Steel rail crack acoustic emission signal detection method based on improved long-short-term memory network

Abstract

The invention discloses a steel rail crack acoustic emission signal detection method based on an improved long-short-term memory network, which solves the problem of automatically filtering complex noise by training a noise model through an LSTM cyclic network in a complex noise environment of a railway site. The method mainly comprises the following steps of: respectively establishing a background noise time sequence model and a crack signal time sequence model for two consecutive times based on an LSTM network with the same structure, getting an upper envelope of error signals predicted by the background noise time sequence model and the crack signal time sequence model, subtracting to remove abnormal noise in the error signals, and finally detecting crack signal components contained innoisy signals. Compared with the prior art, the method has the advantages that (1) the cascaded time sequence model can be used for filtering abnormal noise with an unknown generation mechanism, (2) the noise model is automatically learned without any priori knowledge, and (3) the method can still effectively detect crack acoustic emission signals when the signals are completely submerged by the noise under the background of high-speed strong noise.

Description

technical field [0001] The invention relates to a method in the field of high-speed railway rail crack signal denoising and detection, in particular to a rail crack acoustic emission signal detection method based on an improved long-short-term memory network. Background technique [0002] In today's world, the development of high-speed railways benefits people's livelihood. At present, my country is vigorously promoting the construction of high-speed railway network, which not only brings convenience to people's daily travel, but also promotes the country's economic development. However, the current situation of high-speed rail lines with complex interlacing and large mileage also brings more severe challenges to the task of rail safety monitoring. With the rapid development of high-speed railway, the safety of high-speed railway has been paid more and more attention by people. In the high-speed railway system, the long-term collision and extrusion of high-speed trains are...

AI Technical Summary

Problems solved by technology

However, due to the sensitivity and passive characteristics of acoustic emission detection technology, acoustic emission acquisition signals are more easily affected by noise.

There are usually more complex noise components in the signals collected on the railway site. The background noise is mainly caused by friction and wear accompanied by the mechanical interaction between the wheels and rails, and has obvious stability and timing; sex noise

Images