IEEE PEDS 2011, Singapore, - December 2011 The New Combined Maximum Power Point Tracking Algorithm Using Fractional Estimation in Photovoltaic Systems Dzung Phan Quoc, Quang Nguyen Nhat, Phuong Le Minh, Khoa Le Dinh, Vu Nguyen Truong Dan and Anh Nguyen Bao Faculty of Electrical & Electronic Engineering, HCMC University of Technology, Ho Chi Minh City, Vietnam pqdung@hcmut.edu.vn ; nguyennhatquang29@gmail.com Hong Hee Lee School of Electrical Engineering, University of Ulsan, Ulsan, Korea hhlee@mail.ulsan.ac.kr Abstract- This paper presents an improved algorithm of quick and accurate Maximum Power Point Tracking (MPPT) algorithm which is based on Incremental conductance algorithm, Fractional Open Circuit Voltage and Short Circuit Current The proposed algorithm estimates the short circuit current or open circuit voltage, following by using Fractional Short Circuit Current or Fractional Open Circuit Voltage algorithm to quickly determine point close to MPP MPP will be accurately determined due to Incremental conductance algorithm The proposed algorithm can identify quickly and correctly MPP when the environmental temperature and solar radiation change The results of proposed algorithm are made by simulating with MATLAB/Simulink program and experimenting with kit DSpace DS 1104 Keyword: Maximum power point tracking (MPPT), Incremental Conductance (Inc-cond), Fractional Open Circuit Voltage, Fractional Short Circuit Current, Matlab/Simulink, DSpace DS1104 I INTRODUCTION When a photovoltaic (PV) system is connected to the load, the system will operate at the intersection of the I-V characteristic of the photovoltaic and the load characteristic To increase the effectiveness of photovoltaic system, the photovoltaic system should be operated at the maximum power point Maximum power point is not a fixed point which depends on conditions of environmental temperature and solar radiation Natural environmental conditions are very volatile, so the MPPT controller for photovoltaic systems is very essential The MPPT controller has an impact on the DC-DC converter to inject maximum power at the output before the system is connected to the load or DC-AC converter for grid connection Many MPPT algorithms have been studied and developed ([7], [9], [10]) such as Perturb and Observe (P & O), Incremental conductance (IncCond), Fractional Open Circuit Voltage, Fractional Short Circuit current or ANN based algorithm ([6], [8]) to determine MPPT P & O algorithm is often used in practice because it's simple to implement But this algorithm does not specify MPP exactly when there is rapid change of solar radiation IncCond algorithm overcomes the disadvantages of P & O but the respond time is not fast ANN is a method to determine MPP quickly and accurately ANN algorithm would learn characteristics of a specific photovoltaic panels so when the photovoltaic system is changed, the algorithm must learn a new characteristic In the course of long-term use the characteristics of photovoltaic panel will be changed, resulting in inaccurate algorithm Improvements of MPPT algorithm have been researched and developed in [1-5], [11] Reference [11] proposes a twostage algorithm that offers fast tracking in the first stage and fine tracking in the second stage This method remains a problem of determining VOC (Open Circuit Voltage) This paper presents a new algorithm of MPPT based on improved IncCond algorithm combined with Fractional Short Circuit Current and Fractional Open Circuit Voltage algorithms This method will determine short circuit current if the photovoltaic system operates on the left of I-V characteristics or open circuit voltage if the system works on the right of I-V characteristics Then Fractional Short Circuit Current or Fractional Open Circuit Voltage algorithms will be implemented to put the power around the MPP quickly and then IncCond algorithm is used to determine the exact MPP The proposed algorithm not only identifies quickly and accurately the MPP, but also is not affected by the aging of the system in long term use II THE NEW ALGORITHM The proposed algorithm divides I-V characteristics into three domains: the left, the middle and the right domain as shown in Fig According to initial conditions of whether the photovoltaic system is operating in the left or right domain, the algorithm will determine ISC or VOC At the beginning, if the photovoltaic system operates in the left domain, the algorithm will determine the MPP based on the correspondent value of IREF Assume that in the left domain of the I-V characteristic, there is an almost linear 978-1-4577-0001-9/11/$26.00 ©2011 IEEE 919 Fig I-V Characteristic of a photovoltaic cell relationship between IMPP and ISC of the PV and the value of ISC is determined in the first calculation