Research And Application Of Maximum Entropy Spectrum Method For Weak Signal Recognition In Reflection Seismic

Seismic exploration is an important method for studying deep underground geological conditions and structural conditions in geophysical methods.At present,with the increasing demand for mineral,oil,natural gas and other resources,the degree of exploration is getting higher,and the difficulty of geological exploration work continues to increase,and the imaging accuracy requirements for seismic exploration are getting higher and higher.How to continuously enhance the ability of seismic exploration methods to detect deep earth resources is a serious problem.The classical power spectrum(Fourier spectrum)calculation method uses finite-length observation signals,and some specific assumptions are made on the original time series,such as the implicit assumption that the original signal is filled with zero outside the observation signal window in the autocorrelation method.These assumptions will cause spectrum leakage reducing the resolution,making some weak signals missing,and affecting the effect of seismic signal processing.Therefore,we propose to use the maximum entropy spectrum method.The biggest advantage of the maximum entropy spectrum method is that it has higher resolution than the Fourier spectrum,and it is especially suitable for power spectrum calculation of short time series.The maximum entropy spectrum method does not need to make any specific assumptions on the original time series,and the part outside the time window,the maximum uncertainty of the original signal is guaranteed,so the power spectrum obtained is more reflective of the real power spectrum of the signal.In the process of seismic signal processing,the maximum entropy spectrum method can be used to accurately identify the signal,especially in the weak signal.Therefore,based on the in-depth study of the previous application of maximum entropy spectrum method to process one-dimensional signal data,this paper uses MATLAB software to write the program of maximum entropy spectrum method to process one-dimensional and two-dimensional seismic data.Using this program to process and analyze a variety of data signals,we get the conclusion that the calculation results of the maximum entropy spectrum method are more stable and the resolution is higher.In this paper,we usedGeoeast software to process the 2D seismic data in the Jinding area and obtained the final profile,and then applied the maximum entropy spectrum method to process the above seismic data.Comparedmaximum entropy spectrum method result with conventional processing result,some weak signals are more pronounced,which for the seismic interpretation provides a more accurate seismic profile.