)

200 0.8 kW/m2 Where ܸ௖భ and ܸ௖మ are capacitors voltage of impedance network
150 0.6 kW/m2 which are the same due to circuit symmetry. B is the boost
100 0.4 kW/m2 factor of ZSI. ܸ௚ and ܸௗ௖ denote the input and output voltages
0.2 kW/m2 of impedance network respectively [8,10].
50

0
0 5 10 15 20 25 30 35 40
Voltage (V)
Fig. 4 P-V characteristics of PV module
3rd CISTEM’18 - Algiers, Algeria, October 29-31, 2018

III. MPPT ALGORITHMS E(k), οE(k) and D(k) at a sampling instant k are related by the
As shown in Fig.3 the MPP for the PV module changes following equations:
dynamically with the variation of climate conditions (irradiation
and temperature). Therefore the implementation of the MPPT ‫ܧ‬ሺ݇ሻ ൌ ሾܲሺ݇ሻ െ ܲሺ݇ െ ͳሻሿȀሾܸሺ݇ሻ െ ܸሺ݇ െ ͳሻሿ
algorithm is necessary, to locate the MPP. ο‫ܧ‬ሺ݇ሻ ൌ ‫ܧ‬ሺ݇ሻ െ ‫ܧ‬ሺ݇ െ ͳሻ
The different MPPT methods that have been analyzed for this ‫ܦ‬ሺ݇ሻ ൌ ‫ܦ‬ሺ݇ െ ͳሻ ൅ ο‫ܦ‬ሺ݇ሻ
purpose are Perturb and Observe (P&O), and Fuzzy Logic
Controller (FLC) methods. Where P(k) and V(k) are the output power and voltage of
the PV module. ο‫ܦ‬ሺ݇ሻ is the change in the duty ratio. E(k)
represents the slope of the P-V curve [1,7].
A. Perturb and Observe (P&O) Therefore, the sign of E(k) expresses the location of operating
The (P&O) method is a simple method and can be point at instant k either on the left or on the right of MPP on the
implemented easily. The operating principal of the P&O is the PV module’s PV curve, while the input οሺሻ expresses this
perturbation of the duty cycle, by incrementing or decrementing point’s moving direction. The FLC consists of three basic
the step size ο‫ܦ‬, and observes the resulting change of PV power components: fuzzification, fuzzy inference engine, and
(οܲ). If οܲ >0, the operating point has shifted closer to MPP, defuzzification [1,7]. The membership functions (MF) of input
thus the duty cycle is more perturbed in the same direction, in variables are shown in Fig.8, Fig.9. The fuzzy control rules as
the other case, the direction will be reversed. This process is given in Table II was done with Mamdani Methods and the
stopped when the MPP is located. The algorithm oscillates the defuzzification was done with the centre of gravity method to
operating point around MPP and slightly reduces PV efficiency. calculate the out of FLC (duty cycle) [1,9], by the following
equation:
start ௡ ௡

ο‫ ܦ‬ൌ ෍ሾο‫ܦ‬௜ ൈ ߤሺο‫ܦ‬௜ ሻሿ൙෍ ߤሺο‫ܦ‬௜ ሻ

௜ୀଵ ௜ୀଵ
Measure V(t), and I(t)

Calculate the power P(t)

No Yes
P(k)>P(k-1)

Fig 8. MF of input variables E.

No Yes No Yes
V(k)<V(k-1) V(k)>V(k-1)

D(k)=D(k-1)+ο‫ ܦ‬D(k)=D(k-1)+ο‫ܦ‬ D(k)=D(k-1)-ο‫ ܦ‬D(k)=D(k-1)+ ο‫ܦ‬

Duty cycle D Fig 9. MF of input variables ο‫ܧ‬.

Fig. 6 Flow chart of P&O algorithm.

