多類算法融合的基坑沉降組合預(yù)測(cè)分析

打開文本圖片集

中圖分類號(hào)：U459 文獻(xiàn)標(biāo)志碼：A 文章編號(hào)：1005-8249（2025）03-0128-06

DOI：10.19860/j.cnki.issn1005-8249.2025.03.023

Abstract：Toachieve high-precision predictionoffoundationpitsettlement，basedontheon-sitesettement monitoring

resultsof the foundation pit，thepolarsymmetricmodedecompositionalgorithmis firstusedtodecomposethesetlementdata，

obtaining several modal componentsand trend components.Then，thegrey wolf algorithmandgatedrecurrent unit neural

networkareused toconstructacombined prediction model，and thismodel isused topredictthedeformationof each modal

componentandtrendcomponent，inordertoobtainthecombined predictionvalueoffoundationpitsetlement.Theanalysis

results indicatethatcommonsetlement monitoring itemsduringfoundationpitconstructionincludesurfacesettlement，pittop

setlement，andbuildingsetlement；According tomonitoringdataanalysis，theremaining deformationspaceofthepittop setlementisrelativelythelargest，followedbybuildingsettlementandsurfacesetlement；Overall，theremainingdeforation spaceforthethree typesofsetlementanddeformationprojectsisstillrelativelyoptimistic，withonlyasmall numberof monitoring points having limitedremaining deformationspaceincertainareas；Intheprocessofdataprocessing，thepole symmetric mode decompositionalgorithmhasstrongdataprocessingability，anditsabilityissignficantlybetterthanthedata procesing efectofwaveletdenoisingand mode decompositionmethods.According toGWO-GRU prediction，therelativeerror mean of the prediction results of this model in three types of setlement projects is around 2% ，which has high prediction accuracy.Te predictionresultsshowthatthedevelopmenttrendofthethree typesofsetlementdeformationisrelatively consistent inthefuture，showingasmallrate increase trendandobviousconvergencecharacteristics.Thisindirectlyverifiesthe goodoperational efectoffoundationpitsupportmeasures．Throughresearch，itcanprovidetechnicalreferencesforsimilar projects and has certain practical significance.

Key words：excavation pit；setlement deformation；data decomposition；grey wolf algorithm；combination prediction

0 引言

近年，隨著城市化進(jìn)程的加快，基坑工程數(shù)量越來(lái)越多，基坑開挖會(huì)引發(fā)基坑周邊土體變形，基坑沉降變形屬基坑施工過(guò)程中的必測(cè)項(xiàng)目，姜偉玲等[1]、仇安兵[2]認(rèn)為開展基坑沉降變形的相關(guān)研究具有較強(qiáng)的現(xiàn)實(shí)意義。（剩余6234字）