Review on the Theory and Methods of Photovoltaic Output Forecasting
The large-scale grid connection of photovoltaic power generation has brought great challenges to the stability of power system operation in China.Therefore,the accurate pre-diction of photovoltaic power generation output is very important.This paper systematically summarizes the theories and methods of photovoltaic output prediction.Firstly,the photo-voltaic output prediction is classified,especially divided into point prediction and uncertainty prediction according to the prediction form.Secondly,the photovoltaic output prediction is further elaborated through physical methods,statistical methods,artificial intelligence methods and combination methods.The methods of artificial intelligence are introduced in detail from two aspects:machine learning and deep learning.Then,the evaluation indexes of point prediction and uncertainty prediction are combed,and the optimization techniques of the artificial intelligence prediction model are summarized.Finally,according to the develop-ment status of photovoltaic output prediction in China,the future research trend prospects.
PV power generationArtificial intelligence algorithmuncertainty forecaste-valuation indexoptimization of forecast model
万方数据
维普
