Font Size: a A A

Research On Multiple Suppression Method In Curvelet Domain Based On Compressed Sensing

Posted on:2021-04-29Degree:MasterType:Thesis
Country:ChinaCandidate:Y L ZhaoFull Text:PDF
GTID:2530307109460084Subject:Geological engineering
Abstract/Summary:
With the increasing complexity of exploration target area and the rising of offshore exploration,multiple suppression has become a strong and difficult problems to be solved.Multiple suppression based on the theory of signal filtering methods often cannot meet the requirements of seismic exploration,and multiple wave suppression method based on wave theory considering the kinematic and dynamic characteristics of the multiple wave propagation,and almost don’t need any priori information can be processed,not only avoided some of the limitations of suppression method based on signal domain,and can obtain good suppression effect,so showed its advantages,has the very good development space and application prospect.Based on the SRME(Surface-Releated Mutiple Elimination)multiple suppression method based on wave theory,this paper divides the multiple suppression into two stages:prediction and subtraction.Aiming at the defects of SRME method,the prediction and adaptive subtraction stages are improved by using the theory of curve transform and compressed sensing.Using multiple wave model prediction of SRME prediction stage,but SRME method need full wave field data,and data loss will have a big impact for multiple model prediction,aimed at the defects of the missing data by compressed sensing theory reconstruction,and using the model test data,and aimed at the lack of different levels of refactorings,comparing before and after the reconstruction of multiple forecast models.In adaptive subtraction stage,due to the conventional adaptive subtraction based on least squares to suppress multiple effect is not good and also damage the effective signals,so this article use curvelet domain of a wave and multiple separation method of curvelet transform has a unique superiority in the expression of seismic signal,the multiple wave and a wavelet transform to the curvelet domain,in the characterization of different scales and different directions in different ways,through a nonlinear optimization method in the original data can be directly separated respectively,a wave and multiple so that differs from ordinary subtractive method,do less damage of effective signal.In order to suppress multiple waves and recover effective signals,this paper introduces the method of complex curviline transformation and combines it with the theory of compressed sensing into the method of curviline domain separation.It has both the advantages of compressed sensing theory and curviline transformation,and finally obtains a set of flow of multiple suppression.Finally,through the processing of model and actual data,the validity and practicability of the method are proved.The research of multiple wave suppression method needs the efforts and efforts of more geophysical explorers.It is believed that with the increase of the difficulty of exploration and the development of computer technology,the existing multiple wave suppression technology will be gradually improved,and new theories and methods will also be born to better serve the seismic data processing.
Keywords/Search Tags:SRME, adaptive subtraction, curvelet, compress sensing, complex-curvelet
Related items