基于稀疏约束的地震数据高效采集方法理论研究  

Study on highly efficient seismic data acquisition method and theory based on sparsity constraint

作　　者：王汉闯[1,2] 陶春辉[1,2] 陈生昌[3] 丘磊[1,2] 任浩然[3] 周华敏[3] WANG Han-Chuang TAO Chun-Hui CHEN Sheng-Chang QIU Lei REN Hao-Ran ZHOU Hua-Min(Second Institute of Oceanography, State Oceanic Administration, Hangzhou 310012, China Key Laboratory of Submarine Geosciences , State Oceanic Administration, Hangzhou 310012, China School of Earth Sciences, Zhejiang University, Hangzhou 310027, China)

机构地区：[1]国家海洋局第二海洋研究所,杭州310012 [2]国家海洋局海底科学重点实验室,杭州310012 [3]浙江大学地球科学学院,杭州310027

出　　处：《地球物理学报》2016年第11期4246-4265,共20页Chinese Journal of Geophysics

基　　金：国家自然科学基金项目(41374001;41506078);国家重点基础研究发展计划(973计划)(2012CB417305);中国大洋专项(DY125-11-01;DY125-11-05);中央级公益性科研院所基本科研业务费专项(JG1608)资助

摘　　要：随着地震勘探目标复杂化和精细化程度的提高以及"两宽一高"等采集技术的广泛应用,当前地震数据采集的时间越来越长、成本越来越高.针对此问题,本文基于压缩感知理论开展了地震数据高效采集方法的改进和探索研究.根据波动方程解的一般表示式,从波场传播的角度给出了地震数据具有稀疏性的数学物理依据及寻找适应地震数据稀疏变换的一般方案;在稀疏性先验信息的指导下,发展了具有"蓝色噪声"频谱特征的改进的分段采样方法,并基于最优化理论提出了地震数据重建方法.地震数据的稀疏性理论、稀疏约束下的高效采集方法以及地震数据的重建方法构成了相对完善的地震数据高效采集理论.把该理论用于指导地震数据采集,即利用稀疏约束的随机采样方法改变常规规则密集测网中炮点和检波点(或二者之一)的分布,设计了三种随机且均匀的高效采集测网,提出了利用相应测网获取的地震数据重建为常规规则密集测网地震数据的针对性方案,并使用重建精度、高效采集数据的直接成像和重建后再成像的结果对比证明了上述重建方案的有效性.基于Marmousi模型的高效采集试验检验了本文构建的基于稀疏约束的地震数据高效采集方法理论框架在提高当前地震数据采集效率、降低勘探成本上的优势以及方法的有效性和可行性.With the high-density,high-fold and wide-azimuth seismic data acquisition methods being widely used to deal with the increasingly complex and sophisticated situations of exploration targets,the acquisition period is becoming longer and longer and the acquisition cost is becoming higher and higher than before.To tackle the above problems,we carry out the study on the highly efficient seismic data acquisition method based on the theory of compressive sensing.Weinvestigate the mathematical and physical foundation of seismic data′s sparsity and the corresponding sparse representation method according to the general formula of wave equation′s solutions.Under the guidance of sparsity prior information,we present an improved piecewise random sampling method with a spectral characteristics of ＂blue noises＂.Then,the reconstruction approaches with the sparsity constraint are developed based on optimization theory.Sparse representation theory,highly efficient acquisition methods and reconstruction methods with sparsity constraint of seismic data have been developed（or improved）and the improved theory of highly efficient seismic data acquisition is established.Three highly efficient acquisition networks are designed by applying the random sampling scheme with sparsity constraint to change the regular and dense layout of sources and detectors in the conventional acquisition networks.Reconstruction procedures（including three approaches corresponding to the data on the efficient acquisition networks）are proposed which can reconstruct the seismic data in the highly efficient acquisition networks to the regular and dense acquisition networks.It satisfies the requirements of the subsequent processing in accordance with the existing conventional means.Furthermore,the precision of reconstruction,the contrast of the acquisition data′s imaging results and reconstruction data are taken to validate the reconstruction methods.The numerical examples on the Marmousi model show that the theoretical framework of highly effi

关 键 词：勘探成本 稀疏约束 分段采样 高效采集 数据重建 

分 类 号：P631[天文地球—地质矿产勘探]