Abstract
Accurate estimation of photovoltaic (PV) energy production is essential for ensuring the optimal design, operation, and economic viability of solar installations. However, discrepancies often arise between predicted and actual energy yields, particularly in regions characterized by variable meteorological conditions. The paper presents a synthesis of methods for estimating energy production of photovoltaic systems. For two types of residential systems located in Craiova, a comparative analysis is presented between the estimated values obtained using the most efficient methods and the actual measured values. The research combines simulation-based predictions with real operational data collected over a defined monitoring period to evaluate the accuracy of commonly used estimation models. To estimate solar radiation and energy production, some of the most common specialized software for renewable energy sources were used. The findings reveal that estimation tools tend to overpredict energy output during periods of high temperature and unstable atmospheric conditions, while performing more accurately during stable seasonal intervals. These results highlight the importance of localized calibration of prediction models and the integration of detailed climatic data for improving accuracy. The study contributes valuable insights for system designers, operators, and policymakers seeking to enhance PV system reliability and advance renewable energy planning in Romania and similar climatic regions.