霍西煤矿区地表滑坡灾害敏感性数值建模与等级划分

发布时间：2018-02-27 13:19

本文关键词： 滑坡敏感性 GIS 支持向量机交叉验证霍西煤矿区　出处：《地球信息科学学报》2017年12期 　论文类型：期刊论文

【摘要】：本文以山西省霍西煤矿区为研究区,利用遥感和GIS方法对滑坡灾害的敏感性进行了数值建模与定量评价。利用交叉检验方法构建了径向基核函数支持向量机滑坡敏感性评价模型,并基于拟合精度对模型进行了定量评价;对各评价因子在模型中的重要性进行对比分析;基于空间分辨率为30 m的评价因子,通过径向基核函数支持向量机模型获得了霍西煤矿区滑坡敏感性指数值,并利用分位数法将霍西煤矿区的滑坡敏感性分为极高、高、中和低4个等级。结果表明:拟合精度建模阶段和验证阶段分别为87.22%和70.12%;与滑坡敏感性关系最密切的5个评价因子依次是岩性、距道路距离、坡向、高程和土地利用类型;极高和高敏感区域分布了93.49%的滑坡点,面积占总面积的50.99%,是比较合理的分级方案。本研究不仅可以为研究区人工边坡调查和煤矿资源合理开采提供借鉴,对相似矿区的相关工作也具有参考价值。
[Abstract]:This paper takes Huoxi Coal Mine area of Shanxi Province as the research area, Using remote sensing and GIS methods, the sensitivity of landslide was modeled and evaluated quantitatively, and the radial basis function support vector machine (RBF SVM) was used to evaluate the sensitivity of landslide. The model is evaluated quantitatively based on fitting accuracy, the importance of each evaluation factor in the model is compared and analyzed, and based on the evaluation factor with spatial resolution of 30 m, The sensitivity index value of the landslide in Huoxi coal mine is obtained by using the support vector machine model of radial basis function kernel function, and the sensitivity of the landslide in the Huoxi coal mine is divided into very high and high by using the quantile method. The results show that the fitting precision modeling stage and verification stage are 87.22% and 70.12 respectively, the five evaluation factors most closely related to landslide sensitivity are lithology, distance from road, slope direction, elevation and land use type. 93.49% landslide points have been distributed in extremely high and highly sensitive areas, accounting for 50.99 of the total area, which is a reasonable classification scheme. This study can not only provide a reference for the investigation of artificial slopes in the study area and the rational mining of coal resources. It also has reference value to the related work of similar mining area.
【作者单位】：太原理工大学矿业工程学院测绘科学与技术系;山西省地质环境监测中心;
【基金】：国家重点研发计划“地球观测与导航”专项(2016YFB0502601) 国家自然科学基金项目(41371373;41301469)
【分类号】：P642.22;TD325.2

【相似文献】