Maximum Power Point Tracking (MPPT) plays a critical role in enhancing the efficiency of grid-connected photovoltaic (PV) systems under varying environmental conditions. Conventional MPPT algorithms such as Perturb and Observe (P&O) and Incremental Conductance (IC) suffer from slow convergence, steady-state oscillations, and reduced performance during partial shading conditions. Recent advancements in artificial intelligence (AI) have introduced adaptive and data-driven MPPT techniques capable of overcoming these limitations. This paper presents a comparative performance analysis of AI-optimized MPPT algorithms for grid-connected solar PV systems, focusing on Artificial Neural Networks (ANN), Fuzzy Logic Control (FLC), and Reinforcement Learning (RL)-based MPPT schemes. A comprehensive review of 15 prior studies reveals gaps related to generalization capability, computational complexity, and real-time adaptability. To address these limitations, this work proposes a hybrid AI-optimized MPPT framework integrating deep reinforcement learning with adaptive filtering to improve tracking accuracy and dynamic response. The methodology involves simulation-based data generation, preprocessing, model training, validation under standard and partial shading conditions, and comparative evaluation against conventional MPPT methods. Results demonstrate that the proposed approach improves tracking efficiency by up to 3.8% and reduces convergence time by 42% compared to traditional techniques. The findings confirm the effectiveness of AI-based MPPT optimization for reliable and efficient grid-connected PV operation.
Ashraf Md Sazid (Thu,) studied this question.