围岩流变模型参数识别的启发式-支持向量机方法

发布时间：2018-06-14 08:27

本文选题：隧道 + 围岩　；参考：《现代隧道技术》2016年04期

【摘要】：针对传统反分析算法在精度与效率方面的缺陷,文章引入支持向量机、启发式智能优化算法及三维数值模型对围岩流变参数进行反演。其具体过程为:采用均匀设计方法,利用三维数值模型构造训练样本集,通过启发式智能算法搜索最佳的支持向量机模型参数,建立岩体力学参数与岩体位移之间的非线性支持向量模型;对任意一组给定的岩石力学参数,利用支持向量机的外推能力代替数值计算软件的正向计算,通过启发式算法对力学反演参数进行调整,即可迭代求出岩体力学参数。将该方法运用于广甘高速公路杜家山隧道工程流变参数的反演中,结果表明,该反演方法合理可靠,可有效地指导隧道工程参数设计和施工稳定控制。
[Abstract]:Aiming at the shortcomings of traditional inverse analysis algorithm in precision and efficiency, support vector machine (SVM), heuristic intelligent optimization algorithm and 3D numerical model are introduced to inverse the rheological parameters of surrounding rock. The concrete process is as follows: using uniform design method, using three-dimensional numerical model to construct training sample set, and searching the best support vector machine model parameters by heuristic intelligent algorithm. The nonlinear support vector model between rock mass mechanics parameters and rock mass displacement is established, and for any given rock mechanics parameters, the extrapolation ability of support vector machine is used to replace the forward calculation of numerical calculation software. By adjusting the parameters of mechanical inversion by heuristic algorithm, the parameters of rock mass mechanics can be obtained iteratively. The method is applied to the inversion of rheological parameters of Dujiashan Tunnel on Guanggan Expressway. The results show that the inversion method is reasonable and reliable and can effectively guide the design of tunnel engineering parameters and construction stability control.
【作者单位】：西南交通大学交通隧道工程教育部重点实验室;
【基金】：国家973计划(2010CB732105) 煤炭联合基金重点项目(U1361210) 国家科技支撑计划课题(2013BAB10B04)
【分类号】：U451.2

【相似文献】