Modeling and simulation of a gird connected PV generation system With MPPT fuzzy logic control30

2010 7th International Multi-Conference on Systems, Signals and Devices

Modeling and simulation of a gird connected PV generation system With MPPT fuzzy logic control

F.Bouchafaal, D.Beriberl, M.S.Boucherit2

I Laboratory of Instrumentation, Faculty of Electronics and Computer, University of Sciences and Technology

Houari Boumediene, BP 32 E/-Alia 16111 Bab-Ezzouar Algiers, Algeria.

2 Laboratoire de commande des processus. Departement Genie Electrique et Informatique.

Ecole Nationale Poly technique d'Alger-10, rue Hassen Badi, El Harrach, Alger BP 182, ALGERIA

Email: [email protected]@[email protected].

Abstracts

This paper presents a simulation model of the electric part of a grid connected photovoltaic generation system. The model contains a detailed representation of the main components of the system that are the solar array, and the grid side inverter multilevel inverter Neutral Point Clamped (NPC) VSI. In order to extract the maximum amount of from the photovoltaic generator, we propose an intelligent control method for the maximum power point tracking (MPPT) of a photovoltaic system under variable temperature and insulation conditions. This method uses a fuzzy logic controller. As part of our work, we will focus on voltage inverter at three levels to NPC structure. The latter can increase the voltage supplied to the load (network) through their topology. Thus, they can generate more voltage sinusoidal possible and improve the total harmonic distortion through the high voltage levels provided by the structure of this new converter. The grid interface inverter transfers the energy drawn from the PV module into the grid by keeping common dc voltage constant. The PQ control approach has been presented for the multilevel inverter. The simulation results under MatlablSimulink show the control performance and dynamic behaviour of grid connected photovoltaic system.

Keywords: PV, MPPT, fuzzy logic control , NPC inverter, clamping bridge, enslavement, gird.

1. INTRODUCTION

In recent years, the efforts to spread the use of renewable energy resources instead of pollutant fossil fuels and other forms have increased. Photovoltaic systems are increasing in size as they become more affordable and supporting schemes start to include larger installations. In a near future, photovoltaic systems are going to be very common, and it is expected that they will contribute with a significant share to power generation. One of the most common control strategies structures applied to decentralized power generator is based on Power direct control employing a controller for the dc link voltage and a controller to regulate the injected current to the utility network. The system components and power control scheme were modelled in terms of dynamic behaviours. The proposed models were implemented in Matlab/Simulink. This paper describes

978-1-4244-7534-6/10/$26.00 ©2010 IEEE

the dynamic performance of the PV generator connected through inverters to distribution network under standard climatic conditions (1 OOOW 1m

2 and 25°C).

As part of our work, we will focus on voltage inverter at three levels to NPC structure. The latter can increase the voltage supplied to the load through their topology. Thus, they can generate more voltage sinusoidal possible and improve the total harmonic distortion through the high voltage levels provided by the structure of this new converter.

2. SYSTEM CONFIGURATION

Fig. 1 shows the configuration of the grid-connected PV system, which consists of solar cell array and the three phase multilevel inverter NPC VSI. The control structure of the grid-connected PV system is composed of two structure control: 1. The MPPT Control, which main feature is to extract the maximum power from the PV generator.

2010 7th International Multi-Conference on Systems, Signals and Devices

ipv: I I

Fig I. General diagram of grid connected photovoltaic system

2. The inverter control, which has the main goals:

- Control the active and regulate the reactive power injected into the grid; - Control the DC bus voltage; - Ensure high quality of the injected power;

3. ELECTRICAL MODEL OF

PHOTOVOLTAIC CELL

To find the model of the photovoltaic generator, we must first find the electrical equivalent to that source. Many mathematical models have been developed to represent their highly nonlinearity resulting from that of semiconductor junctions that are the basis of their achievements. Found in the literature several models of different photovoltaic generators them through the procedure and the number of parameters involved in the calculation of voltage and current end-of photovoltaic generator. We will present our work in the model with two diodes; in fact this model takes into account the different internal resistance of the PV cell (Fig.2) [1].

