基于局部峰值约束的快速正交匹配追踪地震数据分解方法

Fast orthogonal matching pursuit seismic data decomposition method based on local peak constraint

地震信号稀疏分解在地震数据处理中起着重要作用,可以获取地震资料中蕴含的各种有用信息。正交匹配追踪算法能够较好地实现地震数据自适应分解与重构。正交匹配追踪的目标是选取与信号最匹配的子波,然而在通过内积实现时往往只着重全局尺度上的相似性,并不能保证子波与信号具有相同的位置及振幅。这里采用局部峰值约束的方法来解决这一问题,构建峰值距离的目标泛函,通过快速扫描算法进行高效地求解计算。模型测试结果表明,该方法可以高效实现地震数据的高精度分解与重构,实际资料处理结果显示,本文方法分解后数据波组特征合理,符合地质规律,可以得到精度较高的多尺度数据。

Seismic signal sparse decomposition plays an important role in seismic data processing,which can obtain various useful information contained in seismic data.The orthogonal matching pursuit algorithm can be used to realize the adaptive decomposition and reconstruction of seismic data.The goal of orthogonal matching pursuit is to select the wavelet which best matches the signal residual.However,only the similarity on the global scale is often emphasized when inner product is used.This can not guarantee that the wavelet and residual have the same position and amplitude.In this paper,the local peak constraint method is used to solve this problem.An objective function of peak distance is constructed,and the fast scanning algorithm is used to solve the problem efficiently.The model tested results show that this method can efficiently realize the high-precision decomposition and reconstruction of seismic data.The actual data processing results show that the wave characteristics of the decomposed data are reasonable and conform to the geological law,and the multi-scale data with high precision can be obtained.