Machine Learning-Based Performance Prediction Model For Solar PV Systems Using Meteorological Inputs

Abstract

Accurate performance prediction of solar photovoltaic (PV) systems is crucial for efficient energy planning and grid integration. This study develops a machine learning-based prediction model that utilizes key meteorological inputs—such as solar irradiance, ambient temperature, humidity, and wind speed—to forecast PV power output. Using advanced regression and ensemble learning techniques, the model is trained and validated on real-world datasets to ensure robustness across varying climatic conditions. Results show that the proposed approach significantly improves prediction accuracy over traditional empirical models, supporting better operational planning and integration of renewable energy into the power grid. This work demonstrates the potential of data-driven models in enhancing the reliability and efficiency of solar PV systems, contributing to sustainable energy transitions.