cycles I −I (1) I sc = I1 + (−V1 ) V2 − V1 In which I REF = K1I SC (2) where K1 is a constant chosen at random with constant K1 = 0.75- 0.92 in Fractional Short Circuit Current algorithm In the next calculation cycles, the photovoltaic system will operate in the middle domain; Incremental conductance algorithm is used to determine IREF so that it operates at MPP At the beginning, if the photovoltaic system operates in the right domain, the algorithm will determine MPP based on the correspondent value of VREF Assume that in the right domain of the I-V characteristic, there is a near linear relationship between VMMP and VOC of the PV array and the value of VOC is determined in the first calculation cycles Fig Flowchart algorithm determines IREF or VREF V −V Voc = V1 + (− I1 ) I − I1 In which VREF = K 2VOC (3) (4) where: K2 is a constant chosen at random with constant K2 = 0.72: 0.78 in Fractional Open Circuit Voltage algorithm In the next calculation cycles, the photovoltaic system will operate in the middle domain; Incremental conductance algorithm is used to determine VREF so that it operates at MPP When environmental conditions vary, the MPP will be changed by photovoltaic system, depending on the variance of current value at the point of change, the algorithm will determine MPP according to the value of IREF or VREF The working principle of the proposed algorithm can be explained using a flowchart show in Fig Let dI dV I b=− V a= (5) The algorithm for determining the value of IREF or VREF being relative to ISC or VOC is shown in Fig.3 III SIMULATION RESULTS OF THE PROPOSED ALGORITHM System model of the proposed algorithm is developed on Matlab/Simulink and SimPowerSystems The photovoltaic system model consists of PV array, the buck - boost converter, load R and MPPT controller as shown in Fig Components of the system include: PV Array: 01 module PV SX 3200 VOC = 30.8 V, ISC = 8.7 A (normal radiation); DC-DC Converter: Buck-Boost Converter with parameters: C1 = 2500µF, L = 1.5 mH, C2 = 5000µF; Load : resistance load Fig Flowchart of the proposed algorithm Case 1: Simulation results of the time response of the proposed algorithm and traditional IncCond algorithm 920 Fig 11 Experimental model using proposed algorithm Fig Block diagram of MPPT controller Case 2: Ability of the proposed algorithm to response to the changes of the environmental temperature conditions Irradiance is constant with λ = 1kW/m2 Ambient temperature in the simulation is: − Time: t = 0s – 0.3 s: T = 250C − Time: t = 0.3s – 0.55s: T = 350C − Time: t = 0.55s – 1s: T = 300C The response is shown in Fig.7 and Fig.8 Case 3: Ability of the proposed algorithm to response to the changes of solar irradiation Ambient temperature is constant: T =250C Irradiation in the simulation is: Fig PV Power Response Curve with Inc algorithm − Time: t = 0s – 0.4s: λ = kW/m2 − Time: t=0.4s–0.8s: λ = 0.2 kW/m2 − Time: t=0.8s–1.2s: λ = 0.8 kW/m2 The response is shown in Fig.9 and Fig.10 Comparing with traditional algorithm, the proposed algorithm has faster response and higher accuracy in case of changing the environmental temperature or radiation intensity Fig PV Power Response Curve (the proposed algorithm) 160 140 P (W ) 180 Time (s) 0.2 0.4 0.6 0.8 IV Fig Power Response Curve with temperature changes (the proposed algorithm) Fig P-V Curve with temperature changes (the proposed algorithm) 100 0 Fig PV Power Curve with irradiation changes (the proposed algorithm) P (W ) 200 Voltage (V) 10 15 20 25 Fig.10 P-V Curve with irradiation changes (the proposed algorithm) Irradiation and ambient temperature in the simulation are: T = 250C and λ = kW/m2 The proposed method tracks to the MPP faster than the conventional IncCond algorithm The proposed algorithm (Fig.5) reaches to the MPP in 0.02s, while the IncCond algorithm (Fig.6) reaches to the MPP in 0.2s EXPERIMENTAL RESULTS The proposed algorithm is implemented on the experimental Kit DSpace DS 1104 to test the ability of the algorithm Components of the system include: PV Array: module PV H-Tech 13W VOC = 10.9 V, ISC = 1.2 A (radiation of lamp); DC-DC Converter: Buck-Boost Converter with parameters: C1 = 3900µF, L = 1.5 mH, C2 = 5000µF; Load: resistance load; Controller MPPT: Kit dSPACE DS1104 set on computer and can communicate with program MATLAB/Simulink The model is shown in Fig 11 Case 1: Proposed algorithm determine MPPT according to VREF Information of photovoltaic system using proposed algorithm is demonstrated in Control Desk Proposed algorithm determine MPP according to VREF for 0.02s (Fig 12) The initial photovoltaic system operates at open circuit voltage and then immediately operates at maximum power point (Fig.13-14) 921 Case 2: Proposed algorithm determines MPP according to IREF Information of photovoltaic