Intelligent Control MPPT Technique for PV Module at Varying Atmospheric Conditions Using MA TLABI SIMULINK

Transcription

1 Intelligent Control MPPT Technique for PV Module at Varying Atmospheric Conditions Using MA TLABI SIMULINK L. Zaghba Unite de Recherche Appliquee en Energies Renouvelables, URAER,Centre de Developpement Des Energies Renouvelables, CDER, 7133 GhardaYa, Algeria N. Terki Electrical Engineering Department, University of Biskra, Algeria Abstract-Maximum power point tracking (MPPT) must usually be integrated with photovoltaic (PV) power systems so that the photovoltaic arrays are able to deliver the maximum power available. This paper proposes two methods of maximum power point tracking using a fuzzy logic and a neural network approach for photovoltaic (PV) module Kyocera KCGT using MAT LAB software. The two maximum power point tracking controllers receive solar radiation and photovoltaic cell temperature as inputs, and estimated the maximum power point and the current and voltage corresponding to it as outputs. The new method gives a good maximum power operation of any photovoltaic array under different conditions (varying atmospheric conditions) such as changing solar radiation and PV cell temperature. From the simulation results, the Neural Network approach can deliver more power and provides a response time response from the tracking system from the point of maximum power and pics lower than the fuzzy logic control. Keywords-PV Module, Kyocera KCGT; Maximum Power Point Tracking MPPT; Artificial Neuronal Networks; Fuzzy System. I. INTRODUCTION Renewable energy sources play an important role in electricity generation. Various renewable energy sources like wind, solar, geothermal and biomass can be used for generation of electricity and for meeting our daily energyneeds.photovoltaic generation is becoming increasingly important as a renewable source since it offers many advantages such as incurring no fuel costs, not being polluting, required little maintenance, and emitting no noise, among others. The photovoltaic voltage-current (V -I) characteristic is nonlinear and changes with irradiation and temperature. In general, there is a point on the V-lor voltage-power (V-P) curves, called the Maximum power point (MPP), at which PV operates with maximum efficiency and produces its maximum output power. The stateof the art techniques to track the maximum available output power of PV systems are called the maximum-powerpoint tracking (MPPT). Controlling MPPT for the solar array is essential in a PV system. There are many techniques have been developed to implement MPPT, these A. Borni Unite de Recherche Appliquee en Energies Renouvelables, URAER,Centre de Developpement Des Energies Renouvelables, CDER, 7133 GhardaYa, Algeria A. Bouchakour Unite de Recherche Appliquee en Energies Renouvelables, URAER, Centre de Developpement Des Energies Renouvelables, CDER, 7133 GhardaYa, Algeria techniques are different in their efficiency, speed, hardware implementation, cost, popularity [1 ][]. One of the most widely used techniques in MPPT is P&O due to its simple and easily implementation. In this paper, intelligent control technique using fuzzy logic control is associated to an MPPT controller in order to improve energy conversion efficiency and compared with Artificial Neuronal Networks method. Simulation and analysis of Artificial Neuronal Networks and fuzzy logic control are presented. II. MODELING AND CHARACTERISTIC OF SOLAR PANEL 1. Modeling a/solar panel The equivalent circuit of the general model which consists of a photo current, a diode, a parallel resistor expressing a leakage current, and a series resistor describing an internal resistance to the current flow, is shown in Fig. 1. The voltage-current characteristic equation of a solar cell is given as: - P - s exp -a- - - R;h (1) IPH is a light-generated current or photocurrent, IS is the cell saturation of dark current, q (= 1. x 1-19 C) is the electron charge, k (= 1.3 x K) is Boltzmann constant, T is the cell working temperature, A is the ideal factor, RSH is the shunt resistance, and RS is the series resistance. Rs I - I h I ( (V+RSI) 1) V+RsI Fig. I The equivalent circuit of a PY cell. TABLE I. ELECTRICAL SPECIFICATIONS OF THE W MUL TT CRYSTALLINE PHOTOYOLTAIC MODULES KYOCERA KCGT Parameter Abbreviation Value Maximum Power PI'V W Tension at Pmax YMPP.3 V IMl'p 7.1A Voc 3.9V Isc.1A NS 5 Idealit factor A I /1/$31. 1 IEEE

