A Prediction of Slip Surface Depth by using A Fuzzy Inference Method with Genetic Algorithm

Abstract

A decision of slip surface depth is the most important subject on the discussion of landslide mechanism and the making plan of prevention measures. This paper discusses the probability of prediction of the slip surface depth by using a fuzzy inference method. In this analysis, firstly, a fuzzy relation matrix method is applied to the selection of factors using in the fuzzy inference. Secondary, the most suitable membership shape meaning as a probability distribution is decided from the self-tuning of fuzzy inference by using a genetic algorithm with the selected factor data. The result obtained from the inference method constructed by the above described procedures has good correlation with the observed depth. Weight Coefficients of the inference rules show the significant factors in the proportion to the arrangement order of landslide width, length and uneven ground counter.