改进秃鹰算法优化ELM的短期电力负荷预测研究

打开文本图片集

DOI：10.16652/j.issn.1004-373x.2025.18.017引用格式：，.改进秃鹰算法优化ELM的短期电力负荷预测研究[J].现代电子技术，2025，48（18）：107-113.

关键词：短期电力负荷预测；改进秃鹰搜索算法；极限学习机；Piecewise混沌映射；莱维飞行策略；动态权重因子中图分类号：TN86-34；TM715 文献标识码：A 文章编号：1004-373X（2025）18-0107-07

Researchonshort-term power loadforecastingbasedon ELM optimizedbyIBESalgorithm

ZHANG Xu， WANG Fanrong

（SchoolofElectricalandElectronicEngineering，HubeiUniversityofTechnology，Wuhan43Oo74，China）

Abstract：Inalusiontotheincreasinglycomplexpowerenvironmentatthepresentstage，andtheshort-termpowerload forecasting methodshavelowforecastingaccuracyandslowconvergencespeed，ashort-termpowerloadforecasting modelbased onextremelearning machine（ELM）optimizedbyimprovedbaldeaglesearch （IBES）algorithmisproposed.Theoriginalbald eaglesearch（BES）algorithmispronetofallintolocaloptimumwhendeterminingtheconnectionweightsandimplicitlayer thresholdsintheELM，andtheconvergencespeedisslow，whichresultsintheporpredictionaccuracy.Onthisbasis，the Piecewisechaoticmapping isusedtoinitilizethebaldeaglepopulationandincreasethediversity.TheLevyflightstrategyis introduced toexpandthesearchrangeof thepopulation，sothatitcanjumpoutof thelocaloptimumintime.Thedynamic weightingfactorisintroducedtoimprovethelocalsearchabilityofbaldeagle.TheIBESalgorithmisusedtooptimizethetwo stochasticparametersofELM，soastoestablishtheBES-ELMshort-termpowerloadforecastingmodel.Teforecastinganalsis is conductedbycombining withtheactualpowerloaddataofaregion.Theresultsshowthat，incomparison withELM，BESELM，PSO-ELM，andDBO-ELM，the improved model has animprovement in forecasting accuracyand convergence speed.

Keywords：short-termpowerloadforecasting;IBES；ELM;Piecewisechaoticmapping;Levyflightstrategy；dynamic weighting factor

0 引言

电力负荷预测是指综合利用历史负荷数据、气象数据、经济数据等相关信息，预测未来一段时间内电力负荷的变化趋势和具体数值，为电力系统运行和规划提供依据。（剩余7228字）