基于小波包分解及AR模型的单通道地震波信号初至点检测

Onset point identification of single-channel seismic signal based on wavelet packet and the AR model

地震波信号的初至点检测在核爆地震的自动识别中起着重要作用,本文描述了一种单通道地震波信号初至点检测方法,它是基于小波包分解及对大地噪声信号进行统计分析后建立的大地噪声模型———AR模型而提出的.文中对该方法进行了理论分析及实际地震波信号检验.研究结果表明,该方法对地震波信号具有很强的自适应能力,能定量确定初至点的位置.

Distinguishing onset points of seismic signals is important in the automatic identification of underground nuclear explosions and natural earthquakes. This paper presents a method of onset point detection of single channel seismic signals based on wavelet packet and the AR model. Theoretical analysis and test of seismic signals indicate that this method is strongly self-adaptive to seismic signals and can detect the position of the onset in seismic trace precisely.