基于支持向量机的地震灾害自动成图方法  

Automatic Mapping Method of Earthquake Disaster Based on Support Vector Machine

作　　者：陈小芳[1] 戚洪飞 俞岗[1] CHEN Xiaofang;QI Hongfei;YU Gang(Key Laboratory of Earthquake Monitoring and Disaster Mitigation Technology,Guangdong EarthquakeAgency,Guangzhou 510070,China;Wuhan ZONDY CYBER T&S CO.,LTD,Wuhan 430000,China)

机构地区：[1]广东省地震局地震监测与减灾技术重点实验室,广东广州510070 [2]中地数码科技有限公司,湖北武汉430000

出　　处：《灾害学》2021年第1期117-121,共5页Journal of Catastrophology

基　　金：广东省科技计划项目(2017B030314082);广东省防震减灾现代化试点省项目(2060301)。

摘　　要：地震灾害风险评估中有许多定性图件,目前常用的成图方法是根据研究区的离散点钻孔结果人工勾画,费时费力;或对离散点钻孔结果强制赋予0、1数值,然后进行插值成图,其从逻辑上不能真实反映场地的客观情况。基于支持向量机技术,该文提出了使用分类工具SVC,将不同结果的离散点钻孔在空间上进行合理分类的自动成图方法。相较于传统方法,该方法效率高,成本低。There are many qualitative maps in the risk assessment of earthquake disaster.At present,they are commonly manual sketched based on the borehole results of discrete points in the study area,which is time-consuming and laborious.Or the borehole results of discrete points are forced to be given values of 0 and 1,and then the interpolation mapping is carried out.It can’t really reflect the objective situation of the site logically.Based on the technology of support vector machine,this paper proposes aautomaticmapping method by using the classification tool SVC to classify the discrete points with different results reasonably in space.Compared with the traditional method,this method has high efficiency and low cost.

关 键 词：地震灾害 风险评估 钻孔 定性灾害图件 自动生成 

分 类 号：X43[环境科学与工程—灾害防治] X915.5[环境科学与工程—安全科学]