基于IR—ADMM组合技术对地震随机噪声的压制

稀疏表示是一种现行有效的随机噪声压制方法，常采用交替方向乘子法逐道分解地震信号，但实际应用中交替方向乘子法计算效率高但精度不足，难以满足高保真地震数据处理的要求。通过结合迭代重加权和交替方向乘子法2种算法，提出了一种新的基于迭代重加权交替方向乘子法的联合稀疏表示方法，兼具收敛速度快和重建精度高的优点。共偏移距道集地震数据具有水平同相轴结构，满足共稀疏性条件，将联合稀疏表示算法应用于共偏移距道集就能够利用信号的空间相干性，提高去噪算法性能。理论和实际资料试算结果表明，所提算法具有较好的应用效果。

Suppression of Seismic Random Noise Based on IR-ADMM Composite Technology

Sparse representation is an effective method to suppress random noise，which often uses the alternating direction method of multipliers to decompose seismic signals trace by traceAlthough the alternating direction method of multipliers has the advantages of high computational efficiency in practical application，its accuracy is insufficient and it is often difficult to meet the requirements of high-fidelity seismic data processingAiming at this problem，a new joint sparse representation based on iterative reweighted alternating direction method of multipliers is proposed by combining the iterative reweighted algorithm and the alternating direction method of multipliers，which has the advantages of faster convergence and better reconstruction accuracy than the candidatesThe common offset gather approximately satisfy the common sparsitySo，the joint sparse representation algorithm can be applied to the common offset gathers to improve the performance of the denoising by using the spatial coherence of tracesThe experimental results show that the proposed algorithm exhibit better efficiency than conventional implementation