基于BP-PSO的倾斜面层加筋挡土墙筋材内力分析  

Internal force analysis of inclined reinforced retaining wall based on BP-PSO

作　　者：刘中南 肖豪 汪磊 LIU Zhongnan;XIAO Hao;WANG Lei(School of Software Engineering(Nanchang),Jiangxi University of Science and Technology,Nanchang 330013,China;Hunan Chemical and Geological Engineering Survey Institute Co.Ltd.,Changsha 410004,China)

机构地区：[1]江西理工大学软件工程学院(南昌),江西南昌330013 [2]湖南化工地质工程勘察院有限责任公司,湖南长沙410004

出　　处：《贵州科学》2025年第2期86-91,共6页Guizhou Science

基　　金：江西省教育厅科学技术研究项目(GJJ200865);高层次人才科研启动项目(jxncbs19002);江西理工大学校级研究生创新专项资金项目(XY2023-S239)。

摘　　要：返包式面层倾斜加筋挡土墙作为一种轻型支挡结构,对其稳定性分析在工程上是非常重要的,通常体现在对其土压力分析、倾斜面层稳定性分析、筋材内力分析、抗滑稳定性分析等。其中筋材内力计算是复杂的力学问题,它涉及筋-土相互作用、土-面板相互作用以及筋材-面板相互作用等[1]。基于筋材内力计算的复杂性,提出了一种基于PSO算法优化的BP神经网络预测方法,以大量模拟数据和实测数据构建训练样本,以BP神经网络建立筋材最大拉力的高精度非线性映射关系,以PSO算法逆向推求进而确定一个最优的映射关系。结果比NCMA规范计算[2-5]和K刚度法[6-7]等计算方法所得的计算结果误差更小,能够满足工程实际应用。As a kind of light support structure,the stability analysis of inclined reinforced retaining wall is very important in engineering,which usually includes soil pressure analysis,inclined surface stability analysis,reinforcement internal force analysis,anti-slip stability analysis,etc.The calculation of the reinforcement internal force is a complex mechanical problem,which involves tendon-soil interaction,earth-panel interaction,and tendon-panel interaction,etc.Aiming at the complexity of internal force calculation,this paper proposes a BP neural network prediction method based on PSO algorithm.By constructing training samples with a large number of simulated data and measured data,the high-precision nonlinear mapping relationship of the maximum tensile force is established by BP neural network,and the optimal mapping relationship is determined by reversing PSO algorithm.Compared with the results of NCMA standard calculation and K stiffness method,the calculation error of the proposed method is smaller,which can meet the requirements of practical engineering application.

关 键 词：倾斜加筋挡土墙 BP神经网络 PSO粒子群算法 最大筋材拉力 

分 类 号：TU41[建筑科学—岩土工程]