Innovative Hybrid Intelligent Robust MPPT for PV Systems in Dynamic Environments

Abstract

This paper proposes a novel hybrid robust intelligent MPPT Control system for enhancing Photovoltaic (PV) system efficiency. Traditional MPPT methods struggle in complex environments. Our solution combines Artificial Neural Networks (ANN) for accurate MPP prediction using historical and real-time data and Sliding Mode Control (SMC) for robust and adaptive control. The proposed sliding surface, based on three error components, ensures rapid convergence to the predicted Maximum Power Point (MPP), even in challenging conditions. Our approach's effectiveness is confirmed through validation under both steady-state conditions and real-world Australian experimental data. The obtained results clearly demonstrate the exceptional performance of our proposed method, achieving almost flawless tracking with a minimal relative error of 0.0001. It exhibits a rapid response time of 38.9 ms, with a difference of 15.16 ms compared to an ANN technique and boasts an impressive efficiency rate of 99.99%, with just a 1.39% difference when compared to an ANN technique under steady-state conditions. When tested with a database of experimental data, our method excels in reducing overshoot and exhibits a power ripple that is only half that of the ANN method. Additionally, it maintains a high efficiency of 98.13%, with a marginal difference of 24.3% compared to the ANN method. Furthermore, it showcases enhanced robustness and stability.