Modeling of Adaptive Neuro-Fuzzy Inference System (ANFIS)- based MPPT controller for a solar Photovoltaic Powered Water Pumping System

Abstract

This work develops and simulates an adaptive neuro-fuzzy inference system (ANFIS) controller for maximum power point tracking (MPPT) in a photovoltaic (PV)-driven water pumping application, comparing its performance under dynamic irradiance conditions to the traditional Perturb & Observe (P&O) algorithm. The study addresses delayed response time and reduced efficiency in solar PV-powered Permanent Magnet Synchronous Motor (PMSM) drive systems for irrigation. The PV array's DC output is converted to AC for the PMSM drive via a DC-AC inverter, with speed management achieved through field-oriented controllers (FOCs) that compare the motor's actual speed to a reference. Simulation results in MATLAB/Simulink demonstrate that the controller significantly improves the system's speed responsiveness, reduces harmonic distortions, and delivers enhanced power to the irrigation load. Under fluctuating irradiance, the ANFIS-based MPPT increased output power by 12.5% compared to P&O, achieving a maximum power output of 270 kW at 480 V and 10 A from the PV system, outperforming the P&O approach which produced 240 kW at 475 V and 9.5 A, thereby demonstrating superior power extraction capabilities.