基于三维数值模拟的地下大型洞室锚固参数智能优化  

INTELLIGENT OPTIMIZATION OF ANCHORING PARAMETERS FOR LARGE UNDERGROUND HOUSES BASED ON THREE DIMENSIONAL NUMERICAL SIMULATION

作　　者：姜谙男[1] 冯夏庭[2] 刘建[2] 刘红亮[1] 

机构地区：[1]东北大学资源与土木工程学院,沈阳110006 [2]中科院武汉岩土力学研究所岩土力学重点实验室,武汉430071

出　　处：《岩石力学与工程学报》2004年第10期1700-1705,共6页Chinese Journal of Rock Mechanics and Engineering

基　　金：国家重点基础研究发展规划(973)项目(2002CB412708);国家自然科学基金重点项目(59939190)资助课题。

摘　　要：地下大型洞室锚固参数优化具有以下特点:稳定性目标和经济性目标是矛盾的,洞室的稳定性分析非常复杂,评价指标要求全面反映锚固效果,锚固参数组合的方案数量多,优化计算量大。针对这些特点,在采用FLAC3D进行较大规模的三维数值模拟的基础上,综合应用遗传算法、神经网络建立了进化神经网络演化有限差分方法,确定了约束条件及优化的指标,给出了优化的方法和步骤。利用该方法,对水布垭地下厂房3～6层施工的上、下部锚索、锚杆、喷层等锚固参数进行了优化。工程实例表明所提方法是可行的。The optimization of anchoring parameters for large underground houses is of some characteristics. For example, the object of stability is incompatible with the object of economics, so that the stability analysis of large underground houses is very complicated, and the assessment indexes are required to denote anchoring effect roundly. Further more a lot of schemes with different anchoring parameters will occur and a great deal of calculation is needed. Aiming at these characteristics, large scale numerical simulation is made with FLAC3D, combined with genetic algorithm and NN, and a new evolutionary NN limited difference method is proposed to optimize the anchoring parameters for large underground houses. Constrained conditions and optimization goals are constituted, and the optimization process is proposed. Using above method, the anchoring parameters of Shuibuya underground houses between the third and sixth construction step are optimized for the underground houses of Shuibuya Hydro-Power Station, which proves that the proposed method is feasible.

关 键 词：地下洞室 三维数值模拟 锚固 智能优化 神经网络 

分 类 号：TU457[建筑科学—岩土工程]