An efficient MPPT controller using differential evolution and neural network

Performance of the photovoltaic (PV) system is highly dependent on the ambient conditions i.e irradiation and temperature. It has non-linear P-V characteristics that will vary with irradiation and temperature, which will affect the output power of PV array. This nonlinear behavior becomes more complex in partial shading and rapidly changing irradiation conditions. Conventional Maximum Power Point Tracking (MPPT) methods fail to track and extract the maximum power from the PV array in such conditions. Another problem with the conventional methods is the steady state oscillations. All these factors result in power losses. This paper presents a new method for the tracking of Maximum Power Point (MPP) based on Differential Evolution (DE) and Artificial Neural Network (ANN). DE has the capacity to optimize the non-linear problem without the use of gradient and ANN has the ability to model complex relationship between the inputs and outputs. Combining both techniques will result in a better controller. The proposed controller will adjust the Duty ratio `D´ of the Boost converter to track maximum power from PV array and gives the constant output voltage. The proposed MPPT method has been developed and simulated using the MATLAB software package. Analysis and comparison show that proposed controller can track the MPP in less time compared to conventional MPP methods and without any fluctuation in steady state. The robustness of the proposed controller has been demonstrated in the partial shading and rapidly changing irradiation conditions.