引水渠道边坡材料参数智能反演及稳定性动态分析

Intelligent Inversion for Material Parameters of Diversion Canal Slopes and Dynamic Stability Analysis

利用有限元方法对引水渠道边坡进行数值模拟计算,将位移计算值和土体材料力学参数值代入径向基函数神经网络(RBF神经网络)的输入层和输出层进行训练,得到多层土体材料参数与位移之间的非线性关系,建立材料参数智能反演模型,根据不同时间的位移监测值反演得到了对应土体的力学参数,将反演得到的参数进行有限元正分析,验证反演模型的精确度。结合反演结果以有限元强度折减法计算引水渠道边坡不同时间的稳定性安全系数,研究边坡稳定性动态变化规律。

The finite element method was used to simulate diversion canal slopes. The calculated displacement values and soil material mechanical parameter values were substituted into the input layer and output layer of the RBF neural network for training. The nonlinear relation between material parameters of multi-layer soils and displacement was obtained, and an intelligent inversion model of material parameters was established. According to monitoring displacement values at different times, the corresponding soil mechanical parameters were obtained by inversion analysis, and the parameters were used for finite element analysis to verify the accuracy of the inversion model. Combined with the inversion results, the safety factor of stability for diversion canal slopes at any time was calculated by the strength reduction method of finite elements, and the dynamic change law of slope stability was studied.