大型地下洞室群施工期围岩力学参数实时动态反演

综合考虑大型地下洞室群施工现场的开挖进度信息、支护进度信息及新出露的地质信息,建立能够实时更新且反映工程实际进度的三维全尺度数值仿真模型,结合均匀设计和人工神经网络技术,基于C#.NET+Python混合编程技术对通用数值软件ABAQUS进行二次开发,提出大型地下洞室群施工期围岩力学参数实时动态反演分析方法。以黄登水电站为例,对其地下洞室群施工期围岩力学参数进行实时反演。结果表明:各施工分期测点计算位移与实测位移随施工期的曲线变化规律一致,实测值与计算值吻合度较高,第五期的平均误差仅为5.5%,得到较好的反演结果,验证了此法的合理性和可操作性。

Real-time dynamic inversion of surrounding rock mechanical parameters during construction of large-scale underground cavern group

Based on information regarding the excavation progress, support progress, and newly outcropped geology at the construction site of a large-scale underground cavern group, a three-dimensional numerical simulation model, which can be updated in real time and reflect the practical process of construction, was established. Using the techniques of uniform design and artificial neural networks and the mixed programming technique of C#. NET+Python, the secondary development of ABAQUS proceeded, and a method of real-time dynamic inversion of surrounding rock mechanical parameters was developed for a large-scale underground cavern group during the construction period. Using the Huangdeng Hydropower Station as an example, real-time dynamic inversion of surrounding rock mechanical parameters during the construction of a large-scale underground cavern group was conducted. The results show that the variations of calculated displacement and measured displacement of measuring points with construction time showed the same pattern of change in different construction periods and were in good agreement. The average error was only 5. 5% in the fifth construction period. This good inversion result verifies the rationality and maneuverability of this method.