COMPARATIVE PERFORMANCE ANALYSIS OF A FEED-FORWARD NEURAL NETWORK-BASED MPPT FOR RAPIDLY CHANGING CLIMATIC CONDITIONS

Abstract

Rapid and abrupt changes in climatic conditions present a challenge to classical MPPT techniques as they drift from the MPP, resulting in loss of power. This paper presents a new MPPT technique based on a feed-forward artificial neural network (FFANN) and a direct control technique. In the proposed approach, FFAAN estimates the optimum value of the PV output voltage V_MPP, while the direct control technique achieves an optimal adjustment of the duty cycle making the operating point at MPP. To evaluate the performance of the proposed technique, the accurate electrical model of the system parts was built and simulated in MATLAB/Simulink environment. The simulation results are collected under rapidly changing climatic conditions. Simulation results show that the proposed MPPT technique achieves higher performance in terms of tracking efficiency and convergence speed compared to both the IC-based MPPT and FL-based MPPT systems. The results show that the proposed technique accurately estimates V_MPP, achieving a tracking efficiency of 99.9%, while the tracking efficiency is 94% when using FL-based MPPT and 91.5% when using IC-based MPPT. This demonstrates that the proposed technique exhibits superior performance under rapidly changing climatic conditions and increases energy production efficiency compared to classical techniques.