A high-speed rail source seismic signal extraction method based on equidistant spectral features

Abstract

The invention discloses a high-speed rail source seismic signal extraction method based on equal interval spectrum features, which estimates the interval of the equal interval spectrum in the amplitude spectrum of the obtained high-speed rail source seismic data; , to construct a new conversion method with interval constraints to represent the high-speed rail source seismic signal more sparsely, and use the continuous wavelet transform to sparsely represent the strong background noise in the high-speed rail source seismic data, forming an over-complete dictionary; using the block coordinate relaxation algorithm , the high-speed rail source seismic signal and strong background noise in the high-speed rail source seismic data are separated and extracted, and the random white noise is filtered out, so that the extracted high-speed rail source seismic signal has a more complete structure and a higher signal-to-noise ratio.

Description

technical field [0001] The invention belongs to the technical field of seismic exploration data processing, in particular to a method for extracting high-speed rail source seismic signals based on equidistant spectral features in massive seismic data. Background technique [0002] At present, most studies are devoted to suppressing or circumventing these effects, but this will waste the huge amount of seismic data generated by the daily operation of high-speed trains. Mining these data to extract effective signals inspired by the operation of high-speed trains can not only analyze the running status of high-speed trains or serve as basic data for other subsequent applications, but also a practical application that turns waste into treasure. [0003] The existing trap method needs to know the harmonic frequency to design the trap, and multiple traps need to be designed to separate the narrow-band signal, which will also cause damage to the narrow-band high-speed rail source s...

AI Technical Summary

Problems solved by technology

[0003] The existing trap method needs to know the harmonic frequency to design the trap, and multiple traps need to be designed to separate the narrow-band signal, which will also cause damage to the narrow-band high-speed rail source seismic signal when there are many frequency bands; The existing sparse modeling of high-speed rail source seismic signals by morphological component analysis has realized the suppression of broadband background noise. Since the equal interval characteristics of the amplitude spectrum of high-speed rail source seismic signals have not been utilized, the extracted high-speed rail source The signal-to-noise ratio of the seismic signal is low, containing some background noise and random white noise

Images