PV output prediction based on hybrid method of ensemble learning and Gaussian process
Ensemble learning is widely used to time-series industrial application,such as photovoltaic(PV)output forecasting,but it suffers from low fitting accuracy and point prediction only without enough training dataset and industry knowledge.To solve this problem,a hybrid method based on ensemble learning and Gaussian process to predict PV output is proposed in this paper.Regarding with the point prediction of ensemble learning of several algorithms,the Gaussian process algorithm is utilized to provide confidence intervals,which has better generalization in prediction.By actual case from PV platform,it illustrates the application of the proposed method.
万方数据
维普
