| Abstract: |
This explosion of grid-connected solar photovoltaic (PV) systems has prompted the formulation of increasingly complex design frameworks, as well as advanced automation strategies, to provide reliable, efficient, and stable power delivery. To clearly understand how to design and automate grid-connected solar power, this review paper provides a meta-analysis of previous literature based on intelligent control techniques such as fuzzy logic controllers, artificial neural networks (ANNs), model predictive control (MPC), sliding mode control (SMC), and hybrid optimization methods. A targeted investigation of 150+ peer-reviewed articles published from 2005 to 2024, with a focus on maximum power point tracking (MPPT), inverter control, power quality improvement, grid synchronization, fault detection, and energy management systems. The meta-analysis shows that machine learning-based approaches mixed with control architectures are trending toward enhanced dynamic response, tracking execution, and minimization of harmonic distortion than standard proportional-integral-derivative (PID) controllers. An analysis of literature reveals limitations with respect to computational complexity, partial shading robustness, real-time adaptability, and benchmarking standards. This analysis consolidates recent results across various intelligent control paradigms and identifies opportunities for future research on the interplay of deep reinforcement learning, digital twin frameworks, and edge-computing-enabled autonomous grid management. This paper can act as a reference for researchers and engineers involved in design, optimization and intelligent automation of next generation grid-connected solar energy systems. |
