基于Mamdani FIS的中吉乌铁路方案线中段滑坡易发性预测

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

中吉乌铁路是我国西北地区通往中亚、南欧国家的一条国际通路,对其方案线沿线的地质灾害的调查及预测可为其选线提供一定建议。本文基于Mamdani模糊推理系统(Mamdani FIS)对方案线北线AK53-AK130、南线AK61-AK131段研究区的滑坡易发性进行预测。通过区域地质背景资料和遥感影像人机交互解译获取了该区地质环境背景、地形因素以及生态环境3类9种滑坡影响因子,建立768条推理规则,通过Mamdani FIS模型得到区内滑坡敏感度文件,在GIS环境中制作研究区滑坡易发性等级图,将研究区划分为极低易发区、低易发区、中等易发区、高易发区和极高易发区。结果显示,滑坡极高易发区和高易发区分布在研究区东北部的费尔干纳山脉附近以及南部的亚瑟河流域。采用受试者工作特征曲线(Receiver Operating Characteristic Curve,ROC曲线)对结果进行验证,曲线下面积为0.859,表明预测结果精度较高。

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.