Modeling and Optimization of Solar Panel Cooling using Machine Learning and Deep Learning Techniques

Abstract

The efficiency of photovoltaic (PV) solar panels significantly decreases when their surface temperature exceeds optimal operational limits. Conventional cooling methods have the disadvantage of poor adaptation to changing environment and therefore poor energy output. This study presents a novel approach for modeling and optimizing solar panel cooling systems using deep learning algorithms and artificial neural networks (ANNs). The real-time temperature and irradiance and ambient conditions are incorporated into a model that calculates the thermal performance and applies the model to dynamically control the cooling mechanism. The predictive modelling was conducted using convolutional neural network (CNN), and the optimization process of the real-time selection of the most effective cooling strategy was carried out using reinforcement learning (RL). This study has shown the promise of intelligence-focused cooling control powered by deep learning to optimise PV systems and make them more sustainable.