A comprehensive review of slope stability analysis based on artificial intelligence methods

Abstract

For preventing landslide disasters caused by the slope collapse, it is crucial to research on the investigation of slope stability. For the very complex influence factors on the slope stability, nowadays, the studies on slope stability by using the artificial intelligence methods have become the hot topic, and there have been numerous related works in this field. In this study, according to the essential difference, the studies in this field are divided into two main types (computation of slope stability and evaluation of slope stability), from which, the numerous previous studies have been reviewed comprehensively. For the studies on slope stability computation, according to the used artificial intelligence methods, the studies in this field are reviewed from four aspects: studies by quasi-physical intelligence methods, studies by simulated evolutionary methods, studies by swarm intelligence methods, and ones by hybrid intelligence methods. And from the used artificial intelligence methods too, the studies on slope stability evaluation are also reviewed from four aspects: studies by artificial neural network methods, studies by vector machine methods, ones using hybrid intelligence methods, and ones using other intelligence methods. Moreover, the merits and demerits of those studies have been comprehensively analyzed, and their state-of-the-art research advancement has also been summarized. At last, the possible research directions of slope stability investigation based on artificial intelligence methods are also suggested.