关键词: 岩体, 结构面产状, 优势分组, 模拟退火算法, K-means聚类

Abstract: Aiming at the complexity， poor clustering accuracy and low grouping efficiency of the previous methods of dominant grouping of discontinuity orientation， a new method of dominant grouping of discontinuity orientation based on simulated annealing algorithm and K-means clustering (SAK) was proposed. The algorithm is simple and easy to implement. Based on the annealing principle of simulated annealing algorithm， the grouping results of K-means algorithm were optimized， which aimed to overcome the shortcoming of the K-means algorithm’s susceptibility to the initial clustering center. The analysis of discontinuities generated by computer simulation showed that the proposed method is superior to the traditional K-means algorithm. The method was applied to the grouping of measured discontinuity orientation of Xinglong Tunnel on Third Ring Expressway of Chongqing City， and compared with the existing methods. The results showed that this method not only has high clustering accuracy， but also has fast iteration speed and strong engineering practicability.

Key words: rock mass, discontinuity orientation, dominant grouping, simulated annealing algorithm, K-means clustering