2 . Solar panel Characteristic 1 15 G 1Wfm -G W/m G G W/tyt W/m -G W/m T= Co ---.,."\ Fig. V-I Characteristics of PV module at constant temperature (T=5 CO) and varyinginsolation 1===+==+===+==+=lfr=.. :::Jj ""q--= p...,.. %,::==j,5==j to===: r5==-135-j Fig 3.P-V Characteristics of PV module at constant temperature (T=5 CO) 1 and varying insolation G=1Wfm% I T ac" II I II-:: I -T=75C" T 1C" 15 I I :\ '\\ \. '.\\ 3 3. Fig.V-1 Characteristics of PV module at constant insolation (G=1 W/m) and varying temperature ool-ii-i--i---=::::;<i=::;::;::-t---,,.---ii----f z!i!l!l::>q----+-'<-'+l\, i [ j ::1""' T =1 C Fig 5.P-V Characteristics of PV module at constant insolation 3. DC-DC Buck-Boost Converter (G=1 W/m) and varying temperature DC conversion has gained the great importance in many applications, starting from low-power applications to high power applications. In this paper, buck-boost converter is chosen to be used in the MPPT system. Buck-boost converter is used to step down and step up the DC voltage by changing the duty ratio of the MOSFET. If the duty ratio is less than.5, the output voltage is less than the input voltage; while if the duty ratio is greater than.5, the output voltage is greater than the input voltage. Duty ratio is the time at which the MOSFET is on to the total switching time. The buck-boost converter is shown in Figure.The relation between the input and the output voltages of the buck-boost converter is given as follows: [5] [] [7]. Vout = Vin () 1- L CI C fs Resistive Load RL TABLE 1. BUCK-BOOST CONVERTER PARAMETERS ImH 1 flf 33 flf KHZ 5 On applying Kirchhoffs laws, we fmd: dvpv i = pv l_'_d dt Cpv Cpv di L- = (1- D).v + D.vpv dt dv V C - dt -(l-d)i-- = R (3) I is the current through the inductance, the voltage across the capacitor, D is the duty ratio and Vpv is the voltage measured from the photovoltaic panel. Sl l G_$r----r-1 fc_ f 9 I Fig.. The buck-boostconverter circuit. Maximum Power Point Tracking Maximum Power Point tracking controller is basically used to operates the Photovoltaic modules in manner that allows the load connected with the PV module to extract the maximum power which the PV module capable to produce at a given atmospheric conditions.pv cells have a single operating point where the values of the current and voltage of the cell result in a maximum power output. With the varying atmospheric condition and because of the rotation of the earth [3][], the irradiation and temperature keeps on changing throughout the day. So it is a big challenge to operate a PV module consistently on the maximum power point and for which many MPPT algorithms have been developed [1]. The most popular among theavailable MPPT techniques is Perturb and Observe (P&O) method. This method is having its own merits and demerits. The aim of the present work is to develop the Simulink model of artificial neural network MPPT controller and then the fuzzy intelligent control has introduced on it to improve its overall performance. DC/DC VPV Ip" Converter Fig. 7. Block diagram of PV Module with MPPT Controller

