无约束正反向预测误差最小反褶积
|
摘要
由于地震子波的时变性，在假设子波为非时变的情况下，用脉冲反褶积对整个地震道进行处理，所得地震剖面会出现浅层频率高，深层频率低的现象。实践证明，采用开时窗的方法在时窗短时，这种反褶积的效果也很差。Burg反褶积虽然采用滑动时窗，可适应子波的时变性，但由于受Levinson关系的约束，仍然摆脱不了Toeplitz矩阵的性质，不能完全克服脉冲反褶积的缺限。为此，我们提出了无约束正反向预测误差最小反褶积方法。该方法无需解Toeplitz矩阵，也不要求子波是非时变的，并采用Marple递推算法快速实现。理论分析和实际应用的效果表明，该方法是一种对小时窗和大时窗均有效且稳健的反褶积方法。
Under unreasonable presumption that wavelet doesn't vary with time,processing whole seismic trace by pulse deconvolution results in the seismic section thathas high-frequency contents in shallow part but low-frequency contents in deeppart. It has been shown that windowing also brings poor effect of the deconvolutionwhen the time window is short. Burg deconvolution, using running window,is suitable to time-varying property of wavelet;but restricted by Levinson relation,it stillfails to cast off the characteristics of Toeplitz matrix and to remove fully the defectof pulse deconvolution. To cope with the problem, we advance the deconvolutionbased on unconstrained forward-backward prediction error minimization. Themethod needs neither the solution of Toeplitz matrix nor the presumption thatwavelet does not vary with time,and it can be fast achieved by using MARPLE recursion algorithm. Theoretical and real results say that the deconvolution method isfirm and effective on both small window and big window．
Under unreasonable presumption that wavelet doesn't vary with time,processing whole seismic trace by pulse deconvolution results in the seismic section thathas high-frequency contents in shallow part but low-frequency contents in deeppart. It has been shown that windowing also brings poor effect of the deconvolutionwhen the time window is short. Burg deconvolution, using running window,is suitable to time-varying property of wavelet;but restricted by Levinson relation,it stillfails to cast off the characteristics of Toeplitz matrix and to remove fully the defectof pulse deconvolution. To cope with the problem, we advance the deconvolutionbased on unconstrained forward-backward prediction error minimization. Themethod needs neither the solution of Toeplitz matrix nor the presumption thatwavelet does not vary with time,and it can be fast achieved by using MARPLE recursion algorithm. Theoretical and real results say that the deconvolution method isfirm and effective on both small window and big window．
引文
1俞寿朋.高分辨率地震勘探,石油工业出版社,1993
2王宏禹.现代谱估计,东南大学出版社,1990
3Marple L.A new autoregressive spectrum analysys algorithm. IEEE Trans on ASSP,1980,28(4)
4Yilmaz O.Seismic data processing,1987
5钱绍新等.最大熵原理在地震资料处理中的应用.石油地球物理勘探,1984,19(4):295～306
2王宏禹.现代谱估计,东南大学出版社,1990
3Marple L.A new autoregressive spectrum analysys algorithm. IEEE Trans on ASSP,1980,28(4)
4Yilmaz O.Seismic data processing,1987
5钱绍新等.最大熵原理在地震资料处理中的应用.石油地球物理勘探,1984,19(4):295～306
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心