Study of Seepage Properties of Fractured Rock Mass Based on Improved K-Means Clustering Algorithm

Seepage properties of fractured rock mass are of prime importance for hydraulic engineering and accurate description of rock fracture geometry parameters is an important and basic task in rock hydraulics. In this paper, an improved K-means clustering algorithm for structural plane of fractured rock mass was first brought forward and the corresponding Matlab program for discontinuity orientations partitioning was compiled and then used in the fitting analysis of dominant orientations of certain dam foundation rock mass. On this basis, combining calculation formulas of multi-group fractures, the permeability tensor and principle value was calculated for the actual dam foundation. The results provide a theoretical and computational reference for other similar projects.