基于多窗口联合优化的多次波自适应相减方法

自适应相减是预测减去法压制多次波的重要环节。基于匹配滤波器的多次波自适应相减方法通常采用逐个数据窗口进行匹配来估计滤波器。逐个大数据窗口求解长滤波器进行多次波自适应相减会导致较低的计算效率,另一方面在复杂地质构造中,逐个小窗口匹配估计的短滤波器不足以表征预测多次波与真实多次波之间的复杂差异。为进一步有效均衡一次波保护和多次波压制并且提高计算效率,引入多个窗口联合优化来求解滤波器,同时利用多个小窗口的预测多次波信息来匹配原始数据。引入快速迭代收缩阈值算法(FISTA)求解多个小窗口估计一次波的L1范数最小化约束问题,从而有效分离一次波与多次波。模型数据和实际数据处理结果表明,与传统的逐个大数据窗口进行多次波自适应相减的方法相比,基于多窗口联合优化的多次波自适应相减方法在保持计算精度的同时提高了计算效率;与传统的逐个小数据窗口进行多次波自适应相减的方法相比,基于多窗口联合优化的多次波自适应相减方法可以更有效地平衡多次波去除和一次波保护。

Adaptive subtraction is an important step in multiple suppression that uses prediction and subtraction methods.In general,the adaptive multiple subtraction (AMS) method estimates matching filter by matching each data window one by one.Long filters drawn by the AMS method in large data window lead to low computational efficiency,whereas for complex geology,short filters drawn from each small data window cannot effectively represent the real differences between the true multiples and the predicted multiples.In this study,multiple-window joint optimization was introduced to determine the filter,and prediction multiples of multiple small windows were simultaneously used to match the original data.Meanwhile,the fast iterative shrinkage thresholding algorithm (FISTA) was introduced to solve the optimization problem with the L1-norm minimization constraint on the estimated primaries in multiple small windows.Thus,the primaries and multiples can be effectively separated.The processing results of synthetic data and field data showed that the proposed method can improve computational efficiency while maintaining accuracy compared with the traditional adaptive multiple subtraction method based on large window matching.Compared with the traditional adaptive multiple subtraction method based on small-window matching,the proposed method can balance multiple suppression and primary wave protection more effectively.

国家自然科学基金项目“基于卷积神经网络的多次波自适应相减方法(41804110)”、中石油重大科技项目“塔里木盆地深层复杂高陡构造与碳酸盐岩储层地震速度建模及成像关键技术研究(ZD2019-183-003)”、湖北省科技厅项目(2021CFB119)和湖北省教育厅青年人才基金项目(Q2021204)共同资助。