Font Size: a A A

Research On The Spectral Inversion Methods Of Seismic Signal

Posted on:2016-06-25Degree:MasterType:Thesis
Country:ChinaCandidate:M ZhaoFull Text:PDF
GTID:2180330473459732Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
With further exploration and development of oil-gas reservoirs, problem of thin layers has become increasingly prominent. Improving identification and prediction accuracy of thin layers has become an urgent problem to which geologists pay close attention. Based on high-resolution frequency technology, seismic spectral inversion provides a new way to improve the resolution of seismic data. As a result, prediction accuracy of thin layers is improved. For the past few years, along with the continuous deepening study of spectral inversion conducted by domestic and foreign scholars, spectral inversion to solve the prediction of thin layers becomes a hot issue. However, it is still a short period since the research began, there is a series of technical problems didn’t reach consensus.This paper carries out a series of studies about seismic spectral inversion method, including further study of spectral inversion theory. A set of relatively complete process of inversion algorithm is established. Main work includes:1. Theory of seismic wavelet is studied. The influence of wavelet extraction methods on spectral inversion algorithm and inversion results is discussed. Among them the Ricker wavelet used as seismic wavelet for inversion and the extraction method of statistical wavelet are mainly studied, and the flowchart of wavelet extraction is shown at the same time.2. Representation of seismic record in the time domain is studied. The flowchart of seismic data’s Fourier transform is summarized and the method of adding time window function and inversion step in Fourier transform is introduced at the same time.3. Influence of the odd and even parts of reflection coefficients on the resolution of inversion results is studied and the decomposition method of reflection coefficient for odd and even parts is illustrated.4. Building process of the objective function of spectral inversion is derived in detail and the classic method for solving the objective function- Conjugate Gradient Method is introduced.5. The optimization methods of spectral inversion are researched mainly. Separable surrogate functionals(SSF) method and orthogonal matching pursuit(OMP) algorithm are studied, the related process and simulation examples are listed out, and the analysis of how to set values of important parameters in inversion is done.6. Coding of all algorithm modules is finished, simulation tests for several kinds of typical geological model are done. Field data from different work areas are inversed, and the seismic inversion results of the two kinds of inversion algorithm are analyzed and compared.
Keywords/Search Tags:reflectivity inversion, SSF, OMP, wavelet extraction, seismic frequency spectrum analysis
PDF Full Text Request
Related items