Accurate photovoltaic (PV) performance modelling is essential for reliable power prediction, energy yield assessment, and system-level decision-making. Various models have been developed for this purpose, but most of them face difficulties in obtaining the full set of parameters required for accurate PV prediction. Conventional empirical and one-diode-based models offer a simple approach with reasonable accuracy for PV modelling; however, they often suffer from deviations under high cell temperature and strong irradiance conditions, particularly for thin film PV technologies. This study has enhanced the one-diode PV performance model by incorporating additional empirical coefficients to improve temperature-dependent behavior while maintaining practical model complexity. The proposed model has been comprehensively evaluated across four PV technologies (monocrystalline silicon (c-Si), multi-crystalline silicon (mc-Si), cadmium telluride (CdTe), and amorphous silicon (a-Si)) and compared to the well-established empirical models (PVWatts and Sandia) as well as conventional one-diode-based formulations under a representative subtropical climatic condition. Results demonstrate that the proposed model consistently improves prediction accuracy across all technologies, with particularly pronounced gains for thin-film modules. Annual nRMSE values are reduced to approximately 2.0% for CdTe and 3.1% for a-Si, significantly showing better accuracy than both empirical and conventional one-diode models. Moreover, the proposed model substantially reduces annual cumulative energy prediction errors, with relative improvements of approximately 30–50% compared with conventional one-diode formulations. These improvements translate into substantially enhanced robustness at the system level, especially for large-scale PV installations. Overall, the proposed model provides a robust and practical framework for accurate PV performance and energy prediction across a variety of technologies and operating conditions. • PV performance modelling is essential for reliable power prediction and energy yield assessment. • The one-diode PV model is enhanced using empirical coefficients to improve temperature effects. • The model is validated across four PV technologies and benchmarked against PVWatts and Sandia. • The model offers a robust and practical framework for accurate PV energy prediction.
Liu et al. (Sun,) studied this question.