西藏邦铺矿区地应力测量及三维地应力反演分析  

Inverse Analysis on Three-dimensional Stress and In-situ Stress Measurement of Bangpu Mining Area in Tibet

作　　者：刘飞[1] 胡斌[1] 宋丹[2] 甘露[3] 欧光劲 

机构地区：[1]中国地质大学(武汉)工程学院,湖北武汉430074 [2]西安科技大学地质与环境学院,陕西西安710054 [3]成都建筑材料工业设计研究院有限公司,四川成都610011 [4]四川金顶(集团)股份有限公司,四川峨眉山市614224

出　　处：《矿业研究与开发》2015年第3期47-51,共5页Mining Research and Development

基　　金：国家自然科学基金面上项目(41172281);国家重点基础研究发展计划(973计划)项目(2011CB710604);中央高校基本科研业务费摇篮人才计划项目(CUGL100413);中央高校基本科研业务费特色学科团队项目(CUG090104)

摘　　要：西藏邦铺矿区处于青藏高原腹地,初始地应力场复杂。根据矿区地应力实测资料和矿区地质构造条件分析,初步确定研究区域内初始地应力场特征。运用神经网络建立应力值与位移边界条件的非线性映射关系,通过粒子群算法从全局搜索最优的反演参数。将地应力反演值与实测值对比分析,两者在量值上比较接近,表明基于粒子群优化算法的初始地应力场反演方法能够准确有效的获得岩体的初始地应力场。通过反演分析,得到了西藏邦铺矿区合理的初始地应力场的分布规律。The initial stress field of Bangpu mining area located in central Tibetan Plateau was complex.Based on the analysis of in-situ stress measured data and geological structure conditions,the characteristics of the initial stress field were preliminarily determined.The nonlinear mapping relationship between the stress and the displacement boundary condition was established by neural network,and the optimal inversion parameter was searched from global by particle swarm optimization.The inversion values of in-situ stress were similar to the measured values by comparative analysis,which indicated that the in-situ stress field of rock mass could be accurately and effectively obtained by inversion method based on the particle swarm optimization.Finally,the reasonable distribution of initial stress field in Bangpu mining area was received through inversion analysis.

关 键 词：初始地应力场 非线性反演 神经网络 粒子群算法 

分 类 号：TD311[矿业工程—矿井建设]