基于遗传算法的BP神经网络隧道施工参数正反演分析与应用

摘要

相对于监测数据采集的高效性,隧道施工中对现场突发状况缺乏高效的应对措施.本文结合工程实例,采用正交表及对应三维数值计算模型,得到隧道施工参数与对应隧道变形的样本集,应用基于遗传算法的BP神经网络M atlab程序,通过对施工参数进行正演分析,实现相对高效的施工反馈;在实测数据基础上,通过进一步的反演分析,可优化施工参数,实现施工工艺的经济优选.工程应用结果表明,该方法的分析结果能够满足工程施工精度要求,有效提高施工过程中突发状况的应对效率,同时也为设计中参数的优化选择提供参考,为建立隧道工程施工的高效反馈机制提供新方法和新思路.%Compared with the high efficiency of monitoring data collection during tunnel construction ,effective counter measures are not available to deal with emergencies on the construction site .In this paper ,combined with practical work ,a sample set of the tunnel construction parameters and the corresponding tunnel deforma-tion was established through an orthogonal layout and corresponding numerical 3D calculation model .The Mat-lab program of a genetic algorithm based BP neural network was applied to forward analysis of the tunnel con-struction parameters in order to achieve relatively efficient construction feedback .Based on the measured data , the construction parameters can be optimized through further back analysis to realize economic optimization of construction technology .The results showed that this method satisfied the precision requirements of construc-tion ,effectively improving the efficiency to deal with emergencies in the construction process .Furthermore ,it can be applied to the optimization of parameters during design ,offering a new way in establishing high-efficien-cy tunnel construction feedback mechanism .