Fig.2. Model of a ohotovoltaic cell with two diodes

It consists of a current source iph that models the conversion of light energy flow, electrical resistance Rp shunt is a consequence of leaks by the side effect on the photovoltaic cell, a series resistance R,;, representing the various resistance contact and connection and two diodes 01 and D2 in parallel model the PN junction [2].

The current generated by the module is given by the following equation:

- - .( q·(V+R,·I) J- .

( q·(V+R,·I) J V+R,·I I -Iph IS] eXJl A. K T 1 IS2 eXJl

A,. K T I D ] _''1'

(I)

Where V and I represent the output voltage and current of the PV, respectively; Rs the series parasitic resistance of a solar array and Rsh its shunt parasitic resistance; q is the electronic charge; Iph corresponds to the light­generated current of the solar array. ISb2 represent the current saturation of the two diodes ; AJ,2 is ideality factor of the junction of 01 and D2, K the Boltzmann's constant, T the cell temperature.

From equation (1) we notice that the output current of the PV module depends on the photocurrent itself, which itself depends on the solar insulation and the junction temperature of the cells of the module, consequently the power which a module can deliver depends on the solar insulation and the temperature of the junction. The characteristics of a PV cell of changes in current and power based on the voltage of the PV cell is shown in figure 3.

, , . � ---------- r ---------- t ---------- t -------

· . . · . . I I , , �� ----------� ---------- T---------- --------- , --- ----- � " ,

� ----------�----------�---- -----�----------�-----· . · . , " 1.5 -- - -- - -- - -r - -- - -- - - - T---------- T ---------- , ------I , , I I I I I · . .

I __________ � _ _ ______ � __________ � __________ � ________ _ � • I , I , , , I I I , ___ �--------_-�--------_-�------_---�--------- 10

Fig.3. Power and current voltage characteristics

2010 7th International Multi-Conference on Systems, Signals and Devices

4. FUZZY LOGIC MPPT TRACKING

CONTROLLER

The maximum power that can be delivered by a PV panel depends greatly on the insulation level and the operating temperature. Therefore, it is necessary to track the maximum power point all the time. Recently fuzzy logic controllers have been introduced in the tracking of the MPP in PV systems (Fig.4). They have the advantage to be robust and relatively simple to design as they do not require the knowledge of the exact model. They do require on the other hand the complete knowledge of the operation of the PV system by the designer. The two FLC input variables are the error E and change

e

ce

CE SCE Fig. 4: General diagram of a fuzzy of error CE at sampled times k defined by[3]: { E(k) == P(k) -P(k -1)

V(k)-V(k-I) CE (k) == E(k) -E(k -I)

(2)

Where P(k) is the instantaneous power of the photovoltaic generator. The input E(k) shows if the load operation point at the instant k is located on the left or on the right of the maximum power point on the PV characteristic, while the input CE(k) expresses the moving direction of this point. The fuzzy inference is carried out by using Mamdani's method, (Table 1), and the defuzzification uses the centre of gravity to compute the output of this FLC which is the duty cycle:

I (daJ-daj da == j�l (3)

IIl(daJ j�l Where ex is duty cycle the associated fuzzy sets involved in the fuzzy control rules are defined as: NB: Negative big; NS: Negative small; ZE: Zero; PS: Positive small PM; PB: Positive big. The control rules are indicated in Table 1 with E and CE as inputs and dex as the output. These two variables and the control action ex for the tracking of the maximum power point are illustrated in figure 5 [4].

