一种指数平滑分数阶离散灰色模型及其应用

打开文本图片集

中图分类号：TP18；N941.5 文献标识码：A 文章编号：2096-4706（2025）22-0137-08

An Exponential Smoothing Fractional-order Discrete Grey Model and Its Applications

WU Xinyun （ChinaWestNormalUniversity，Nanchong637oo9，China）

Abstract：Because the total energyconsumption data shows a large jump and has significant nonlinear characteristics， the existingtimeseriesprediction，artificialinteligencepredictionandgreymodelpredictionarenotidealfordatasimulation with suchcharacteristics.Inorder tobetersimulateandpredictthedevelopment trendofenergyconsumption，anexpoential smoothing fractional-orderDiscrete GreyModelisproposedinthispaper.Theexponentialsmoothing termisdynamically adjusted，andtheParticle SwarmOptimizationalgorithmisusedtorealizetheoptimalsearchofparameters.Theapplication scopeof thetraditional greymodelisextended toapproximate non-homogeneousexponentialsequence，exponentiallinear combinationsquenceandexponentialnonlinearcombinationsequence.Ten，theeffectivenessofthenewmodel isverifed by the numerical simulationoffourdifferent characteristicsequences andtheexampleof wind power generationin China.Finaly the new modelisusedtopredictthe totalenergyconsumptionin China.Theresultsshowthatthefitingand predictionaccuracy of the new model is higher.

Keywords： exponential smoothing; fractional-order accumulation; Discrete Grey Model; Particle Swarm Optimization algorithm

0 引言

随着生产力提升与科技进步，中国经济水平不断提高，城市化飞速发展，人民对能源的需求居高不下[。（剩余10843字）