3 .1 MP PT using Fuzzy Logic Control Fuzzy logic controllers have been introduced recently in the tracking of the MPP in PV systems. They have the advantage to be robust and relatively simple to design as they do not require complete knowledge of the exact model and it can handle nonlinearity O. The proposed fuzzy logic MPPT Controller, shown in Figure, has two inputs and one output. The two input variables are the error E and change of error CE at sampled times k defined by eq. and 7, where P, V are the PV panel power and voltage respectively at instant k: [][9][1] P(K)pv-P(K-l)pv E(k) = () V(K)pv-V(K -l)pv CE(K) E(K) - E(K - 1) = (5) Where: P(K)pvand V (K)pv are the power and the voltage of the PV generator respectively at instant k. The power of the Pv system: P(K) i(k). V(K) = () The input E(k) shows if the operation point at the instant k is located on the right or on the left of the MPP on the PV characteristic curve as shown in figure 7, while the input CE(k) shows moving direction of this point. Where the control action D is duty cycle of PWM signal that control the Buck Boost converter [11][1][13]. according to professional experience and the operation of the system control. The fuzzy rule algorithm includes 5 fuzzy control rules listed in table I. Fuzzy inference engine is an operating method that formulates a logical decision based on the fuzzy rule settingand transforms the fuzzy rule base into fuzzy linguistic output. In this paper Mamdani's fuzzy inference method, with Max-Min operation fuzzy combination has been used [13][1]. O.r O '-'\.-+_ /--+_ -- _+ /_+_ -- _+I_ -- l O.1---, l.1 Rules E Fuzzil1c<:ltioll Inf'e. ence,..., Dcffuzification D ' I CE / o 5...I Fig.. Block diagram of the fuzzy controller p 7 1 /' V./ V 15 5 Voltage Fig.9. PV cell power curve dp/dv_ '\ \ P' The fuzzy controller design contains the three following steps: Fuzzijication The fuzzification is the process of converting the system actual inputs values E and CE into linguistic fuzzy setsusing fuzzy membership function. These variables are expressed in terms of five linguistic variables (such as ZE(zero), PB (positive big), PS (positive small), NB (negative big), NS (negative small)) using basic fuzzy subsetsas shown in Fig.lO [11][1][13]. Rule base & inference engine Fuzzy rule base is a collection of if-then rules that contain all the information for the controlled parameters. It isset 3 \ 35 < Defuzzijication Fig.IO. Membership function of E, CE and D Defuzzification of the inference engine, which evaluates the rules based on a set of control actions for a given fuzzy inputs set. This operation converts the inferred fuzzy control action into a numerical value at the output by forming the union of the outputs resulting from each rule. The center of area (COA) algorithm is used for defuzzification of output duty control parameter. i.e If E is NB and CE is ZO then crisp D is PB, it means that if the operating point is far away from the MPP by the right side, and the variation of the slope of the curve is almost Zero; then increase the duty cycle. The Output is duty cycle D, is expressed by [][9][1][ 11] [1]: 'LJ=lf.l(Dj).Dj D = (7) 'LJ=l f.l(d j) E/CE NG NP ZE PP PG TABLE 3 FUZZY RULES TABLE NG NP ZE PP PG ZE ZE PG PG PG ZE ZE PP ZE PP PP ZE ZE ZE NP NP NP NP ZE ZE NG NG NG ZE ZE

4 input unit is connected to all radial units on the hidden layer and each radial units on the hidden layer is connected by weighted synapses (represented by w) to the output layer. The synaptic weights are modified during training phase in order to teach the networks the non-linear relationship that exists between inputs and output [15][1][17]. Fig. Il.The input-output surface wavefonn of the FLC. MP PT using artificial neural network.. J Artificial neural network architecture Radial basis functions represent a class of functions whose value increases or decreases as a function of distance to a central point [1,]. They are employed for tasks of interpolation of sets of points in multidimensional spaces. Such a problem is characterized by mapping a vectorial space x of multiple dimensions in a uni-dimensional vectorial space t. The data set consists of N input vectors X', and its corresponding values of f. The goal is to find a function h(x), as in Eq. (): hex) = t. n = N. () The use of radial basis