system using proposed algorithm is demonstrated in Control Desk Proposed algorithm determines MPPT according to IREF for 0.03s (Fig.15) The initial photovoltaic system operates at short circuit current and then immediately operates at maximum power point (Fig.16-17) Fig 18 PV Power Curve in case Fig 19 I-V Power Curve in 3 Case 3: Traditional IncCond algorithm determines MPP Information of photovoltaic system using IncCond algorithm is demonstrated in Control Desk IncCond algorithm determines MPPT according to VREF for 0.45s (Fig.18) The initial photovoltaic system operates at open circuit voltage and then step by step operates at maximum power point (Fig.19-20) The proposed algorithm can determine MPP accurately by both reference current and Fig 20 P-V Curve in case voltage When there is a change in ambient temperature or radiation intensity, the proposed algorithm has good response, even at low radiation intensity (light sun) V Fig 12 PV Power Curve in case Fig 13 I-V Curve in case Fig 14 P-V Curve in case CONCLUSION The proposed algorithm satisfies the two essential elements in determining MPP, which are fast and accurate response in case of rapid change of environmental conditions of temperature and solar radiation, compared to traditional algorithms The main utility of the algorithm: − Not affected by the property of the photovoltaic − No additional photovoltaic system to determine the open circuit VOC or short circuit current ISC − Good response even when the solar irradiation is low − Identify MPPT by the current and voltage, the calculation algorithm can be implemented easily in the DSP microcontroller in the future ACKNOWLEDGMENT The authors gratefully acknowledge the HCMUT – VNU (Vietnam) and Network-Based Automation Research Center of University of Ulsan (Korea) for providing excellent supports and facilities Fig 15 PV Power Curve in case Fig 16 PV Power Curve in case REFERENCES [1] A.J.Mahdi, W.H.Tang and Q.H.Wu, “Improvement of a MPPT Algorithm for PV Systems and Its Experimental Validation”, International Conference on Renewable Energies and Power Quality Granada, 23 rd to 25 th March, 2010 [2] Carlos A P Tavares, Karla T F Leite, Water I Suemitsu, Maria D Bellar, “ Performance Evaluation of Photovoltaic Solar System with Different MPPT Methods”, ICON 09, 35th Annual Conference of IEEE Industrial Electronics, p.719-724 [3] C Liu, B Wu, R Cheung, “Advance Algorithm for MPPT Control of Photovoltaic Systems”, Canadian Solar Building Conference Montreal, August ,20-24, 2004 [4] C.Hua, C.Shen,“ Study of Maximum Power Tracking Techniques and Control of DC/DC Converters for Photovoltaic Power System”, Fig 17 P-V Curve in case 922 [5] [6] [7] [8] [9] [10] [11] PESC98 Record,29th Annual Conference of IEEE Power Electronics Specialists,Vol.1,p.86-93 J.H.Lee, H.Bae, B.H.Cho, “ Advanced Incremental Conductance MPPT Algorithm with a Variable Step Size”, EPE-PEMC 2006,12th International Conference of IEEE Power Electronics and Motion Control, p.603-607 M.S.Kaiser, S.K.Aditya, R.K.Mazumder, “ Performance Evaluation of A Maximum power Point Tracker (MPPT) For Solar Electric Vehicle Using Artificial Neural Network” Daffodil International University Journal of Science and Technology, Volume 1, issue 1, July 2006 M.A.S.Masoum, M.Sarvi, “ Voltage and Current Based MPPT of Solar Arrays under Variable Insolation and Temperature Conditions”, Universities Power Engineering Conference, 2008 P.Q.Dung, L.M.Phuong, H.H.Lee, L.D.Khoa, N.T.D.Vu, “ The New ANN Maximum Power Point Tracking Algorithm Using ANN- Based Solar PV Systems”, Tencon2010, Fukuoka, Japan T.Esram, P.L.Chapman,“Comparison of Photovoltaic Array Maximum Power Point Tracking Techniques”, IEEE Transaction on Energy Conversion, Vol 22, No 2, June 2007 W.Xiao, W.G.Dunford, P.R.Palmer and Antoine Capel, “ Application of Centered Differentiation and Steepest Descent to Maximum Power Point Tracking”, IEEE Transactions on Industrial Electronics, Vol 54, No.5, October 2007 X.Wang and A.P.Hu, “An Improved Maximum Power Point Tracking Algorithms for Photovoltaic Systems”, Australasian Universities Power Engineering Conference (AUPEC 2004), 26-29 September 2004, Bribane, Australia 923 ... MPP At the beginning, if the photovoltaic system operates in the right domain, the algorithm will determine MPP based on the correspondent value of VREF Assume that in the right domain of the I-V... system, depending on the variance of current value at the point of change, the algorithm will determine MPP according to the value of IREF or VREF The working principle of the proposed algorithm. .. 3: Traditional IncCond algorithm determines MPP Information of photovoltaic system using IncCond algorithm is demonstrated in Control Desk IncCond algorithm determines MPPT according to VREF for