基于DE-SVM三维隧道智能反馈分析研究

为了解决隧道围岩参数不确定性的问题,本文引入支持向量机、差异进化算法和三维数值模拟建立了隧道围岩参数快速反馈分析方法。介绍了方法的原理、步骤和实现过程,开发了隧道监测在线反馈分析软件平台。该方法克服了传统优化方法因为三维数值计算耗时长而无法应用以及容易限于局部最优解的不足,可以与多通道自动监测仪器实时配合使用。将该方法应用于兰州白家坡隧道K8+750到K8+1000桩号之间区域监测分析,搜索得到了围岩力学参数,精度令人满意,本方法提供了隧道围岩三维快速反馈分析的有效途径。

Study on 3D Intelligent Back Analysis of Tunnel Based on DE-SVM

In order to solve the problems of uncertainty of tunnel surrounding rock parameters, this paper introduced support vector machine, difference evolution arithmetic and three-dimensional numerical simulation, so as to construct quick feedback analysis method of surrounding rock parameters for tunnel construction. The theory, steps and realizing process of the method are introduced, and the on line feedback analysis platform is developed. The method has overcome the shortages of consuming too much time and easily being limited in local optimum solution, and can mat with the automatic monitoring instruments in real time. The method is used to analyze the monitoring data of the region between marks K8+750 and K8+1000, and gets the identified surrounding rock parameters with satisfactory precision. The method provides an instrument for three-dimensional quick analysis on tunnel.