边坡临界滑动面搜索的奖惩蚁群算法研究

摘要

Critical slip surface searching is vital to slope stability analysis. Essentially, this is a path searching problem, which can be solved by the ant colony optimization method. To overcome the shortcomings of traditional ant colony optimization, the premium-penalty strategy is introduced, and the pheromone diversity of the good paths and the ordinary ones is increased to polarize pheromone density of all paths. Thus, a new algorithm namely Premium-Penalty Ant Colony Optimization is proposed to discrete and simu- late the problem of the critical slip surface searching, by which the critical slip surface of slope can be lo- cated. Through two typical examples, one with simple slope and the other with complicated slope, and an application example of a slope project of a dam, the efficiency and effectiveness of the new algorithm are verified. The results show that, not only for simple slope but also for complicated slope, the new algorithm can always find the better critical slip surface in shorter time than many previous algorithms, and the new algorithm can be used in engineering practice very well.%边坡滑动面搜索是边坡稳定计算中一项关键的问题，其实质为安全系数最小滑动路径的搜索问题，本文对采用路径搜索的蚁群算法引入奖惩策略，加大较优路径和普通路径上信息素的差异，分化各条路径上的信息素，以克服其收敛速度慢、早熟收敛的缺点。通过把边坡滑动面搜索模型离散化，采用奖惩蚁群算法解决滑动面搜索问题，提出了一种临界滑动面搜索的新方法。最后对一个简单边坡和复杂边坡的典型算例及一个大坝边坡工程的应用实例进行了计算，验证了新算法的有效性及其高效性。结果表明，无论是对简单边坡还是复杂边坡，本文算法都能以更快的速度搜索到结果更好的临界滑动面，工程应用效果良好。