Table 1: Fuzzy rule table

� NB NB ZE NS ZE ZE PS PS NS PB NB

EX&n 1 NB NS

/

0 0.CiJl1 0.0l.6 .o.OOB

NS ZE ZE ZE NS NB

ZE

0

ZE PS PB PB PB PB PS PS PS ZE ZE NS NS ZE ZE NB ZE ZE

PS PB

(a)

0.00$ OC16 I

0_OJ2

E •

CE •

-0.032. ..a.ou; -0.008 o.oos 0.016 0.032

Fig.S . Membership for inputs and outputs

Obviously, it can be deduced that the fuzzy controller is fast controller in the transitional state and presents also a much smoother signal with less fluctuations in steady state. A fast and steady fuzzy logic MPPT controller was obtained. It makes it possible indeed to find the point of maximum power in a shorter time runs (Fig.6).

- -

50 - - - - �--- - -� ----1-----l- - - - j -- - - -�-----1--- ;- - - - - � ----

40 - - - - �--- - -� - - - - � -- - - -� - - - - � - - - - -: - - - � - - - - -1- - - - - r ----

30 - - - - �--- - -� - - - - j -----l- - -� -- - - -�-----l-- - - -1- - - - _ 1 _ ---. . ��j

20 - - - - -l- - - - + ---- ; - - - + - - - - i - - - - -1- - - - - t - - - - -1- - - - - t ----10 - - - - �- - - -� - - - - j - - - - -� - - - - j - - - - -�- - - - - � -- - - -1-- - - - f ----

°0 2 4 6 8 10 12 14 16 18 20 'v (v:> Fig.6. PV characteristic MPPT under standard climatic conditions.

5. MODELING AND CONTROL OF THE THREE LEVEL NPC VSI

The three-level NPC VSI, presented in figure 1, is one of the most commonly applied multilevel topologies [5].This type of VSI has several advantages over the standard two-level VSI, such as a greater number of levels in the output voltage waveforms, lower dV/dt, less harmonic distortion and lower switching frequencies. The main draw-back of this type of

2010 7th International Multi-Conference on Systems, Signals and Devices

converter is the voltage imbalance produced in the capacitors of the DC-link when one of the phases is connected to the middle point or Neutral Point (NP). Each of three legs of the converter consists of four power switches, four freewheeling diodes and two clamping diodes that limit the voltage excursions across each device to half the input dc-bus voltage [6].

5.1 Knowledge and Control model of a three-level NPC VSI

In order to deduce the knowledge model of the inverter, we introduce the connection function Sis of the switch which describes the state of every switch (1 =c1osed, O=opened). In this function i is the number of the of the commutation cell, iE {I ,2,3} s: is the number of the semi-conductor. The output voltages of the inverter relatively to the middle point M with using the connection functions of

the half-arm «S �1 , S�o) are defined as follows

(4)

In order to develop a control model of this inverter, we define the average model of the knowledge one. Thus, we define the generating function:

/ ) [ 1 r(k+l)T, ] \ug(t) = -J! U (t)dt ,kE N T kT,

e Te--70 (5)

The control model of the three phases three-level NPC VSI, deduced from the system (4), is given by the following equation where <VA>,<V8> and <Vc> represent the mean value of the instantaneous ones.

(6)

Figure 7 represents the simple output voltage of the three-level NPC VSI controlled by the proposed digital strategy with two bipolar carriers. We notice that, for even values of m, the simple output voltage has symmetry relatively to the quarter of the period, and we have only odd harmonics. But for odd values of m, we have no symmetry, and then even and odd harmonics exist. The voltage harmonics gather by families centred around frequencies multiple of 2mf.

6. DC LINK AND POWER CONTROL

To remedy to the problem of the instability of the output DC voltage of the photovoltaic cell, we propose to use a clamping bridge of the input DC voltages of multilevel inverter constituted by a transistor and a resistor in series connected in parallel of capacitor.

