Improved maximum power point tracking of solar pv systems for faster and accurate convergence

Abstract

The maximum power of a PV module represents variations due to temperature, irradiation and load. In the conventional mode maximum power point tracking algorithms are applied to maximize efficiency, reliability by constantly extracting maximum power. The conventional methods that mentioned in literature have several disadvantages in terms of efficiency, accuracy and flexibility specifically under varying weather conditions. It is mainly because of non-linearity in PV module current-voltage characteristics as well as DC-DC converters. Under this project new intelligent control methods for maximum power point tracking will be tested. Basically, fuzzy logic-based hill climbing method will be proposed and tested to obtain faster and accurate converging to the maximum power point during steady state and varying weather conditions. This artificial intelligence approach would simplify exiting methods and provide with proper modeling of nonlinear systems. In achieving this goal, maximum power point system consisting PV module, buck, boost, buck-boost converter, fuzzy logic controller is designed and simulated using Mat lab Simulink and experimentation studies would be carried out. In the latter stage, it is proposed to extend this project by combining fuzzy logics and neural networks so that the system can identify its maximum power point by itself through self-learning rather implementing only an algorithm as in conventional methods.