Optimal hybrid power dispatch through smart solar power forecasting and battery storage integration

This study presents a strategy to optimize hybrid power system dispatch for commercial sectors in South Africa while utilizing the day-ahead method to forecast solar photovoltaic (PV) power. The approach utilizes numerical weather prediction (NWP) models obtained from open weather maps and incorporates power plant specifications to generate predictions of the PV power plant's output. These predictions are then integrated into an optimal control strategy incorporating battery storage. The use of optimal algorithms helps manage PV power plant curtailment during periods of over-generation. It is crucial to optimize PV power systems and ensure a continuous power supply for solar power plants, even during unfavorable weather conditions. Besides, the study develops a model that solves the challenging questions of combining solar power forecasting with an optimal dispatch and demand management scheme. Therefore, there is a need to incorporate battery storage systems through the developed optimal control method to maximize the energy from the PV system and minimize the power from the utility grid. The obtained results demonstrate the effectiveness of the developed model. The winter season presented a lower MAE of 21 kW , an RMSE of 35.4 kW , and a MAPE of 3,1% for PV power output forecasting, showing that the errors during prediction are lower compared to other seasons. It has been observed that 60% of the load is supplied through a combination of PV power and battery storage. Therefore, evidence of the developed optimal hybrid power dispatch with an innovative solar power forecasting model suggests that accurate forecasting can improve system planning and mitigate the necessity of procuring grid power at high electricity prices.