基于粒子群算法优化的灰色预测模型路基沉降预测分析

打开文本图片集

中图分类号：U213.1 文献标志码：A 文章编号：2095-2945（2025）21-0030-05

Abstract：Thispaperfocusesonthepredictionofsubgradesetlementofhigh-speedrailway.Inviewofthekeyimpactof subgradesetlementonthestabilityandsmoothnessoftheline，thegreypredictionmodelisselectedaftercomparingvarious predictionmethods.TheprincipleofGM（1，1）modelandtheoptimizationprocessofparticleswarmoptimization（PSO）areitroduced in detail. The 69～339 day settlement observation data of two sections K417+523 and K417+573 in the first work area of ZH sectionofJinan WestRailwayStationofBeijingShanghai highspeedrailwayaretakenasexamples tocarryoutthecaseanalysis. TheresultsshowthatthepredictionefectofPSO-GM（1，1）modelisbeterthanthatofGM（1，1）modelandtheaveragefiting erorsathetwosectionsare3.8%and3.9%respectivelyTheresidualerortestandclassratiodeviationtestshowthatithas higheraccuracyandbeterstability.Thisresearchprovidesanewideaforgreypredictionmodeltodealwithcumulativeors， and proves that PSO-GM（1，1）model has high reliabilityandapplication value in predicting subgrade settlement.

Keywords：high-speedrailway;subgradesettlement; GM（1，1）model;settlementprediction;particleswarmalgorithm

高速铁路以其迅捷的运行速度，对线路的稳定性和平顺性提出了极高的要求。（剩余5326字）