Photovoltaic System Performance Model for Output Power Forecasting

One of the challenges of using photovoltaic (PV) energy is its intermittent nature. Due to this, managing the electrical grid network becomes difficult. Hence, it is important to be able to predict the output power of a PV system. Being able to do so would help plant managers to effectively manage the supply of energy to meet the demand. In this study a photovoltaic system model composed of a PV array model, a DC cable model, and an inverter model was developed, and was used in conjunction with an irradiance forecasting algorithm to predict the output power of a 1 kW grid-connected photovoltaic system. A total of three system models were evaluated, each differing in the inverter model used. When using measured irradiance as input to the system models, the best model obtained a root mean square error (RMSE) of 32.957 W and a coefficient of determination (R2) of 0.978. When using predicted irradiance as input to the system models, the best model obtained an RMSE ofllS.041 W and an R2 of 0.710.