Research On Spectrum Rearrangement Method Of Seismic Signals Based On Hilbert-Huang Transform

In seismic data processing,data processing workers often need to adopt time-frequency analysis technology to accurately describe the characteristics of signal frequency changing with time,which is determined by the inevitable non-stationarity of seismic signal.We call this algorithm of frequency-time relationship of response signal as seismic signal spectrum decomposition technology.In actual research and production,there are higher requirements for the accuracy and resolution of spectral decomposition technology,especially in the increasing demand for onshore oil and gas development in China,and the increasing difficulty of onshore and offshore seismic exploration,the traditional and conventional spectral decomposition technology has gradually failed to meet the actual demand of this high accuracy,especially in thin layer exploration,due to the increasing demand for onshore oil and gas development in China and the increasing difficulty of onshore and offshore seismic exploration.The instability and low resolution of conventional algorithms often lead to wrong conclusions,which affect the overall exploration quality and plan.Therefore,it is of great significance to find a spectral rearrangement method with better accuracy,higher resolution and more stable algorithm at this stage.Starting from the field of conventional time-frequency analysis technology,many seismic spectral decomposition methods have been fully applied and developed.Firstly,several classical theories of time-frequency analysis of seismic signals are reviewed,such as short-time Fourier transform,Gabor transform,S transform,continuous wavelet transform,Wigner transform and matching pursuit algorithm.Traditional time-frequency analysis algorithms are limited by low accuracy and resolution,and often have some defects in dealing with practical problems.As a new algorithm,the improved Hilbert-Huang algorithm proposed in this paper not only has obvious advantages over other algorithms in accuracy,focusing and resolution,but also has significant improvement in computational efficiency.In order to alleviate the shortcomings of conventional TIME-SPECTRUM decomposition methods,such as low time resolution and frequency resolution,the general idea of the new Hilbert transform is to decompose the signal statically at first.By static decomposition,the non-stationary and non-linear seismic signal is decomposed into a finite number of intrinsic mode functions and a residual term.Finally,each intrinsic mode function is processed by Hilbert transform.By completing the above steps,we can get the instantaneous frequency of seismic signal,and then we can get the time,frequency and amplitude of the three-dimensional seismic time spectrum.In this paper,the Hilbert-Huang transform algorithm is elaborated and deduced in detail.After the improvement of the algorithm,it is applied to the actual seismic data processing.The actual test shows that the improved algorithm has significantly improved the accuracy,resolution and focusing compared with the conventional algorithm,and the computational load has also been compressed.The computational efficiency can reach 2-3 times of the past.On this basis,the effect of spectral rearrangement of seismic signals improved by Hilbert-Huang transform algorithm is discussed in this paper.As a result,the algorithm steps,principles and effects of Gabor transform,one of the traditional time-frequency analysis techniques,are fully compared,and applied to the processing of actual seismic signals.The imaging results of fitting data and actual data show that the results of spectral rearrangement of seismic signals based on Hilbert-Huang transform can better describe the nonlinearity and non-stationarity of seismic waves.The calculation process is simpler,more efficient,with strong adaptability,high accuracy and resolution,and good focusing.Compared with conventional algorithms,Hilbert-Huang algorithm is more suitable for solving the problem of non-linear signals such as seismic signals.Finally,we analyse and process the actual oil and gas reservoir data with time-frequency analysis,and the results show that the improved spectral rearrangement algorithm has better time-frequency resolution and timefrequency focusing effect and higher accuracy;it can better reflect the specific horizons of hydrocarbons in oil and gas reservoirs,and it is more precise and more obvious when dealing with thin reservoirs.