A Seismic Wave Noise Reduction Method Based on the Combination of Adaptive Filtering and Wavelet Transform

Abstract

The invention provides a seismic wave noise reduction method based on combination of adaptive filtering and wavelet transform. The method settles problems of extracting an effective signal which facilitates analysis and interpretation from a mixed signal, compressing noise, highlighting an effective signal, eliminating an inference signal as possible and realizing separation between the signal and the noise. The method comprises the steps of firstly finding a useful signal in a time delay estimating method, wherein the useful signal still comprises the noise in partial frequency band; and then eliminating the partial noise which is superposed with the frequency spectrum of the useful signal in a wavelet de-noising method, thereby realizing a fine de-noising purpose and finally obtaining clean and effective seismic wave data. A true condition of an underground geological structure can be reflected. Furthermore the geological condition can be correctly and reasonably tested.

Description

technical field [0001] The invention relates to a seismic wave noise reduction method based on adaptive filtering and wavelet transform, and belongs to the technical field of seismic wave data processing. Background technique [0002] Geological disasters such as subgrade collapse and ground fracture pose a greater threat to people's lives. Therefore, it is very important to carry out regular safety geological inspections on the built projects to prevent disasters in the geological substructure of subgrades and bridge foundations. At the same time, as the country's investment in major infrastructures increases, a large number of geological exploration work needs to be carried out. In addition, my country belongs to the high-incidence and mass-occurrence areas of geological disasters, which provides a broad market for the research and development and application of engineering seismograph equipment. [0003] In the process of using the geophone to collect seismic wave data, ...

AI Technical Summary

Problems solved by technology

Most of these noises are Gaussian white noise, and its spectrum is distributed in the entire frequency domain, which must overlap with the spectrum of useful signals. Here, frequency-based filtering methods cannot be used for noise reduction.

[0004] In the prior art, the methods for filtering seismic waves mainly include: (1) Inverse Q filtering method: simulate the reverse process of forward propagation of seismic waves, and eliminate the amplitude attenuation and phase distortion of seismic waves during propagation by supplementing the amplitude, but in In the compensation process, as the travel time and frequency increase, the amplitude compensation term becomes larger and larger, that is, the noise is amplified in the compensation process

(2) Inverse filtering, also called deconvolution, is an inverse process of filtering, which compresses the seismic wavelet into the source pulse shape to form an ideal seismic record to offset the effect of ground filtering; the commonly used deconvolution algorithm is pulse Deconvolution, but pulse-pulse deconvolution cannot distinguish signal and noise. Considering the total spectrum, the vertical resolution can be significantly improved, but the signal-to-noise ratio will be reduced while improving the resolution.

(3) Short-time Fourier transform. The resolution of short-time Fourier transform in the full-time and frequency domain is consistent, and it has the characteristics of single resolution, but it requires a very large amount of calculation and low calculation efficiency.

Images