PERFORMANCE ANALYSIS OF NEURAL NETWORK BASED MPPT CONTROLLER FOR SOLAR PV SYSTEM WITH CONVENTIONAL AND CASCADED BOOST CONVERTERS

Abstract

Among all renewable energy sources, solar photovoltaic (PV) represents a very important and reliable energy source. However, the output of the PV module is limited. The system performance in renewable energy sources is improved using DC - DC converters. Boost converters are used if output voltage higher than the PV module is desired. If further higher voltage step-up ratio is required by the solar PV system for which the performance of traditional boost converter declines, then cascaded boost converter configurations are employed. Also, besides using cascaded converters for voltage improvement, it is desired that photovoltaic (PV) power systems extract maximum power from the solar module for efficient operation. The issue of operating the Solar PV Module at the maximum power point at all operating conditions is resolved by applying maximum power point tracking techniques. This project is focused on the implementation of P&O based MPPT controllers in solar PV systems for conventional, quadratic and double cascade boost converters connected to a resistive load and investigates their performance at varying solar radiation and ambient temperature. Then the project is carried forward using Artificial Neural Network as a controlling technique for the MPPT algorithm in order to provide a fast and efficient response from the controller. Furthermore, an impedance load of resistive-inductive type is used to record the performance of the system. All the work in this project is accomplished through simulation of the power and control circuits using MATLAB/Simulink software and results and waveforms have been recorded accordingly.