基于支持向量机岩溶塌陷的智能预测模型  

Intelligent Model of Karstic Collapse Prediction Based on SVM

作　　者：赖永标[1] 乔春生[1] 

机构地区：[1]北京交通大学土木建筑工程学院,北京100044

出　　处：《北京交通大学学报》2008年第1期36-39,43,共5页JOURNAL OF BEIJING JIAOTONG UNIVERSITY

基　　金：国家自然科学基金资助项目(50478061);铁道部科研究开发计划项目(2003G036)

摘　　要：在综合分析了各种岩溶塌陷预测方法和介绍支持向量机的基础上,提出基于支持向量机的岩溶塌陷预测方法,运用Matlab语言编程,建立了相应的岩溶塌陷预测模型.以已有的岩溶地面塌陷实例为学习样本,进行学习测试,得到训练效果较佳的预测模型,并用此模型对某市岩溶塌陷进行预测.结果表明,支持向量机预测模型具有较高精度,在岩溶塌陷预测研究中具有广阔的应用前景.Based on comprehensively analyzing different methods for Karstic collapse Prediction and introducing the support vector machine （SVM）, the article proposed the method of Karstic ground collapse Prediction with SVM and built prediction model with Matlab language. By using the karstic ground collapses as learning samples, a good Prediction model was obtained by studying and training. Then this model was used in prediction of the karstic ground collapses in one city. The result suggests that the SV M prediction model has high prediction accuracy, and it has Wide application prospects in the research of karstic collapse Prediction.

关 键 词：岩溶塌陷 支持向量机 预测模型 

分 类 号：P642[天文地球—工程地质学]