基于神经网络模型的斜坡地质灾害易发性评价

所属分类:GIS/地图编程
开发工具:Python
文件大小:16931KB
下载次数:6
上传日期:2020-04-11 21:10:42
上 传 者顾林丶
说明:  吉林省永吉县存在大量的斜坡地质灾害,为了给永吉县斜坡地质灾害的防治和预警提供高效直观的分析模型,将吉林省永吉县作为研究区,选取高程、坡度、坡向、剖面曲率、平面曲率、距断层距离、岩性、距河流距离、年均降雨量、地形湿度指数和植被覆盖指数等11 个评价因子,利用神经网络模型进行区域斜坡地质灾害易发性分析,再选用频率比、支持向量机模型进行对比。利用ROC曲线对模型的准确性进行验证分析,得出神经网络、频率比和支持向量机模型的成功率分别是91. 3%、89. 3%、90. 2%,预测率分别是87. 3%、84.3%、85. 6%。结果表明: 神经网络模型的精度最高,更适用于永吉县斜坡地质灾害的易发性评价
(There are a lot of slope geological disasters in Yongji County, Jilin Province. In order to provide an efficient and intuitive analysis model for the prevention and early warning of slope geological disasters, Yongji County, Jilin Province is selected as the research area, which includes elevation, slope, slope direction, section curvature, plane curvature, distance from fault, lithology, distance from river, annual rainfall, terrain humidity index and vegetation coverage index 11 evaluation factors, using neural network model to analyze the vulnerability of regional slope geological hazards, and then using frequency ratio, support vector machine model for comparison.)

文件列表:
基于神经网络模型的斜坡地质灾害易发性评价_以吉林永吉为例.pdf (18986536, 2020-02-01)

近期下载者

相关文件


收藏者