Short-time photovoltaic power prediction considering relative influential factor based on typical weather conditions

Photovoltaic (PV) power output is directly related to the weather conditions and the prediction accuracy of PV power decreases for cloudy and rainy weathers. Based on the categories of typical weather conditions, a PV power forecast model is proposed with consideration of the predictive relative tolerance and the relative influential factor (RIF). At first, the climate data are categorized according to the definition of typical weather conditions and the RIF is put forward; Then the probabilistic model of predictive relative tolerance is established by using the t Location-Scale distribution and the Latin hypercube technique is used for sampling of the predictive relative tolerance; Finally, the predictive relative tolerance and the predicted value are superimposed to obtain the final prediction results. A case study has proved the feasibility and effectiveness of the model.