A new neural networks MPPT controller for PV systems

In this paper, an Artificial Neural Network (ANN) MPPT controller has been proposed. The data required to generate the ANN model are obtained from the principle of Perturbation and Observation (P&O) method. The neural network MPPT controller is developed in two modes: the offline mode required for testing different set of neural network parameters to find the optimal neural network controller (structure, activation function, and training algorithm) and the online mode which the optimal ANN MPPT controller is used in PV system. The inputs variables for ANN are the output power derivate (dP) and voltage derivate (dV) corresponding to a given insolation and operating cell temperature conditions, which they have significant influence on the ANN response; the output variable of ANN is the corresponding normalized increasing or decreasing duty cycle (+1 or -1). The proposed neural network MPPT is tested and validated using Matlab/Simulink model for different atmospheric conditions. Results and analysis are presented, many contribution have been demonstrated (response time, MPPT tracking, Overshoot).