边坡临界滑面搜索的连续转移蚁群算法  

A continuous transfer based ant colony algorithm for searching critical slip surface of slope

作　　者：刘江[1] 

机构地区：[1]中国电力工程顾问集团 中南电力设计院,湖北武汉430071

出　　处：《交通科学与工程》2013年第3期58-62,共5页Journal of Transport Science and Engineering

基　　金：国家自然科学基金项目(50178011)

摘　　要：边坡临界滑面的位置和形状与其安全系数至少具有同等重要的地位,它们对工程加固措施更具有直接的指导意义.非圆弧任意临界滑面的搜索是一个多极点、非凸性且目标函数常常无法显式表达的非线性规划问题,采用传统的数学优化方法会遇到很多困难.蚁群算法是一类实现复杂非线性优化问题的重要全局仿生算法.基于离散点集的蚁群算法已被成功应用于边坡滑动面的搜索,其效果显著,但离散点的存在也导致对滑动面形态分辨率较低.通过在边坡条分线上引入插值技术来实现蚂蚁转移目标连续化,克服了离散转移目标对滑动面形状分辨率低的缺点.算例的分析结果表明:这种改进的效果非常有效.The position and the shape of the critical slip surface of a slope can be provid- ed as the significant directive to slope reinforcement, which has at least equal impor- tance with the Factor of Safety （FOS） of the slope. The search of non-circular slip face is in essence a non-linear programming problem. The existed classic optimization meth- ods have difficulties in solving this problem. The ant colony algorithm is a modern global optimization algorithm and has been successfully used in the search of the critical slip surface of slope. In existed research, the slope is required to be discretized to points. But the introduction of discretized points lead to the lower representation of the slope. This is overcome by employing the interpolation technique on the slice line of slope, which can provide a continuous destination set for the ants to transfer. Results from the verifica- tion examples show the effectiveness of the method presented.

关 键 词：边坡 临界滑动面 蚁群算法 连续转移目标 

分 类 号：O319.56[理学—一般力学与力学基础]