0.8 0.8

0.6 0.6 m=12 m=27

OA OA

0.2 0.2

00

Fig.? The digital strategy of the three-level NPC VSl

In this part, the model of intermediate filter with clamping bridge is defined by following equation

C dU cl I . I 1· -cIt = pv -

l d1

-"

C dU ,2 = I + . -

I 2· dt pv 1d2 ,2

(7)

In this part, the intermediate filter with clamping bridge is controlled by following algorithm: jUe> -Ueer = E,

si E, >0 ana

si E, <0 on a

Avec iE {1,2}

Ti=l => lri='Ti.Uei Rei

Ti =' 0 => Iri =' 0

(8)

In order to maintain multi DC bus link voltage Ue, respective of reference Ueref and load variations, we propose to enslave it using fuzzy logic control to stabilize the multi DC bus link voltage controller [7]. In Park frame the network model, is given by the following relation:

(9)

Where 0) is the angular frequency of the three-phase voltage, V dneb V d, id and V qneb V q, iq are the components of Vnet and inet in the d and q axis respectively.

2010 7th International Multi-Conference on Systems, Signals and Devices

In equation (9), the variables have been translated into d-q coordinates and they should be understood as averaged variables. The d-q rotating axes are synchronized to make zero component of the voltage grid (V qne!=O). As a result, the q channel current reference, Iqref is set to zero in order to achieve unity power factor [8].

The term coLne! is current coupled, which will deteriorate the global behaviour of the system. To avoid this problem, a decoupling control scheme is proposed in figure 8. We control the network current of the phase d-q by proportional plus Integral controller [7,8].

Fig.8. Decouple PI-d-q controJIer current

"

GPV

ipv

Direct Power Control

The algorithm of this current loop is given in figure 9. We use a controller fuzzy logic for voltage.

.1 Simulation result

We use the algorithm enslavement elaborated previously (fig.9) to control the active and reactive power of the cascade we applied load variation between two instants t=1.5s and t=2s. This system is simulated by the software Matlab/Simulink environment. Figure 10 shows the simulation results when we use full clamping bridge and feedback control with PI controller. Thus this figure illustrates well the good decoupling of the voltage loops V d and V q because the V q voltage remains practically constant (equal to zero) during the variation. This enabled us to obtain high dynamic performances of the controllers and decoupling. The power active and reactive follows quietly those references respectively. The currents iq and id are practically proportional respectively then active and reactive power. The currents idl is the opposite of the current idZ• The inverter current ido has a mean value practically null.

N e t

Fig.9. Enslavement algorithm of P-Q control

2010 7th International Multi-Conference on Systems, Signals and Devices

____ ..l _____ 1 _____ 1. _____ I _____ L ___ _

----1- - - - -1- - - - - - - - - - -1 - - - - - 1- ----- - - - -t - - - - -1- - - - - + - ----; - - - - - t- - ---

____ ...1 _____ 1 _____ 1. ____ -' _____ L ___ _ I I I I ----1---- -1- - - - -

T - - - - -1 - - - - - 1- ----

v.

-�o�----�.-----_t------�------t------,�-k--v�,�

- - - - ...J - - - - -1- - - - - }-------,J _____ l- - ---I I I I I

- - - -"'T ---- -1- - - - T - - - - 1- - - - -r ----____ ...l _____ 1_ _ _ _.L _ _ _ _ _ ____ L ___ _

I I I I I - - - - 1" - --- -1- - - - T -- - - 1- - - - - r - ---

____ ..1 _____ 1 _____ 1. - --- -"-______ -L ____ --j I I I I I

----,---- -----T----�-----r- p-

%����.-----_t------�------t_----_,��----�P.�.�

, � -J -1

, � -4 -

-� , , , , , 1 - ---- , ---------- - ---------- r ----330 - - - - , - - - - - - - - - - T - ----:- - - - - r - ---320 - - - - "1 - - - - - - - - - - 1" - - - - -1- - - - - t" - - - -310 - - - - -t - - - - - - - - - - 1" - - - - -1- - - - - t- - ---

T ' T�' " " "i"� 290 - - - - -I - - - - - - - - - + - - - -1- - - - - +- - - - --�. � ' . : .,. : � 270 - - - - .J __________ .J. _____ 1 _____ .L ___ _ _ 0 ____ J __________ � __________ L_ U�

I I Uc2 o O.S 1 1.5 2 2.5 3

8000

----1-,

- - - - -t -- - - -1- - -,

- - - - -4 - - - - -I

-; - �

,

t- r

1.'

