相关向量机在地震滑坡敏感性分析中的应用

发布时间：2018-04-02 09:37

本文选题：相关向量机　切入点：遗传算法　出处：《科技导报》2017年15期

【摘要】：地震滑坡敏感性分析是地震次生灾害研究的重点内容之一。数据量大且致灾因素复杂是研究地震滑坡问题的难点。在对已有敏感性分析模型研究的基础上,以芦山地震为例,选取地面高程、坡度、坡向、地层、斜坡形态、斜坡结构、距断层平均距离、距水系平均距离、地震峰值加速度9个地震滑坡评价因子,建立基于遗传算法的相关向量机(GA-RVM)敏感性分析模型,生成地震滑坡敏感性区划图,统计结果显示滑坡正确率为99.74%,滑坡密度在极高敏感区达到27.4057个/km2。结果表明,相对于基于遗传算法的支持向量机,GA-RVM获得了更高的预测精度,可为进一步完成地震灾害预防提供依据。
[Abstract]:The sensitivity analysis of earthquake landslide is one of the important contents in the study of earthquake secondary disasters.It is difficult to study the problem of earthquake landslide with large amount of data and complicated factors.Based on the existing sensitivity analysis models, taking the Lushan earthquake as an example, the surface elevation, slope, slope direction, stratum, slope shape, slope structure, average distance from fault and average distance from water system are selected.Nine seismic landslide evaluation factors with peak acceleration of earthquake are established, and the sensitivity analysis model of correlation vector machine GA-RVM based on genetic algorithm is established, and the sensitivity map of seismic landslide is generated.The statistical results show that the correct rate of landslide is 99.74 and the density of landslide is 27.4057 / km2 in extremely sensitive area.The results show that the prediction accuracy of GA-RVM is higher than that of GA-RVM based on genetic algorithm, which can provide the basis for the further completion of earthquake disaster prevention.
【作者单位】：中国地质大学(武汉)地球物理与空间信息学院;武汉工程大学资源与土木工程学院;长安大学地质工程与测绘学院;
【基金】：国家高技术研究发展计划(863计划)项目(2012AA121303) 国家自然科学基金青年基金项目(41301386)
【分类号】：P642.22

【相似文献】