针对岩石边坡稳定分析中常规聚类算法存在收敛速度慢、易陷入局部最优的局限性,基于蚁群信息素的K均值聚类法,提出一种解决边坡稳定性的新方法,分析了三峡库区36个边坡数据资料,并结合工程类比综合判断了边坡的稳定状态。结果表明,该法的聚类效果优于常规聚类法,计算效率高,为边坡稳定性分级的聚类分析评价提供了新途径。 For the stability analysis of rock slope, the conventional clustering algorithm has some limitations such as slow convergence rate and easy falling into local optimum. Based on K-means clustering method of ant colony pheromone, a new method to solve the slope stability is proposed. The data of 36 slopes in the Three Gorges Reservoir Area are combined with the engineering analogy to judge the slope stability. The results show that this method is superior to the conventional clustering method in clustering efficiency and has high computational efficiency. It provides a new way for cluster analysis and evaluation of slope stability classification.