High Precision Spectrum Inversion Method Analysis And Application

With oil and natural gas exploration work developing, the geological conditionbecomes more and more complex and changeable. The work we look for complex oiland gas reservoir, and fine describe underground reservoir will face more challenges.So we need to develop and improve the level of oil and gas exploration technology.Especially, high resolution inversion technology，as the main basis for exploringoil-gas reservoirs, needs more attention.To the small thickness reservoir, the conventional inversion method can not reachthe exploration precision. It is difficult to distinguish the top and bottom reflectivesurface of oil reservoirs. So we put forward a kind of method which is different fromtraditional inversion method, named high precision spectrum inversion method. Highprecision spectrum inversion method develops on the basis of spectrum decompositionnonlinear algorithm. Its main characteristic is doing seismic data spectrumdecomposition to get different frequency band information firstly, and then adjust theweighted coefficient to obtain the high resolution inversion results.The foundation of Spectrum inversion method is the spectrum decomposition.Using the spectrum decomposition, we can obtain instantaneous properties from theseismic data; these attributes can be used to improve the resolution of seismic data. Forexample, by using the relationship of the instantaneous frequency and thin layer, we canrecover the thin layer of information which is effected by tuning function.These seismicsection contains more strata information, in under the restriction of the logging data,can further improve the resolution of seismic data. In theory, using this method, throughadjusting the spectrum decomposition window, we can get any small thin layer ofinformation.In high precision spectral inversion, inversion algorithm will directly influence theinversion result. So how to choose appropriate inversion algorithm is an importantwork. As we know, the geophysics problems are most nonlinear problem. The geophysicists put forward many nonlinear algorithms to solve these problems. In orderto improve the precision of the inversion, get a more credible inversion result, wecompared the four nonlinear algorithms; finally, we choose the simulated annealingmethod as the inversion algorithm.This paper briefly analyzing why the high precision spectral inversion canimprove the resolution, deducing the objective function of the inversion, using thewavelet transform instead of the conventional time-frequency analysis algorithm, thencombined with simulated annealing method to get high precision information, and byusing the actual data inversion to prove the feasibility of this method.