蚁群算法与遗传算法融合及其在边坡临界滑动面搜索中的应用

临界滑动面搜索是边坡稳定性分析中一项非常重要的内容。相对于圆弧滑动面的确定，只需要圆心和半径3个未知量，非圆弧滑面的确定则需要找出若干个控制点，是一个多维空间的优化问题。非圆弧滑动面优化搜索问题相当复杂，常规优化算法往往达不到要求。改进了蚁群算法，使其具备在连续空间的搜索能力，并与遗传算法融合，形成优势互补，克服了遗传算法的无反馈能力导致无用的冗余迭代、求解效率低以及蚁群算法初期信息素匮乏导致算法速度慢的不足。通过与商用软件GEO-SLOPE的算例求解结果对比，来说明本算法的有效性。

Hybrid of ant colony algorithm and genetic algorithm and its application to searching critical slope slip surface

Searching critical slip surface is very necessary in slope stability analysis. Noncircular slip surface search is a multi-dimensional optimization problem to find out several control points, and more complex than searching circular slip surface, in which only center and radius of circular slip surface need to be determined. Common optimization methods can not search critical noncircular slip surface very well due to the complexity of the problem. To overcome the multiple redundancy and low efficient solving of genetic algorithms as a result of no feedback ability, as well as low speed of ant colony algorithms owing to absence of original pheromone, a modified ant colony algorithm has been improved, which is capable of searching in continuous space, to integrate with genetic algorithm to complement each other’s advantages. Finally, a suppositional example calculated by text-mentioned methods and GEO-SLOPE, is illustrated to explain effectivity of the algorithm.