基于GCN-LSTM的分布式光伏超短期出力預(yù)測方法研究
制造業(yè)自動(dòng)化
頁數(shù): 7 2024-08-25
摘要: 氣象數(shù)據(jù)是光伏出力預(yù)測的重要依據(jù),氣象數(shù)據(jù)的質(zhì)量對(duì)預(yù)測的準(zhǔn)確性至關(guān)重要。但對(duì)于分布式光伏系統(tǒng),往往缺乏氣象監(jiān)測裝置,難以對(duì)每個(gè)站點(diǎn)分別提供準(zhǔn)確的氣象數(shù)據(jù)。針對(duì)這一問題,提出一種分布式光伏超短期出力組合預(yù)測方法,將圖卷積神經(jīng)網(wǎng)絡(luò)(Graph Convolutional Network,GCN)與長短期記憶神經(jīng)網(wǎng)絡(luò)(Long Short-Term Memory,LSTM)耦合構(gòu)建預(yù)... (共7頁)