TABLE II
FUZZY RULE BASE TABLE
The P&O algorithm can be easily understood by studying ο‫ܧ‬ BN NS Z PS PB
the flow chart shown in Fig. 6. [1,6] E
NB Z Z PB PB PB
B. Fuzzy Logic Controller (FLC) Method
NS Z Z PS PS PS
A functional block diagram of the FLC for located the MPP Z PS Z Z Z NS
is shown in Fig. 7, in which error E(k) and οE(k) are the inputs PS NS NS NS Z Z
to the fuzzy processor (FP), while the control output is D(k). PB NB NB NB Z Z

Rule base
IV. SIMULATION RESULTS
To evaluate the performance of the proposed system,
simulation models have been established using MATLAB
E(k) Inference οD(k)
Fuzzifier Defuzzifier SIMULINK software, where we used the Conergy PowerPlus
οE(k) engine
214P as solar PV module. The electrical parameters of this
module are listed in Table I. A resistive load R and a Z-source
Fig. 7 Structure of fuzzy logic controller.
inverter are used to evaluate the proposed algorithms . The main
3rd CISTEM’18 - Algiers, Algeria, October 29-31, 2018

parameters of the Z-source inverter and MPPT algorithm are

listed in Table III. The two MPPT algorithms explained
previously are simulated and compared in terms of their
behaviors in tracking capability with the variation of the
irradiation.

Table III
SIMULATION PARAMETERS
parameters R ୸ ୸ ˆୡ ο‫ܦ‬୫ୟ୶ ο‫ܦ‬୫୧୬

value 20 5 mH 3.3 mF 10 kHz 0.01 0.005

Fig.13 Transient response PV’s power for G=500 W/m2.

Fig.14 represent the variation of the duty cycle in the steady-
stat for G=500 W/m2 , Fig. 11 represent the response of the
PV’s power using two tracking algorithms at the different
atmospheric conditions shown in Fig.10 . From the results
shown in Fig. 11 and Fig. 14 we can see that the FLC based
MPPT applied in the z-source inverter at the steady state has a
better stability’s performance and less oscillation of the
operating point around MPP than the P&O. The inconvenient
of the FLC based MPPT applied in the z-source inverter is that
it has a bad transient response, and track the MPP slower than
the P&O as shown in Fig.12. This inconvenient due to the Fig.14 Steady-state response PV’s power for G=500 W/m2.
complexity of this algorithm. Fig.15 show the variation of the
duty cycle for each algorithm, as we can see the duty cycle of
the FLC oscillate more than the P&O’s duty cycle and has a
rapid variation, which decrease the oscillation of the PV’s
power in the steady-state.

Fig.15 Duty cycle variation for G=500 W/m2.

V. EXPERIMENTAL VERIFICATIONS
The experimental solar generation system is the same as the
Fig.10 Irradiance variation. simulation status. The parameters of the z-source inverter and
the load, can be obtained as those listing in Table III. A
simulator solar array “Chroma 62150H-600S/1000S” is used to
generate the required P-V and I-V characteristics. The electrical
parameters of the PV module Conergy PowerPlus 214P used in
this study are listed in Table I. The two MPPT algorithm was
implemented using a dSpace MicroLabBOX, where the
converter switching frequency is 10 kHz.

A. Mppt P&O
Fig.11 Power variation.

Fig.16 I–V and P–V curves of the PV’s array

for G=600 W/m2.

Fig.12 Duty cycle variation.

3rd CISTEM’18 - Algiers, Algeria, October 29-31, 2018

B. MPPT FLC

(a)

Fig.20 I–V and P–V curves of the PV’s array

for G=600 W/m2.

(b)

(a)

(c)
Fig.17 The measurement of Ipv (a) , Vpv(b), Ppv(c)
for G=600 W/m2.
(b)

(c)
Fig.18 I–V and P–V curves of the PV’s array Fig.21 The measurement of Ipv (a) , Vpv(b), Ppv(c)
for G=200 W/m2. for G=600 W/m2.

(a)
Fig.22 I–V and P–V curves of the PV’s array
for G=200 W/m2.