functions has proved to be appropriate for the task of interpolation, with the use of sets of N basis functions, one for each point, being the functions of the form given in Eq. (9) (9) Where ifj is some kind of non-linear function. The argument of function ifj is basically a Euclidean norm (a distance) between two vectors. A kind of radial basis function widely employed is the Gaussian function given in Eq. (1): <fj//i) _exp(llx:i") = (1) Where x corresponds to one of the input vectors x ", having elements Xi, Pj is a vector that specifies the hyper-center for function ifji' having elements fjij, and (J represents a parameter that defines the spread of the function. RBF networks were proposed by the work of Broomhead and Lowe (19), and comprise a class of multi-layer neural networks in which the activation function of each neuron in the intermediate layer is a radial basis function. The concept of an RBF network is illustrated in Fig 1. The radial function in use is usually a Gaussian function, of the kind shown in Eq. (11), in which vector X corresponds to the input vector of the radial unit and fj J represents the center of the radial function. The output layer usually contains neurons that calculate the scalar product of its inputs. In a RBF network having k radial units in the intermediate layer and one output, this is given by Eq. (11): (11) where x and P are defined as in Eq. (), ifj represents the activation function of the radial units, as, for instance, the Gaussian function represented by Eq. (), W, represents the weight values by which the output of a radial unit is multiplied in the output layer and Wo is a constant factor [1].... Artificial neural network training Artificial neural network have memory, which corresponds to the weights in the neurons. The weights and biases of the network are adjusted by the learning rate in order to move the network output closer to the targets. The 'newff function allows a user to specify the number of layers, the number of neurons in the hidden layer and the activation function used as described below. After training, the network weights are set by the back-propagation learning rule. The number of epochs for this example is set to 1 and the learning rate is.. During training, the input vector will be passed through the neural network and the weights will be adjusted 1 times. The learning rate of the network is also set [19][]. The following Matlab code creates a radial basis network [1]: net = newrb(pr,tr,[io, 1], {'tansig','purelin'},'trainlm', 'Iearngdm','msereg'); neurainparam.epochs = 5; neurainparam.goal =.1; net= train(net,pr,tr); gensim(net); III. SIMULlNK MODEL OF PY SYSTEM WITH ARTIFICIAL NEURAL NETWORK MPPT AND Fuzzy LOGIC CONTROLLER C... I.. I(l) V)ou,...? Voll.... ( Fig 1. Schematic diagram of Radial Basis Function Networks The figure shows a typical RBF composed of three layers: an input layer composed of three radial units, a hidden layer where non-linear processing (represented by function ifj) is carried out, and an output layer, containing a single unit. Each Fig 13.simulation Block Diagram of MPPT PY systems using Artificial Neural Network and Fuzzy Logic Controller.

5 A. Operation underconstantconditions In this case the temperature and irradiation are considered constant. It takes the values of standard conditions: temperature5 Cand irradiation in looow/m. B. Operation on Variables Conditions In this case the temperature and irradiation are changing with time under different weather condition. Fig. 1 presents how the irradiance is changing for the PV solar panel. The voltage and the current vary depending on irradiance. The curve of variable irradiance is plotted using a signal builder, where the irradiance is not very realistic, because these are instantaneous changing irradiance. The simulation resultats are shown in the next figures. : I ' Output ofthe Buck Boost Converter - Output ofthe Pv Panel 1 WI ' W/m1 N/m' :,. r- a