mostoftheresearchesonPVpowergenerationforecastingmethodshaveproblemssuchaslongtimeformodeltrainingandproposeanoptimization.UsingtheBP(backpropagation)neuralnetwork,thislearningalgorithmismainlyapplicabletomulti-input,multi-outputnetworks.Itcanrelyonready-madedataandinputandoutputwithou