论文部分内容阅读
谱分解是应用于油气藏描述的一种有效方法.谱分解方法如短时傅里叶变换、连续小波变换、S变换等因使用预先选定的窗函数和基函数限制了其时频分辨率.对于双线性Wigner-Ville分布(WVD)谱分解,其在计算时频分布不要求选定窗函数,时间和频率分辨率均很高,但存在交叉噪声.在本文中,我们提出利用迭代反演方法计算WVD谱分解.其关键思想是,利用复正则化非稳态回归技术拟合解析地震道和它的傅里叶分量,并且将时变的傅里叶系数定义为WVD时频分布.理论合成数据表明,基于复正则化非稳态回归WVD谱分解方法有较高的时间频率分辨率,并且能有效地压制交叉噪声.最后,我们将本文方法应用于三维实际地震数据进行河道检测.
Spectral decomposition is an effective method applied to reservoir description. Spectral decomposition methods such as short-time Fourier transform, continuous wavelet transform, S-transform, etc. have limited their time-frequency resolution due to the use of pre-selected window functions and basis functions . For bilinear Wigner spectral decomposition (WVD), its time-frequency distribution does not require the selected window function in time-frequency and frequency resolution, but crossover noise exists.In this paper, we propose to use iterative The key point of the inversion method is to calculate the WVD spectral decomposition.The key idea is to fit the seismic trace and its Fourier components by using the regularization unsteady regression technique and define the time-varying Fourier coefficients as WVD time-frequency distribution The theoretical synthesis data show that the WVD spectral decomposition method based on regularization unsteady regression has higher time-frequency resolution and can effectively suppress cross-noise.Finally, we apply the method to three-dimensional real seismic data for river detection .