Multiple wave is one of the important factors affecting the signal-to-noise ratio of marine seismic data. The model-driven-method (MDM) can effectively predict and suppress water-related multiple waves, while the quality of the multiple wave contribution gathers (MCG) can affect the prediction accuracy of multiple waves. Based on the compressed sensing framework, this study used the sparse constraint under L0 norm to optimize MCG, which can not only reduce the false in the prediction and improve the image accuracy, but also saves computing time. At the same time, the MDM-type method for multiple wave suppression can be improved. The unified prediction of multiple types of water-related multiple waves weakens the dependence of conventional MDM on the adaptive subtraction process in suppressing water-related multiple waves, improves the stability of the method, and simultaneously, reduces the computational load. Finally, both theoretical model and practical data prove the effectiveness of the present method.

中文翻译：

---

基于L0范数稀疏约束的多波预测与抑制

多重波是影响海洋地震数据信噪比的重要因素之一。模型驱动方法（MDM）可以有效地预测和抑制与水有关的多波，而多波贡献集（MCG）的质量会影响多波的预测精度。在压缩感知框架的基础上，利用L0范数下的稀疏约束对MCG进行优化，不仅可以减少预测中的错误，提高图像的准确性，而且可以节省计算时间。同时，可以改进用于多波抑制的MDM型方法。多种与水有关的多波的统一预测削弱了常规MDM在抑制与水有关的多波时对自适应减法过程的依赖性，提高了方法的稳定性，同时减少了计算量。最后，理论模型和实际数据均证明了该方法的有效性。