文档名：SSAMLP模型在岩质边坡稳定性预测中的应用
摘要：岩质边坡的力学参数量化及稳定性分析对岩质边坡灾害的防治具有重要意义.Hoek-Brown(H-B)准则是一种用于确定岩体力学参数的经典方法,能反映出边坡岩体变形和位移的非线性破坏特征.在此基础上,首先,提出一种麻雀搜索算法(SparrowSearchAlgorithm,SSA)改进多层感知器(Multi-LayerPerceptron,MLP)的神经网络模型,并用于边坡稳定性预测、指标敏感性分析及参数反演.其次,将收集的1085组岩质边坡的几何参数和H-B准则参数等作为输入变量,极限平衡理论Bishop法求解的安全系数作为输出变量,对SSA-MLP模型进行训练学习和性能评估.最后,将该模型运用于25个边坡实例,验证模型的有效性.结果显示,该模型收敛速度快、精度高,为边坡稳定性分析和参数量化提供了一种新思路.
Abstract：Inthispaper,anewmethodofrockslopestabilitypredictionandparameterinversionisproposedbycombiningtherockstrengthcriterion,swarmintelligencealgorithm,andartificialneuralnetworktheory.Hoek-Brown(H-B)strengthcriterionisaclassicalmethodusedtodeterminethemechanicalparametersofrockmass,whichcanreflectthenonlinearfailurecharacteristicsofrockmassdeformationanddisplacement,andhasgoodapplicabilityinthestabilityanalysisofrockslopes.Therefore,basedontheparametersoftheH-Bcriterion,theindexsystemofrockslopestabilitypredictionisconstructed.Then,accordingtothecomplexnonlinearcharacteristicsofslopeengineeringstabilityproblem,apredictionmodelofMulti-LayerPerceptron(MLP)neuralnetworkimprovedbySparrowSearchAlgorithm(SSA)isproposedforthepredictionofsafetyfactorandparameterinversionofrockslope.Theparameteroptimizationfunctionofthesparrowsearchalgorithmisusedtooptimizetheconnectionweightsandthresholdsofthemulti-layerperceptronneuralnetwork,andtheSSA-MLPneuralnetworkmodelwithhigherpredictionaccuracyisobtained.Besides,accordingtothegeometricparametersandH-Bcriterionparametersof1085setsofrockslopescollectedasinputvariables,andtheslopesafetyfactorsolvedbytheBishopmethodbasedonthelimitequilibriumtheoryasoutputvariables,thetraininglearningandperformanceevaluationoftheSSA-MLPmodelarecarriedout,andcomparedwithothernetworkmodels.Itisanalyzedthatthemodelhadhighfeasibility.Inaddition,thesensitivityanalysisbetweenthesafetyfactorandthecharacteristicindexiscarriedoutbyusingtheSSA-MLPmodelandtheKendallcorrelationcoefficient.Finally,themodelisappliedto25slopecases,andtheparameterinversionofthedisturbancecoefficient(D)andthegeologicalstrengthindex(GSI)inengineeringcasesiscarriedouttofurtherverifytheeffectivenessofthemodelTheresultsshowthatthemodelhasfastconvergencespeedandhighprecision,whichprovidesanewideaforslopestabilityanalysisandparameterquantification.
作者：侯克鹏 包广拓 孙华芬Author：HOUKepeng BAOGuangtuo SUNHuafen
作者单位：昆明理工大学国土资源工程学院,昆明650093;云南省中-德蓝色矿山与特殊地下空间开发利用重点实验室,昆明650093
刊名：安全与环境学报 ISTICPKU
Journal：JournalofSafetyandEnvironment
年，卷(期)：2024, 24(5)
分类号：X43
关键词：安全工程 边坡稳定性 Hoek-Brown准则 多层感知器(MLP)神经网络 麻雀搜索算法 参数反演
Keywords：safetyengineering slopestability Hoek-Browncriterion Multi-LayerPerceptron(MLP)neuralnetwork SparrowSearchAlgorithm(SSA) parameterinversion
机标分类号：TU457TD824.7P642
在线出版日期：2024年6月12日
基金项目：云南省科技厅项目SSA-MLP模型在岩质边坡稳定性预测中的应用[
期刊论文] 安全与环境学报--2024, 24(5)侯克鹏 包广拓 孙华芬岩质边坡的力学参数量化及稳定性分析对岩质边坡灾害的防治具有重要意义.Hoek-Brown(H-B)准则是一种用于确定岩体力学参数的经典方法,能反映出边坡岩体变形和位移的非线性破坏特征.在此基础上,首先,提出一种麻雀搜索算法...参考文献和引证文献
参考文献
引证文献
本文读者也读过
相似文献
相关博文
SSA-MLP模型在岩质边坡稳定性预测中的应用 Application of SSA-MLP model to stability prediction of rock slope
SSA-MLP模型在岩质边坡稳定性预测中的应用.pdf
- 文件大小:
- 6.18 MB
- 下载次数:
- 60
-
高速下载
|
