Optimizing solar energy generation through machine learning: A framework for CUF analysis

This study examines the impact of meteorological variability on the Capacity Utilization Factor (CUF) of solar photovoltaic systems across India, using a multi-method analytical framework that integrates statistical, time-series, and machine learning approaches. Drawing on daily performance and weather data from over 60 solar-enabled facilities including a detailed case analysis of the Sai Mitra solar plant in Prasanthi Nilayam—the study evaluates the influence of 38 weather variables on CUF over a one-year period. The analysis revealed that seasonal effects are pronounced, with winter months yielding the highest average CUF (13.97) and monsoon the lowest (12.93). Maximum temperature was identified as the most significant predictor of CUF (β = 0.71, p < 0.001), while the interaction between short-term temperature anomalies and long-term warming showed a diminishing effect (interaction term β = −0.142, p < 0.001). Lagged performance also emerged as a strong determinant of current CUF, with facility-level regressions indicating that prior CUF values accounted for up to 43.9 % of variance in some locations. Inter-facility dependencies were also observed; for instance, CUF at facility R2 was significantly influenced by lagged performance at R1 (β = 0.177, p < 0.05), highlighting regional coherence in weather effects. These findings underscore the value of integrating diverse real-time historical weather datasets that influence CUF modeling frameworks for enhancing solar performance forecasting. The combined fusion of machine learning, statistical and time-series approaches, and big data in this research study, is particularly crucial for modeling the complex dynamic interdependencies that influence CUF with improved reliability, in comparison to conventional statistical models. Furthermore, the research supports data-informed operational planning, climate-responsive infrastructure design, and the formulation of adaptive energy policies in weather-sensitive regions.