Abstract: When underground or surface excavations are made in rock masses, the behavior of the surrounding rock material will be greatly influenced by the presence of discontinuities. Various modes of rock slope and wedge failure can be attributed to the existence of fractures in a rock mass. In order to make rock mass structure plane group reasonably, and simulate the distribution of rock mass structural plane network by computer, this paper proposes a fusion of clustering method with improved genetic algorithm and support vector machine. First of all, according to the rock mass structural plane orientation information to establish the mathematical model of structural plane grouping, the global optimal clustering center of structural plane is obtained by using the improved genetic algorithm. Then the structural plane is grouped completely with support vector machine. The application of practical engineering shows that the genetic-support vector machine clustering algorithm for grouping of the rock mass structure plane is reasonable and reliable.

Key words: clustering analysis of structural plane, improved genetic algorithm, support vector machine (SVM), dominant orientation