PREDICTION OF LANDSLIDE SUSCEPTIBILITY FOR MIDDLE SECTION OF CHINA-KYRGYZSTAN-UZBEKISTAN RAILWAY PROJECT LINE BASED ON MAMDANI-FIS MODEL

The China-Kyrgyzstan-Uzbekistan railway is an international gateway from northwest China to Central Asia and Southern Europe. The forecast of geological disasters along the railway can provide some suggestions for its route selection. The Mamdani fuzzy inference system(Mamdani FIS)model is applied on the prediction of landslide susceptibility of the study area on AK53-AK130 of the north and AK60-AK131 of the south along the railway project line. Based on the nine kinds of landslide influencing factors,768 rules of inference are established. These factors include geological environment background,topographic factors and ecological environment and are collected by remote sensing and regional geological background data. The aim of the study is to apply the Mamdani FIS model for the China-Kyrgyzstan-Uzbekistan railway landslide prediction,and then divides the research area into extremely high prone zones,high prone zones,moderate prone zones,low prone zones,and extremely low prone zones. The result shows that the extremely high prone area and high prone area of landslide are distributed near the Fergana Mountains in the north and the Arthur River Basin in the south. Based on the results,the area under the curve(AUC)obtained from the Receiver Operating Characteristic Curve(ROC curve) is 0.859,which indicates that the prediction results are successful.