I �

-, I

- T

- -I< .,.-------,-----1 ...J _____ 1 __

2.'

I _ _ L _

- - 1- - - - - � -

- - 1- - - - - r -

I • I.

I I are'

-----;!o�'-.---__+__ ----.,'-�5---___+__-----;!2 .•• -'-- Qdq 3

---- �-----� ----+---- �-----� ----____ � _____ L ____ � ____ J _____ L ___ _

I I I I I ----l-----r----T----�-----r----- - - - -t - --- -1- - - - - 1- - - - - -I - ---- "' __ �U7cC71-c-:U=cC20i1

-2°0 2.5 3

Fig.! O. P-Q power control and performances Systems

The grid voltage and current are in phases then the

power factor at the grid connection is unity. We show the performances of the FLC of the input voltage of the three-level PWM inverter. We note that, the output voltage of PV follows perfectly its reference which is constant and are no effect for the load variation. Therefore the different input DC voltages of the three-level NPC VSI are constant and practically equal by pairs too (Uc1=Ud, become constant after a transient state and insensible to any perturbation. In

consequence the output voltage of the three-level NPC VSI is symmetrical

7. CONCLUSION

In this work, the aim was to control the voltage of the solar panel in order to obtain the maximum power possible from a PV generator, whatever the solar insulation and temperature conditions. Since quite a few control scheme had already been used and had shown some defects, it was necessary to find

2010 7th International Multi-Conference on Systems, Signals and Devices

and try some other methods to optimize the output, fuzzy logic controller seemed to be a good idea. The controllers by fuzzy logic can provide an order more effective than the traditional controllers for the nonlinear systems, because there is more flexibility. A fast and steady fuzzy logic MPPT controller was obtained. It makes it possible indeed to find the point of maximum power in a shorter time runs. In this paper, we have studied stability problem of the input voltage of the three-level NPC inverter. fed by photovoltaic cell. The application of the clamping bridge shows parfait following of the input DC and his reference and the stability of the input voltage of three-level NPC inverter. have shows is possible to conceiver, with frequency charger using in output the three-level inverter. The results obtained are full of promise to use the inverter in high voltage and great power applications as PV generation system with gird connected.

REFERENCES

[1] D.Y.Lee and al."An Improved MPPT converter using current Compensation Method for Scaled PV-Applications", 0-7S03-0/03/$17.00(C) 2003, IEEE. PP.540-545.

[2] V. Salas, E. Olias, A. Barrado, A. Lazaro, "New Algorithm Applied to Maximum Power Point Tracking without Batteries", 21 st European Photovoltaic Solar Energy Conference, 4-S September 2006, Dresden, Germany [3] T. Esran, P.l. Chapman, « Comparaison of photovoltaic

array Maximum Power Point Tracking Techniques », IEEE Transactions of Energy Conversion 2006. [4] G.Y. Ayvazyan, G.H. Kirakosyan, A.H. Vardanyan, "Maximum Power Operation of PV system Using Fuzzy Logic Control", Armenian Journal of Physics, vol. 1, 200S, pp. 155-159. [5] K.Matsui and al. "Application of parallel Connected NPC­PWM Inverters with Multilevel Modulation for AC Motor Drive" IEEE Transaction on Power Electronics, Vo1.15 .N°.5. September 2000. PP.901-907. [6] L.Yacoubi and al."New nonlinear control of three-phase NPC boosts rectifier operating under severe disturbances". Mathematics and computers in simulation 63(2003) 307-320. Published by Elsevier B.V. [7] 1. Rodriguez and al. "Calculation of the DC-bus Capacitors of the Back-to-back NPC Converters", EPE-EPE-PEMC 2006, Portoroz, Slovenia, PP.137-142. [S] Y. Tian, « Analysis Simulation and DSP Based Implementation of Asymmetric Three-Level Single-Phase

Inverter in Solar Power System", degree of Master of Science, Summer Semester 2007.