基于PSO-BP神经网络的热电厂负荷预测策略研究

打开文本图片集

中图分类号：TP311 文献标志码：A 文章编号：2095-2945（2026）01-0032-04

Abstract：Atpresent，theeficientuseofenergyandgreendevelopmenthaveatractedwidespreadatentionfromscholars. Basedonalargeamountofhistoricaldatageneratedbytheenergymanagementsystemof athermal power plant，this paper usesbigdataanalysismethodtocalculatethecoelationcoeffientbetweenthedatatojudgethecorrelationstatusbetweenthe data.APSO-BPneuralnetworkmodelisestablishedtopredicttheheatloadofathermalpowerplantinthenext24hours，in ordertobetterprovideproduction，operationmanagementanddecision-makingservicesforthethermalpowerplant.ThePSO-BP neuralnetworkmodelisproducedbyfusingparticleswarmalgorithmandBPalgorithm.Itnotonlyimprovestheprediction accuracyofBPneuralnetwork，butalsoefectivelysolvestheproblemofslowleamingspeedofBPneuralnetworkalgorithm， easyto fall into local minima，poor stability，etc.

KeyWords： big data analysis;thermal characteristics;prediction model; PSO-BP neural network; prediction accuracy

能源是否能够做到高效利用和绿色低碳发展关系到了人类社会的未来和命运。（剩余4786字）