This study analyzes the performance and predictive modeling of solar photovoltaic (PV) systems at the Bui Generating Station in Ghana using the XGBoost (Extreme Gradient Boosting) algorithm. The predictive model, validated through Monte Carlo simulations, demonstrates measured stability across perturbation scenarios. Distribution analysis confirms appropriate parameter bounds, while error analysis demonstrates consistent pattern preservation across simulation scenarios. The study quantifies the relative influences of environmental factors, particularly the interplay between temperature, irradiance, and humidity (correlations ranging from −0.33 to 0.36). These findings provide insights for system operation while acknowledging the complex, often weak coupling between environmental parameters. Seasonal performance analysis reveals distinct optimization windows, with the Post-Rainy season showing the highest stability (PR: 0.986 ± 0.082) and optimal enhancement potential. Sensitivity analysis identifies critical operational thresholds, including performance transitions at 80 % relative humidity and optimal temperature ranges below 32 °C, where each 1 °C reduction yields 0.45 % efficiency gain. The study establishes specific optimization strategies including automated cleaning systems triggered at 85 % peak irradiance, yielding 2.5 % efficiency improvement, and enhanced inverter response protocols during peak generation periods, achieving a 3.2 % performance gain. These findings inform practical implementation frameworks for performance optimization, contributing to improved energy generation efficiency and system reliability.