锦屏一级水电站地下厂房时空智能反馈分析的实现与应用

Implementation and development of AI back analysis for the underground powerhouse of Jinping-1 hydropower project

本文首先对锦屏一级水电站地下厂房洞室群规模巨大、地质条件复杂、地应力高等不利情况作了简单介绍;由于几乎没有可供借鉴的工程经验,给围岩稳定评价以及加固处理措施的确定提出了极大挑战。系统开展对地下厂房洞室群施工监测与反馈分析研究,是解决上述问题的重要手段。然后简要介绍锦屏一级地下厂房时空双系列反馈分析系统的结构、组成和实现方法。对本项研究提出的初始地应力场分区模拟技术、收敛-增量混合位移目标函数,以及动态松动区模型等新功能开发作较详细介绍。最后介绍反馈分析在锦屏一级水电站地下厂房施工中,对洞室群的开挖顺序、支护参数等进行优化,对洞室群的围岩稳定性进行合理评价和预测预报所起的作用。

Jinping-1 hydropower station is a key underground hydroelectric power project in the southwest of China.The powerhouse of this project is very large and under complicated poor geological conditions with a maximum crust stress of 35MPa,which brought many difficulties and a great challenge to the design and construction and high uncertainties of the caverns safety.A back-analysis of the powerhouse construction process is quite necessary.In this paper,the layout,principle,structure and implementation of the back-analysis system is briefly described,and some new developments are demonstrated in details,such as the zonal modeling of crust stress field,the convergence-increment hybrid target function and variable loosening zonal model.Finally this paper shows how this system works out an optimum and adjusts the excavation and cavern supports,and the roles of back analysis in making prediction and warning to ensure the power house safety.