基于混沌局部扰动蚁群算法的边坡稳定聚类分析

针对蚁群聚类算法用于边坡稳定性分析时存在的缺陷,在传统蚁群聚类算法的基础上,引入混沌特有的随机性和遍历性,利用Logistic映射函数对蚁群分布矩阵进行初始化,以获得较优的初始分布,同时对聚类参数进行局部扰动,以增强跳出局部极值的能力,从而建立了基于混沌局部扰动蚁群算法。以三峡库区的36组边坡资料为例,应用该方法与常规聚类方法、模拟退火算法和常规蚁群聚类算法进行对比分析。结果表明,该方法方便、快捷,聚类效果优于其他三种方法。

Cluster Analysis of Slope Stability Based on Ant Colony Algorithm with Chaotic Local Disturbance

Aiming at defects of the ant colony clustering algorithm used in slope stability analysis, based on the traditional ant colony clustering algorithm, the chaotic characteristics of the randomicity and ergodicity are introduced. The Logistic mapping function is used to initialize ant colony distribution matrix for getting the better initial distribution. At the same time, the clustering parameters are locally disturbed in order to enhance the ability to jump out of local minimum. So, the ant colony algorithm with chaotic local disturbance is proposed. Taking 36 groups of engineering slope data in Three Gorges Reservoir for an example, conventional cluster analysis, simulated annealing and conventional ant colony clustering algorithm are compared with the proposed method. The results show that the proposed method is more convenient and fast, and it is better than other three methods.