Hierarchically weighted rough-set genetic algorithm of rock slope stability analysis in the freeze–thaw mountains

Abstract

The instability of rock slope in the freeze–thaw mountains has abrupt and uncertain characteristics, and its prediction accuracy often is very low. How to accurately estimate the stability of this kind of rock slope is an urgent problem to be solved. In this paper, we presented the evaluation method of rock slope stability in the freeze–thaw mountains coupling the hierarchy analysis, the rough-set theory and the genetic algorithm. Fifty highway rock slopes in Taishun county of Zhejiang province, China, were selected as the examples in the study. The main factors influencing rock slope stability were obtained using the analytic hierarchy process (AHP), and their corresponding weights were given. The evaluation rules were extracted from rock slope instability examples using the hierarchy analysis, the rough-set theory and the genetic algorithm. Then the approximate reasoning method was proposed using the method of inexact reasoning and default inference. The proposed method takes inexact reasoning and default reasoning to deal with incomplete match or absence of information. The results show as follows: it is reasonable and feasible to evaluate rock slope stability using the proposed method; using the proposed method, we can grasp the key factors influencing on rock slope stability, and the reasoning is flexible; the presented method can improve the intelligent prediction accuracy of rock slope instability in the freeze–thaw mountains. It remains to be further studied whether the method presented in this paper is suitable for the stability evaluation of other rock slopes.