An adaptive TS-fuzzy model based RBF neural network learning for grid integrated photovoltaic applications

Abstract

In this research work, an adaptive TS-fuzzy based RBF neural network (ATSFRBFNN) algorithm based maximum power point tracker (MPPT) is employed for grid integrated photovoltaic (PV) system. The novel learning algorithm provides accurate and rapid PV power tracking under fluctuating solar insolation. However, an improved damping circulating current limiting inverter controller is employed to generate gating signal of voltage source inverter and to deliver reduced harmonic constant, less power loss, mitigation of power quality issues regulation of Dc-link voltage utilization for grid integrated PV system under steady, and transient weather conditions. The experimental results justify the improved system performance under varying operating conditions. The behaviour of novel hybrid PV MPPT structure is compared with TS-fuzzy, RBFNN, and GA-RBF for higher tracking efficacy, least tracking deviation, accurate responses, and zero steady-state power oscillation around MPP region.