Fig.17. Input and Output current of the Buck Boost with Artificial Neural Network MPPT controllerat constant temperature (T=5 CO) and varying insolation solutflofsignai Builder: Group 1 11, Fuzzy logic Controller for Maximum Power Point Tracking 1 9 1COOW/rr ! W/m [1 o rum' f COWin' r- -Output ofthe Buck Boost Converter 11" -Output ofthe Pv Panel.9. Time (sec) Fig.1. Variation of irradiance used in simulation. /m' _1 lla.. <uu ",m u, MPPT controller at constant temperature (T=5 CO) and varying insolation '''' r -Output ofthe Buck Boost converter UVVII 1.uu """ - If - 1 Win' r- - Output ofthe Pv Panel Oil' Fig.1. Input and Output Power of the Buck Boost converter with fuzzy WI ' W/m I: I" r-- Output of the Buck Boost Converter - Output of the Pv Panel Fig.15. Input and Output Power of the Buck Boost converter with Artificial.1 Neural Network MPPT controller at constant temperature (T=5 CO) and varying insolation Fig.19. Input and Output Voltage of the Buck Boost converter with fuzzy I\. :r MPPT controller at constant temperature (T=5 CO) and varying insolation 1W/11' W I /m' W/m' W/n { W/m' 3 W/m' O W/m I AA IUY!," ' -Output of the Buck Boost Converter II' -Output of the Pv Panel WI ' Ir Win 1- r"r- -Output ofthe Buck Boost Converter -Output ofthe Pv Panel o C1 (.3. 5 time (5) C (.9 Fig. I.Input and Output Voltage of the Buck Boost converter with Artificial Neural Network MPPT controllerat constant temperature (T=5 CO) and varying insolation Fig.. Input and Output current of the Buck Boost converter with fuzzy MPPT controller at constant temperature (T=5 CO) and varying insolation

6 [] Aurobinda Panda,M.K.Pathak, S.P.Srivastava, "Fuzzy Intelligent Controller for the Maximum Power Point Tracking of a Photovoltaic Module at Varying Atmospheric", Journal of Energy Technologies and Policy, VoU, No., 11. [3] H. E.A. Ibrahim,"Comparison Between Fuzzy and P&O Control for MPPT for Photovoltaic Systern Using Boost Converter", Journal of Energy Technologies and Policy, Vol., No., 1. Fig. 1 Duty cycler variation at constant temperature and varying insolation Figure 15 to presents the results of the simulation model PV panel. The voltage, current and power output of PV panel are represented by red curve and at the output of circuit connected to the photovoltaic panel, they are represented by blue curve. The Irradiance is variable, passing successively through the following values:,, looo, and W/m. To test the operation of the system, the change of solar radiation was modeled. The temperature is fixed at 5 and the level of solar radiation is varied with four levels. The first level of illumination is set at W 1m, at the moment. s the solar radiation level pass abruptly at the second level W/m ), and then the third again looo W 1m), at time O.s. After, he drops at fort level at W/m, at time.7 s and finally passed at the last level G= W/m at time.9 s. According to the simulation results presented above, all quantities to regulate Ipvout, Vpvout and Ppvout converge well to references Ipv, Vpv and Ppv after a time acceptable response t = O.Ols respect to slow dynamics of the profile of the primary source (radiation and temperature). These results show the effectiveness of the algorithm and the relationship between the irradiation and the output power of the PV panel, and show the operation of the Buck-boost converter. From these results we see that the variation of the radiation has a remarkable effect on the functioning of the system. The figures show that the artificial neural network approach provides a response time response from the tracking system from the point of maximum power and pics lower than the fuzzy logic control. CONCLUSION This paper presents the modeling of a photovoltaic (PV) module at varying atmospheric conditions such as irradiation and temperature. It also includes the maximum power point tracking (MPPT) of the PV module using fuzzy logic and a neural network approach. For the