MPPT Design for Photo Voltaic Energy System Using Backstepping Control with a Neural Compensator

Abstract

It is very important to have maximum power point trackers for photo voltaic systems to improve their efficiency. This paper deals with the converter based maximum power point tracking by robust backstepping controller along with the neural network. The neural network provides the output reference PV voltage to the backstepping controller. Back propagation neural network is used for a standalone photovoltaic system under robust environmental conditions. Unlike Conventional   solar-array   mathematical   model, neural network does not require any physical data for modeling since it has the superior potential to derive non-linear models without requiring the physical data of the models. In  this  paper  the maximum power point of photovoltaic module is predicted with the simulation trained back-propagation  neural network using a  random  set  of  data  collected  from  a  real  photovoltaic  array. The neural network based PV system with backstepping controller is modeled in MATLAB/Simulink. At different atmospheric conditions the developed model is simulated. The simulation results of PV system depict that with the proposed converter based controller, the maximum power is tracked accurately and successfully.