【摘要】 研究了从天然地震和人工爆破事件的波形记录中提取出来的能量比特征在天然地震和人工爆破事件的自动识别中的有效性及适用性。对波形记录进行了4层小波变换,然后对变换得到的小波系数提取能量比特征,最后利用支持向量分类机ν-SVC进行识别效果检验。实验证明,由bior2.2小波包分解后提取出来的能量比特征对天然地震和人工爆破事件的识别效果很好,可用于实际的自动识别系统作为识别判据之一。更多还原

【Abstract】 We study the validity and applicability of the energy ratio feature exacted from waveform records of natural earthquake and artificial explosion for automatic recognition of natural earthquake and artificial explosion.We apply the 4-level wavelet transform on seismic wave records and exact the energy ratio feature of the wavelet coefficients obtained from the transform,then,use support vector classification ν-SVC to examine the recognition effect.The results show that the energy ratio feature exacted after the bior2.2 wavelet packet decomposition is a very good feature to discriminate earthquake and explosion,which implies its applicability in related application such as our ongoing project of earthquake and explosion automatic recognition system.更多还原