Abstract:

Spectral decomposition technique as an interpretative processing technology has been paid extensive attention by many geological and geophysical researchers. The effectively frequency division display function of seismic signals for oil gas detection has also become a focus point. The authors propose a spectrum decomposition technique of Wigner bispectrum diagonal slice with an improved kernel function based on high time-frequency resolution of the Wigner higher order spectrum. According to the advantages and disadvantages of the exponential and the cone-shaped kernel functions, the new improved kernel function is given. Then the kernel function filtering method in fuzzy fields is used to suppress the cross-terms of Wigner bispectrum diagonal slice and the center point of the ambiguity function in Wigner bispectrum diagonal slice is corrected. The numerical simulation shows that the new function can suppress the cross-terms of time-delay axis. Finally, the effectiveness of  new technique is verified by indicating the presence of oil and gas area in 20 Hz profile, and in accord with well.

Key words: Wigner bispectrum diagonal slice, cross-term, kernel function, spectrum decomposition, oil and gas detection