基于BP神经网络的隧道围岩力学参数反分析方法

摘要

In order to solve the problems of the model complex and slow speed in problem-solving for all conventional displacement back-analysis, the program of BP neural network compiled by the M language of MATLAB was used for back-analysis of displacements.Aiming at the disadvantage of slow-footed convergence of the traditional BP neural network, optimization arithmetic and normalization method were used to quicken the network training rate.FLAC-3D program is used as a tool of forward process in the simulation of tunnel excavating and combined with the program of BP neural networks and orthogonal design, artificial neural network that can be used for back-analysis of displacements was establish.Furthemoro, this approach was used to inverse the mechanical parameters of a certain tunnel surrounding rock.The inversed results show that the model we made is simple, easy to solve the problem, and can get accurate solutions, and then could be used widely in underground engineering for back-analysis of displacements.Artificial neural network has good nonlinear information storage capacity and adaptability.It is possible to make use of the BP algorithm to inverse the mechanical parameters of tunnel surrounding rock.And it can provide reference for finding parameters in engineering design and calculation.The application of back-analysis method is important in the stability analysis of the tunnel surrounding rock and in informational design.%目的 为解决各种传统位移反分析方法的反分析模型复杂、求解难度大等问题,基于MATLAB的二次开发语言M语言,编写了用于位移反分析的BP神经网络源程序.针对传统BP网络收敛速度慢的缺点,采用优化算法及归一化方法来加快网络的训练速度.方法 隧道开挖模拟采用FLAC-3D数值方法作为正演工具,结合正交设计法和BP神经网络等程序,建立了用于位移反分析人工神经网络方法,并应用该方法对某隧道围岩力学参数进行了反演.结果 反演结果表明,所建立的位移反分析的人工神经网络方法具有模型简单、求解快捷等优点,且其精度亦能达到工程应用要求,因而可以在工程实际中推广应用.结论 人工神经网络有着良好的非线性信息存储能力和自适应性,利用人工神经网络中的BP算法反演隧道围岩力学参数,在实际应用中是完全可行的,可以为工程所需的计算参数提供参考,对隧道围岩稳定性评价及信息化设计具用实际意义.