基于径向基神经网络的边坡可靠度分析  

Reliability Analysis of Slope Based on Radial Basis Function Neural Network

作　　者：夏舒 XIA Shu(Hunan Sangzhi County Traffic Construction Quality and Safety Supervision Institute,Zhangjiajie,Hunan 427200,China)

机构地区：[1]湖南省桑植县交通建设质量安全监督所,湖南张家界427200

出　　处：《黑龙江交通科技》2024年第8期32-35,共4页Communications Science and Technology Heilongjiang

摘　　要：为实现边坡稳定系数的快速求解,提出了一种基于RBF神经网络的边坡可靠度求解方法。利用RBF神经网络的非线性映射能力,对边坡的功能函数进行高精度的拟合。通过构建合适的神经网络模型,并输入边坡的相关参数,有效地模拟边坡的稳定状态,并为其可靠度分析提供基础。采用改进策略改进标准的遗传算法,进一步提升了算法的寻优能力,使其更适用于边坡可靠度的求解问题。以湖南宁道高速边坡为工程背景,验证了该方法的可行性。结果表明,RBF神经网络与有限元计算结果之间实现了较为精确的拟合,RBF神经网络模型能够准确地反映边坡的稳定状态。同时,改进后的遗传算法在寻优过程中表现出了更强的能力,能够更快速地找到边坡可靠度的最优解。采用RBF神经网络方法计算得到的边坡稳定系数为1.19,相比于蒙特卡洛法小17.9%,计算结果更偏保守。In order to achieve the rapid solution of slope stability coefficient,a slope reliability solution method based on RBF neural network is proposed.Using the powerful nonlinear mapping ability of RBF neural network,the functional function of slope is fitted with high precision.By constructing an appropriate neural network model and inputting the relevant parameters of the slope,the stability of the slope can be effectively simulated and the basis for its reliability analysis can be provided.The improved strategy is adopted to improve the standard genetic algorithm,which further improves the optimization ability of the algorithm and makes it more suitable for solving the problem of slope reliability.Taking the slope of Ningdao Expressway in Hunan Province as the engineering background,the feasibility of this method is verified.The results show that the RBF neural network model can accurately reflect the stability of the slope.At the same time,the improved genetic algorithm shows stronger ability in the optimization process,and can find the optimal solution of slope reliability more quickly.The slope stability coefficient calculated by RBF neural network method is 1.19,it is 17.9%smaller,and the calculation results are more conservative.

关 键 词：边坡稳定性 功能函数 RBF神经网络 改进遗传算法 

分 类 号：U416.1[交通运输工程—道路与铁道工程]