基于GWO-VMD-LSTM的熱力系統(tǒng)單參數(shù)時(shí)間序列預(yù)測(cè)方法
熱能動(dòng)力工程
頁(yè)數(shù): 9 2024-06-20
摘要: 針對(duì)熱力系統(tǒng)參數(shù)運(yùn)行數(shù)據(jù)預(yù)測(cè)困難、準(zhǔn)確率低的問題,基于灰狼算法(Grey Wolf Optimizer, GWO)、變分模態(tài)分解(Variational Mode Decomposition, VMD)、長(zhǎng)短期記憶模型(Long Short Term Memory, LSTM)提出一種單參數(shù)時(shí)序預(yù)測(cè)方法。首先,使用改進(jìn)適應(yīng)度函數(shù)的GWO對(duì)VMD的分解層數(shù)和懲罰系數(shù)進(jìn)行尋優(yōu);其次... (共9頁(yè))