节点文献
基于归一化局部互相关算法的相位编码全波形反演
The Phase-Encoded Full Waveform Inversion Based on Normalized Local Cross-Correlation Algorithm
【摘要】 本文提出了一种基于归一化局部互相关算法的相位编码全波形反演方法,该方法细化了每个时窗内各模拟道与实际记录道的相关度,使得残差记录能更精确地反映速度模型扰动对于波场的影响,因此其适应于单边观测地震数据的相位编码全波形反演。模型实验表明:相比于归一化全局互相关算法,归一化局部互相关算法可更有效压制单边观测地震数据残差记录中的观测残差,并且基于归一化局部互相关算法的相位编码全波形反演的反演精度明显优于基于归一化全局互相关算法的相位编码全波形反演,其有望在实际数据的全波形反演中发挥重要作用。
【Abstract】 In this paper,a method of phase-encoded full waveform inversion based on normalized local cross-correlation algorithm is proposed,which can discribe the correlation degree of modelled seismic traces and actual seismic traces in each time window,so the method is more suitable for the phase-encoded full waveform inversion of unilateral observation seismic data since the residual record can reflect the effect of velocity model perturbation on wave field accurately,Numerical experiments show that:the normalized local cross-correlation algorithm can more effectively suppress the observation residuals in the residual record of unilateral observation seismic data compared with the normalized global cross-correlation algorithm,and the inversion accuracy of full waveform inversion based on normalized local cross-correlation algorithm is obviously better than that of full waveform inversion based on normalized global cross-correlation algorithm,so it is expected to play an important role in waveform inversion of actual data.
- 【文献出处】 中国海洋大学学报(自然科学版) ,Periodical of Ocean University of China , 编辑部邮箱 ,2017年07期
- 【分类号】P631.4
- 【被引频次】4
- 【下载频次】191