支持向量机在地震与爆破识别中的应用

Application of SVM to discriminating earthquakes and explosions

为研究天然地震事件和爆破事件识别算法,对上海测震台网记录的上海周边区域天然构造地震和爆破事件记录进行小波包分解,并提取特征向量,提出用支持向量机(SVM)识别天然构造地震和人工爆破的算法。结果表明,基于SVM算法的向量识别分类方法,在天然地震和爆破识别中是可用的,准确率预计达85%以上。

In order to study the recognition algorithm of natural earthquake and explosion events, this paper decomposes and extracts the eigenvectors of the natural earthquakes and explosion events in the surrounding area of Shanghai recorded by Shanghai Seismic Network, and proposes the identification of natural earthquakes and explosion events by Support Vector Machine（SVM）. The results show that the method based on SVM algorithm is available in identification research of natural earthquakes and explosion events, and the recognition accuracy is over 85%.