performance analysis, the simulation of the PV module with MPPT controller is done by using MA TLAB/Simulink. The Simulation results demonstrated the peak power tracking capability of the proposed fuzzy logic scheme. It was also demonstrated that the Neural Network control improves the tracking performance compared to the fuzzy logic method and, thus, avoids the tuning of controller parameters. REFERENCES [1] A.Terki, A. Moussi, A. Betka & N. Terki, 'The Effectiveness Of Fuzzy Logic Control For Pv Pumping System," Courrier du Savoir - N I, Octobre 11, pp [] Ahmed M. Fares I, Belal A. Abo Zalarn, "Cornparison Between Different Algorithms for Maximum PPT inphotovoltaic Systems and its Implementation on Microcontroller", Journal of Energy Technologies and Policy, Vol.3, No.5, 13. [5] Abdullah M. Noman, Khaled E. Addoweesh, and Hussein M. Mashaly,"DSPACE Real-Time Implementation of MPPT-Based FLC Method",Hindawi Publishing Corporation International Journal of Photoenergy Volume 13, Article ID 5973, II pages. [] Rasoul Rahmani, Mohammadmehdi Seyedmahmoudian,Saad Mekhilef and Rubiyah Yusof,"lmplementation Of Fuzzy Logic Maximum Power Point Tracking Controller For Photovoltaic System", American Journal of Applied Sciences, 1 (3): 9-1, 13. [7] Carlos Meza, Domingo Biel, Juan Negroni, Francesc Guinjoan,"Boost Buck Inverter Variable Structure Control for Grid-Connected Photovoltaic Systems with Sensorless MPPT", IEEE ISIE 5, June -3, 5, Dubrovnik, Croatia [] S.Lalouni, D. Rekioua,"Optimal Control of a Grid Connected Photovoltaic System with Constant Switching Frequency", Energy Procedia 3 ( 13 ) [9] K.V.Hari Prasad, CH.Uma Maheswar Rao," Design And Simulation Of A Fuzzy Logic Controller For Buck & Boost Converters", International Journal of Advanced Technology & Engineering Research (ljater), Volume, Issue 3, May 1. [1] [] D. Sera, T. Kerekes, R. Teodorescu, F. Blaadjerg "Improved MPPT Algorithms for rapidly Changing Environmental Conditions",Power Electronics and Motion Control. [II] G.CD. Sousa, B.K. Bose, "A fuzzy set theory-based control of a phase controlled converter DC machine drive", IEEE Trans. Ind. Appl. 3 (I) (199) 3-. [1] N. Drir,L.Barazane and M. Loudini, "Fuzzy logic for tracking maximum power point of photovoltaic generator", Revue des Energies Renouvelables Vol. 1 N I (13) 1-9. [13] An Cheikh, C. Larbes, G.F. Tchoketch Kebir and A. Zerguerras. "Maximum power point tracking using a fuzzy logic control scheme". Revue des Energies Renouvelables Vol. 1 N 3 (7). [1] Messai, A., A. Mellit, A. Guessoum and S.A. Kalogirou, 11. "Maximum power point tracking using a GA optimized fuzzy logic controller and its FPGA implementation", Solar Energy, 5: DOI:1.11/j.solener.1.1. [IS] Christopher A. Otieno, George N. Nyakoe, Cyrus W. Wekesa, "A Neural Fuzzy Based Maximum Power Point Tracker for a Photovoltaic System", IEEE AFRICON, September 3-5, pp.i-, 9. [1] Abdessamia Elgharbil, Dhafer Mezghanil, Abdelkader Mami, "A Maximum Power Point Tracking Method Based On Artificial Neural Network For A Pv System", International Journal of Advances in Engineering & Technology, IJAET, Nov. 1. [17] A.B.G. Bahgat, N.H. Helwa, G.E. Ahmad, E.T. EI Shenaw. "Maximum power point traking controller for PV systems using neural networks", Renewable Energy 3 (5) [1] Hiyama T. "Identification of optimal operating point for PV modules using neural network for real time maximum power tracking control". IEEE Trans Energy Conv 1995; 1():3-7. [19] T. Hiyama, and al. "Neural Network Based Estimation Of Maximum Power Generation From PV Modules Using Environmental Information", IEEE Trans. on EC, Vol. 1, No. 3, 1997, pp.: 1-7. [] H. Demuth, M. Beale, M. Hagan., "Neural network toolbox user's guide for use with MATLAB". Natick, MA: The Math Works Inc. [1] Sedaghati F, Nahavandi A, Badamchizadeh M A, Ghaemi S, Fallah M A. "PV Maximum Power-Point Tracking by Using Artificial Neural Network", Mathematical Problems in Engineering. 1; Vol. 1.