基于分形方法的地震子波提取及应用

Seismic wavelet extraction and application based on fractal technique

成都理工大学,四川成都 610059

Chengdu University of Technology, Chengdu 610059, China

常规自相关子波求取方法假设反射系数功率谱为白噪谱,因而在实际应用中受到限制。以分形理论为基础的地震子波功率谱估计方法,可突破反射系数功率谱为白噪的假设约束,适应反射系数序列具有一定相关性的地震资料处理。理论模型和实际地震资料处理表明,分形技术应用于复赛谱求解地震子波的方法中,可适应非白噪随机反射系数序列地震记录的子波求取。用基于分形技术的复赛谱法求取的子波对地震记录进行反褶积处理,分辨率明显高于自相关子波求取方法。

As we assumed the power spectrum of the reflection coefficient is white noise in traditional method to calculation autocorrelation wavelet, so there is some limitation in practice. The author studies a new method to estimate wavelet power spectrum which based fractal theory, and it breaks though the above assumption, adapts to processing the seismic data which the reflection coefficient series have some cor-relativity. The theoretical model and practical dataprocessing example is demonstrate, the fractal technique apply to calculating cepstrum for solving wavelet, suits to compute the wavelet of non-white noise reflection coefficient series seismic recorder, and the deconvolution processing by wavelet based fractal method, the resolution ratio is obviously higher than the autocorrelation wavelet method.