(b)
(a)

(c)
Fig.19 The measurement of Ipv (a) , Vpv(b), Ppv(c)
for G=200 W/m2.
(b)
3rd CISTEM’18 - Algiers, Algeria, October 29-31, 2018

REFERENCES
[1] Ammar Ghalib, Al-Gizi, Sarab Jwaid Al-Chlaihawi, "Study of FLC
Based MPPT in Comparison with P&O and InC for PV Systems, "
2016 International symposium on Fundamentals of Electrical
Engineering University Politehnica of Bucharest, Romania, June 30-
July2.2016.
[2] Mostefa Kermadi, El Madjid Berkouk, "Artificial intelligence-based
(c) maximum power point tracking controllers for Photovoltaic
Fig.23 The measurement of Ipv (a) , Vpv(b), Ppv(c) systems: Comparative study" Renewable and Sustainable Energy
for G=200 W/m2.
Reviews 69 (2017) 369–386.
[3] Wafaa ABD EL-BASIT, Ashraf Mosleh ABD El–MAKSOOD,
Fouad Abd El-Moniem Saad SOLIMAN " Mathematical Model for
Fig. 16 and Fig.18 shows the I–V and P–V curves of the PV’s Photovoltaic Cells, " Leonardo Journal of Sciences ISSN 1583-
array, for G=600 W/m2 and G=200 W/m2 respectively, where 0233, December 2013 p. 13-28
we can see the MPP tracking of the P&O. [4] Conergy PowerPlus 214P–240P, PowerPlus 214P_240P, Juin_2011,
www.conergy.fr.
Fig. 17 and Fig.19 shows The measurement of Ipv (a) , Vpv(b),
[5] F. Z. peng, "Z-source inverter," IEEE Transactions on Industry
Ppv(c), for G=600 W/m2 and G=200 W/m2 respectively, for the Applications , vol. 39, no. 2, pp. 504-510, 2003.
P&O. [6] Radak Blange, Chitralekha Mahanta, Anup Kumar Gogoi, "MPPT
Fig. 20 and Fig.22 shows the I–V and P–V curves of the PV’s of Solar Photovoltaic Cell Using Perturb & Observe and Fuzzy
array, for G=600 W/m2 and G=200 W/m2 respectively, where Logic controller Algorithm for Buck-Boost DC-DC
we can see the MPP tracking of the FLC based MPPT. Converter", Energy, Power and Environment: Towards Sustainable
Growth (ICEPE), 2015 International Conference on, 12-13 June
Fig. 21 and Fig.23 shows The measurement of Ipv (a) , Vpv(b),
2015.
Ppv(c), for G=600 W/m2 and G=200 W/m2 respectively, for the [7] T.-F.-Wu, C.-H. Chang, Y.-K. Chen "A Fuzzy Logic Controlled
FLC based MPPT. From the Fig 16, 18, 20 and 22, the two Single-Stage Converter for PV Powered Lighting System
algorithms track the MPP, at the different atmospheric Applications" IEEE Transactions on Industrial
condition. From the Fig 17, 19, 21 and 23, we can see that the Electronics ( Volume: 47, Issue: 2, Apr 2000, 287 – 296
simulation results have been verified experimentally, as we can [8] Chaib Ibtissam, Behlouli Asma, Hdjaidji Fatma Zahra, Benkouider
see the FLC based MPPT has a better performance in the steady ouarda, Berkouk El-Madjid, "Comparative study between diffirent
control strategy of the Z-Source inverter" , Electrical Engineering -
state performance and less oscillation of the operating point Boumerdes (ICEE-B), 2017 5th International Conference on, 29-31
around MPP than the P&O. Oct. 2017.
[9] A. A. S. Mohamed, Alberto Berzoy, Osama Mohammed, "Design
and Hardware Implementation of FLMPPT Control of PV Systems
VI. CONCLUSION Based on GA and Small-Signal Analysis", IEEETransactions on
The main purpose of this paper is to track the MPP of the Sustainable Energy, September 4, 2015.
[10] H. Rostami, D. A. Khaburi, "Voltage Gain Comparison of Different
PV module, by the implementation of two algorithms P&O and Control Methods of the Z-Source Inverter ",Electrical and
the FLC based MPPT, using the Z-source as an adapter. The Electronics Engineering, 2009. ELECO 2009. International
algorithm has been developed in Matlab Simulink environment, Conference on, Bursa, Turkey
and verified experimentally, where we used the Chroma
62150H-600S/1000S Solar Array Simulator to program the I–V
curves, and the dSPACE MicroLabBOX to implement the
algorithms in real time. The results have been compared. From
the simulation results and experimental verification, it is found
that the variations of the duty cycle by Fuzzy Logic Controller
technique gives better performance and more stable in the
steady-state in comparison with P&O which has a faster
transient response than the FLC.

View publication stats