A Novel Maximum Power Point Tracking Control of Photovoltaic System Under Partial and Rapidly Fluctuating Shadow Conditions Using Differential Evolution Hamed Taheri*, Zainal Salam*, Kashif Ishaque* and Syafaruddin** *Faculty of Electrical Engineering, Universiti Teknologi Malaysia (UTM), Johor 81310, Malaysia **Department of Computer Science and Electrical Engineering, Kumamoto University, Kumamoto 860-8555, Japan E-mail: taheri@fkegraduate.utm.my Abstract—Photovoltaic (PV) system performance extremely depends on local insolation and temperature conditions. Under partial shading, P-I characteristics of PV systems are complicated and may have multiple local maxima. Conventional Maximum Power Point Tracking (MPPT) techniques can easily fail to track global maxima and may be trapped in local maxima under partial shading; this can be one of main causes for reduced energy yield for many PV systems. In order to solve this problem, this paper proposes a novel Maximum Power Point tracking algorithm based on Differential Evolution (DE) that is capable of tracking global MPP under partial shaded conditions. The ability of proposed algorithm and its excellent performances are evaluated with conventional and popular algorithm by means of simulation. The proposed algorithm works in conjunction with a Boost (step up) DC-DC converter to track the global peak. Moreover, this paper includes a MATLAB-based modeling and simulation scheme suitable for photovoltaic characteristics under partial shading. Keywords—MPPT; Partial shading; Differential Evolution (DE); Photovoltaic; Boost Converter I. INTRODUCTION The ever-increasing energy demand and growing concern about environment has sparked enormous interest in the utilization of renewable energy (RE). One main source of RE is the sun. Since this energy source is free, abundant, sustainable and environmental friendly, solar electricity, or more popularly known as photovoltaic (PV), is envisage being one of the major energy source of the future [1]. PV can be used in a wide range of applications- from power supplies for satellite communications to large solar power stations feeding electricity into the grid [2]. The grid connected PV power system has a very large commercial potential. Despite its rapid growth, there remain several challenges that hinder the widespread use of PV power systems. The main limitation is the high cost of the module. The cost is basically determined by the economics of scale, supply-demand, price of basic the semiconductor material and improvement in the cell manufacturing processes. These issues are extensively discussed elsewhere and are beyond the scope of this paper. Another type of problem is directly related to PV system installation itself. One common situation is the mismatching effects of the PV output due to broken parts or partial shading of the modules. The latter, which is of more serious concern, may be the result of sudden cloud changes in the sky, obstruction of buildings, trees, poles etc. To maintain high efficiency, PV system should be operated optimally in all conditions including during partial shading. To obtain the maximum power, PV system is normally equipped with Maximum Power Point (MPP) tracker in their power converter control algorithm [3]. Many MPP tracking techniques have been proposed in recent years. They are generally categorized into the following groups: 1) Perturbation and observation methods; 2) Incremental conductance methods; 3) Other new approaches (Artificial intelligence method e.g. Fuzzy and Neural Network and etc.) The Perturbation and Observation (P&O), which works satisfactorily when the irradiance varies very slowly, is widely used in photovoltaic systems [4]. However P&O often fails to track global MPP when irradiance changed suddenly. This is due to the alterations in power levels caused by suddenly irradiance changing that may confuse the MPP tracker to find the real peak point. Another drawback of P&O is that the system oscillates consistently about the MPP. The constant oscillation in known be one of the major sources of the reduced MPPT efficiency. It can be minimized by reducing the perturbation step size. But a smaller perturbation size slows down the MPPT execution significantly. Another popular approach is the Incremental Conductance method whereby the MPP are detected by comparing for each step the derivative of conductance with the instantaneous conductance [5]. The derivative of conductance and instantaneous conductance can be calculated by sensing the instantaneous and the previous PV voltage and current values. The increment size determines how fast the MPP is tracked. Fast tracking can be achieved with larger increments but with the expense of accuracy. Furthermore oscillation around the MPP may occur. Recently artificial intelligence methods which include Fuzzy [6] and Neural Network [7] have been applied to track the MPP. The main strength of these methods is their abilities to deal with the nonlinear characteristics of the I-V curves. For Fuzzy, it is necessary to create control rules that meet the output characteristics 2010 IEEE Symposium on Industrial Electronics and Applications (ISIEA 2010), October 3-5, 2010, Penang, Malaysia 978-1-4244-7647-3/10/$26.00 ©2010 IEEE 82