The Method Study To Stratigraphic Absorption Compensation Based On The Fast Matching Pursuit

With the increasing difficulty of exploration,the exploration focus is changing from simple structural reservoir to complex lithologic reservoir,the exploration targets are changing from the shallow large structure to the deep small structure,therefore,the higher and higher requirements are proposed for seismic data.It is known that seismic wave propagation in inelastic underground medium will suffer absorption,which is showed as amplitude attenuation and waveform distortion,reducing the resolution of seismic data and the middle and deep seismic data are especially significant.Now,the widely application method of the seismic data absorption compensation,which is effect on the accuracy of calculating quality factor and the stability of compensation method,can not compensation absorbtion effectively.Considering this,how to compensate the attenuation of seismic data accurately and stably becomes one of the urgent problems.The matching pursuit algotithm,which is a adaptive signal decomposition and reconstruction method,can match decomposition nonstationary signal such as seismic signal and it can realize seismic signal fine characterization that have great value to study the localization in the frequency domain and time domain.Unfortunately,the enormous calculation of the current matching pursuit algorithm restricts its wide range of applications.On the basis of the complex two-parameter fast matching pursuit algorithm,we improve the calculation method of the instantaneous frequency and resolve the problem of the negative instantaneous frequency presence by phase unwrapping.So,the instantaneous frequency can be used in matching pursuit directly,which further improves the algorithm computational efficiency.Through analysis the theory of seismic wave absorption and attenuation,we can see that the absorption and attenuation is related to the product of time and frequency.Considering this,this paper proposed an effective method for absorption compensation based on matching pursuit algorithm.In our method,seismic data are firstly decomposed into a variety of matching wavelets by utilizing of a modified version of matching pursuit algorithm.After that,a scatter-plot is provided by taking the product of the travel time and central frequency of each wavelet to be the x-axis and its corresponding amplitude to be the y-axis.These scatters are then divided into different intervals according to the products of travel times and central frequencies.The amplitude attenuation curve,which is changing with the products of travel times and central frequencies,can be obtained by calculating the mean of the amplitudes of matching wavelets in each interval.Secondly,we can obtain the amplitude gain curve by deal with the amplitude attenuation curve.The amplitude of matching wavelet within its corresponding interval can then be compensated by this amplitude gain curve.Finally,the seismic signal is reconstructed by these compensated matching wavelets.By utilizing dimensionality reduction method,the robustness of our algorithm has been improved,as the information in two dimension merged into one,which is the product of time and frequency.Compared with those absorption compensation methods,the most advantage of our method is that the value of quality factor is not necessary any more.