Research On Separation Of Multi-source And Multi-wavefield Seismic Signal Based On Multi-domain Denoising

Seismic data acquisition is a very important part of geophysical exploration.3D high-density seismic acquisition is often time-consuming and costly.Therefore,with the gradual deepening of oil and gas exploration and development,improving efficiency is very important for oil and gas exploration.Based on this,a multi-source seismic data acquisition method was proposed to make up for the inefficiency of traditional seismic acquisition technology.Multi-source seismic acquisition adopts simultaneous excitation of multiple seismic sources(or excitation at a certain delay)and synchronous reception,which can greatly reduce the acquisition time and reduce the acquisition cost.Although multi-source seismic data acquisition improves the acquisition efficiency,it also brings serious aliasing noise to the seismic data,which makes it impossible to directly process seismic data.Therefore,before seismic data processing,it is necessary to perform wavefield separation on multi-source seismic data to obtain relatively clean single-shot seismic records.In order to better solve the noise problem caused by multi-source seismic data,this thesis proposes to use an improved S-transform domain multi-source seismic signal separation method for multi-source seismic data separation.First of all,the theoretical analysis of multi-source seismic data acquisition is carried out from the perspective of mathematics and physics.Secondly,the aliasing noise and random noise in the multi-source seismic data are pseudo-separated to eliminate the influence of delay on the mixed seismic source data.Then,performing Normal Moveout correction on the pseudo-separated seismic data and converting the gather data after Normal Moveout correction to the S transform domain.Finally,an improved S-transform method is used to design an adaptive filter to denoise the aliasing noise,and the effective signal is continuously enhanced through multiple loop iterations until a relatively clean single shot seismic record is obtained.The theoretical and actual data trial calculations show that the multi-source seismic signal separation method based on the improved S-transform domain studied in this thesis can perform amplitude-preserving separation of multi-source seismic data,and obtain multiples of seismic data similar to traditional seismic acquisition.Greatly improving the acquisition efficiency and providing technical support for the separation and processing of three-dimensional high-density multi-source seismic data.