【摘要】 基于正交试验设计和FLAC3D建立的学习样本以及测试样本,通过工程现场获取的围岩位移信息,用神经网络建立待反演参数与围岩位移之间潜在的映射关系。研究结果表明:利用该神经网络的仿真预测功能,结合遗传算法搜索反演参数的最优解,从而实现位移反分析;可将反演结果反馈于隧道支护结构的设计,实现隧道的信息化施工与设计。更多还原

【Abstract】 The learning samples and test samples were built based on orthogonal experimental design and FLAC3D numerical simulation,using the project site for the surrounding rock displacement information.The potential mapping between parameters and surrounding rock displacement was established using neural network.The results show that the optimal solution of inversion parameters can be derived to achieve the displacement back analysis combined with the genetic algorithm and prediction function of the nerve network.Then inversion results can be the feedback for the design of tunnel support structure to achieve the tunnel construction and design of information technology.更多还原