Optimized prediction of electrical and thermal performance in solar PV/T systems using cloud-based data acquisition and advanced ANN models

Abstract This study develops a cloud-monitored platform for predicting electrical and thermal performance in photovoltaic/thermal systems using back propagation (BP) and radial basis function (RBF) models. Genetic algorithm (GA) and particle swarm optimization (PSO) were applied to optimize the BP and RBF models. Results show that all four models exhibit high prediction accuracy, with the GA-BP model achieving the best performance. The average prediction errors for electrical power, electrical efficiency, thermal storage tank temperature, and thermal efficiency were 0.74%, 0.97%, 0.47%, and 0.32%, respectively, significantly improving precision over conventional models.