Location of the critical slip surface based on monitoring data and genetic algorithm

Abstract

The non-circular search limit equilibrium method (LEM) is a significant methodology in the analysis of slope stability, aimed at identifying the critical slip surface (CSS) and the safety factor (SF). However, the existing LEM does not consider monitoring data, such as the subsurface displacement, in the CSS search process for the actual slopes. In this paper, a new approach is proposed that combines LEM with subsurface displacement data to more accurately identify the CSS. The k-means clustering algorithm is utilized to determine the recognizing slip positions corresponding to each subsurface displacement curve. Additionally, the Spencer’s method is employed to calculate the SF. By treating the recognizing slip positions as constraint conditions, the genetic algorithm (GA) is applied to search for the location of the CSS. To validate the accuracy and usefulness of the proposed method, four numerical examples and an experimental case are presented.