IDENTIFICATION OF SLOPES WITH HIGHER RISK TO SLOPE FAILURES BASED ON INFORMATION PROCESSING TECHNIQUES

Abstract

In recent times, the sediment disasters, such as slope failures, debris flows, and landslides,
caused by typhoons or cloudbursts have occurred in Japan. The progression of global warming will increase
the scale of typhoons and cloudbursts striking the Japanese Islands, and there is a concern that the frequency
of sediment disasters may increase. Therefore, it is important to identify slopes with a higher risk to sediment
disasters to prevent future disasters. In this study, a method based on artificial neural networks and
mathematical statistics was used to identify such slopes. In the proposed method, the self-organizing map
(SOM), cluster analysis, and Hayashi’s second method of quantification are combined. The proposed method
was applied to the data gathered from periodical inspections of road slopes. In the results, slopes with a higher
risk to slope failure were identified and ranked according to their risk.