Font Size: a A A

Study And Integration Of Tomographic Static Correction Algorithm Based On First Break Data

Posted on:2012-01-17Degree:MasterType:Thesis
Country:ChinaCandidate:Q L MengFull Text:PDF
GTID:2120330338493417Subject:Earth Exploration and Information Technology
Abstract/Summary:PDF Full Text Request
Static correction greatly influences the results of seismic exploration.If the problem ofstatic correction isn't solved correctly, we may get the wrong image of the seismic section,and this will affect the work of interpretation and inversion, leading to the wrong recognitionof the structure underground. Currently, most of oil exploration works in our country areconducted in the complicated areas. And in these areas, the surface is very complex, thevelocity and thickness of the low-velocity zone isn't uniform, these conditions can not meetthe traditional static correction methods. But the method of tomographic static correctiondoesn't require some hypothesis, so this method can solve the static problems in complicatedareas perfectly, and it has became the current trend of static correction.Aiming to acquire practical software of static correction, this paper analyzed and studiedevery related parts of this area: the method of picking the first breaking time of seismic data,the method of ray tracing of first arriving wave, the method of tomographic inversion and themethod of static correction. Then, this paper selected the proper method of every part andimproved them in order to achieve the requirement of tomographic static correction. In thepart of first breaking time picking, this paper mainly studied and improved the method ofBack-propagation Artificial Neural Networks, because this method can perfectly deal with thelow signal noise ratio seismic data. As for ray tracing method, this paper selected andimproved the method about high precision fast ray tracing with multi-retracing in order toraise the speed and precision of this method. In the aspect of inversion method, aftersummarizing and comparing several common methods, this paper chooses the LSQR as theinversion method in the tomographic static correction. LSQR method has lots of advantagessuch as fast converging speed, low memory occupancy, high inversion precision and so on.Lastly, we integrated all the main algorithms using the technology of Qt, and got thesoftware of first arriving wave tomography. In this software, we have designed severalpractical modules, such as the loading and editing of first breaking time, building of initialvelocity model, ray tracing, tomographic inversion, QC and so on. This software can not onlyprovide the statics but also the precise near-surface velocity model, not the equivalent model.
Keywords/Search Tags:Tomographic static correction, Back-propagation Artificial Neural Networks, Ray tracing, LSQR inversion, Technology of Qt
PDF Full Text Request
Related items