In the past decade, Artificial Intelligence (AI) has become a revolutionary power in the solar energy industry, dealing with essential issues related to intermittency, resource forecasting, efficiency of systems, and upkeep. This survey thoroughly examines and synthesizes AI methods—ranging from classical models such as artificial neural networks (ANN) and support vector machines (SVM) to current methods like deep learning, hybrid models, and reinforcement learning—that are employed in solar energy optimization between 2013 and 2023. We analyze the technical merits and demerits of each model along with its practical performance across applications like irradiance forecasting, energy management, and predictive maintenance. Experimental comparisons, case studies, and an advocated theoretical framework are included to substantiate the findings. The review delineates research gaps like unavailability of standard datasets, poor interpretability of models, and difficulty in model deployment in data-scarce domains. It concludes with a forward-looking discussion on trends like edge-AI, federated learning, and explainable AI. This piece of work is set to inform researchers, policymakers, and stakeholders in the industry on how to better utilize AI to facilitate global solar energy uptake.
Ishu Anand Jaiswal (Mon,) studied this question.