摘要
以三峡工程为背景 ,在试验、监测、前期分析结果基础上 ,建立了地质与施工的概化模型。以基本力学参数和初始应力场为基本变量 ,按正交设计和均匀设计方法进行计算方案的组合 ,采用考虑开挖卸荷效应的显式有限差分法完成了船闸高边坡的开挖模拟计算 ,并通过神经网络和遗传算法进行优化反演。反演位移计算值与实测值较为吻合 ,表明反演得到的力学参数和地应力场基本合理。
Based on the results of test, monitoring and past analyses of the high rock slope in the Three Gorges Project, a generalized model for geology and construction process is established. By combination of computing schemes of orthogonal and even design methods with basic mechanical parameters and initial stress field as basic variable, the finite difference method is used to complete the excavation simulated computation of shiplock highslope considering the effect of excavation unloading, and an optimized back analysis is conducted by using the neural network and genetic algorithm. The basic agreement of result of the back analysis with the testing data indicates that the mechanical parameters and initial stress field obtained from the proposed back analysis are reasonable.
出处
《岩石力学与工程学报》
EI
CAS
CSCD
北大核心
2000年第z1期987-993,共7页
Chinese Journal of Rock Mechanics and Engineering
关键词
三峡工程
边坡
开挖
卸荷
位移反分析
神经网络
遗传算法
the Three Gorges Project, slope, excavation, unload, displacement back analysis, neural network, genetic algorithm
