为研究永胜县泥石流灾害易发性，以云南省丽江市永胜县为研究区，将其划分为475个子流域单元。采用灰色关联分析法对各个评价因子进行关联度计算，剔除掉关联度最低的因子，使用共线性诊断进行独立性检验，最终保留流域平均坡度、流域多年平均最大降雨量、流域平均植被覆盖度、流域平均高程、流域平均melton比率、流域平均水系密度、流域平均崩滑核密度、流域平均道路密度等8个评价因子。评价因子的信息量通过信息量法计算得出，将关联度值作为各因子叠加的权重值，进一步构建灰色关联分析-信息量模型，开展永胜县泥石流易发性评价研究。结果表明：泥石流灾害极高易发区频率比高达4.06，接受者操作特征曲线（receiver operating characteristic curve, ROC）下的面积为0.818，说明选取的8个评价因子以及灰色关联分析-信息量评价模型对永胜县泥石流灾害易发性评价研究具有较好的预测能力，得到的评价结果也对当地的泥石流灾害防治工作提供了一定的参考作用。

In order to study the disaster susceptibility of debris flow in Yongsheng County, the research area was Yongsheng County of Lijiang City, Yunnan Province, and it was divided into 475 sub-watershed units. Grey correlation analysis method was used to calculate the correlation degree of each factor, and the factor with the lowest correlation degree was eliminated. The independence of factors was tested by collinearity diagnosis. In the end, eight factors including average slope, average annual maximum rainfall, average vegetation coverage, average elevation, average melton ratio, average water system density, average landslide core density and average road density were retained. The information volume of the factors was calculated by the information volume method, and the correlation degree value was taken as the weight value of the superposition of each factor. The grey correlation analysis-information volume model was further constructed to carry out the evaluation research on the vulnerability of debris flow in Yongsheng County. The results show this as follows. The requency ratio of debris flow disaster points in the extremely vulnerable area is as high as 4.06, and the area under the ROC (receiver operating characteristic) curve is 0.818, indicating that the selected eight factors and the grey correlation analysation-information volume method have good forecasting ability for the evaluation of debris flow disaster vulnerability. The results can also be used as reference for the prevention and control of debris flow disaster in Yongsheng County.