针对隧道围岩参数取值的不确定性,以现场量测得到的位移信息量为基础,利用结构分析有限元的正演分析法提取实验样本,构造遗传算法全结构优化的径向基概率神经网络(RBPNN)模型,反演围岩力学参数.结果表明,利用遗传优化径向基概率神经网络反演围岩力学参数能够达到工程应用要求,为隧道施工的理论及其工程化方法提供了有效的途径. According to the uncertainty of tunnel surrounding rock parameters, based on the displacement information measured in the field, the experimental samples were extracted by the forward analysis of structural analysis finite element method, and the full-structure optimized RBF neural network Network (RBPNN) model, the mechanical parameters of surrounding rock are retrieved.The results show that the application of genetic algorithm to inverse the mechanical parameters of surrounding rock can achieve the requirements of engineering application, which provides an effective method for tunnel construction theory and its engineering method Way.