Power Output Prediction of Individual Photovoltaic Modules Under Operating Photovoltaic System Conditions

Abstract

Accurately predicting power output details of individual photovoltaic (PV) modules is crucial for evaluating and controlling operating PV systems. Although many techniques have been developed to address this aspect, accurately detecting and predicting the power output of an individual module in a large-scale PV operating system remains challenging. To improve the accuracy and efficiency of predictive technology, a new method is proposed to extract five physical parameters based on the single-diode model (SDM) of a PV module. The proposed technique only requires the initial electrical performance of a PV module under standard conditions. In addition, the initial voltage and current, which depend on solar irradiance, the average temperature, as well as the degradation rate, are used to solve the physical parameter functions of an operating PV module. Consequently, these five parameters and the maximum power output of an operating module can be calculated with the proposed model. To determine the accuracy of the proposed approach, the relative error (RE) and mean absolute percent error (MAPE) of individual PV modules in operating PV arrays and strings are examined. The results indicate that the minimum RE of the power output of an individual PV module occurs near the maximum power point (MPP). The experimental results of both the PV arrays and strings indicate that the MAPE of power prediction for an individual module at the MPP is lower than that reported in previous research. Moreover, solar irradiance measurement accuracy and stability are the main factors influencing the error in this study. Comprehensive experimental result analysis demonstrates that the proposed real-time technique is suitable for large-scale PV farm testing integrated infrared imaging in terms of accuracy, reliability, and efficiency.