Train wheelset bearing rail edge sound signal separation method based on harmonic-impact Doppler modulation compound dictionary

Abstract

The invention discloses a train wheelset bearing rail edge sound signal separation method based on a harmonic-impact Doppler modulation compound dictionary. Microphones which are arranged at two sidesof the rail and right face the train wheelset bearing are used to acquire sound signals x(t) made in the case of high-speed passing of the train. The detected signals are processed in steps: (1) an over-complete parametric Doppler modulation complex harmonic-impact compound dictionary Datom3 is built; (2) a matching pursuit algorithm is used to carry out sparse decomposition on the rail edge signals x(t) in the well-built over-complete complex compound dictionary Datom3 to obtain a projection dictionary Datom4 and a projection coefficient K; and (3) according to a bearing resonance frequencyband and the geometric position relationship between the microphones and the wheelset bearing, atoms which meet requirements are screened from the dictionary Datom4 to form a dictionary Datom5, and linear combination is carried out to obtain reconstructed fault signals sig. Better match with the fault signals in a time frequency structure are realized, better sparse representation and signal reconstruction are achieved, and the sound source separation effects are enhanced.

Description

technical field [0001] The present invention relates to the technical field of railside acoustic fault diagnosis of high-speed train wheelset bearings, in particular to a method for separating railside acoustic signals of train wheelset bearings based on a harmonic-shock Doppler modulation composite dictionary, which is used for separating trackside signals from the trackside signals The fault signal of the wheel set bearing is separated to eliminate noise and improve the accuracy of fault diagnosis. Background technique [0002] When the train is running at high speed, the sound signal from the wheel set bearing contains information closely related to its health status, and the trackside acoustic fault diagnosis has the characteristics of non-contact monitoring. However, trackside signals are mixed with sound signals from other parts of the train, wheel-rail contact sound and aerodynamic noise, etc., which brings difficulties to effective fault diagnosis. A common denoisin...

AI Technical Summary

Problems solved by technology

[0003] The technical problem to be solved in the present invention is: to overcome the deficiencies of existing technologies and methods, to provide a method for separating acoustic signals of train wheel-set bearing rails based on harmonic-shock Doppler modulation compound dictionary, which has the ability to eliminate in-band noise and It overcomes the problem that single dictionary atoms cannot achieve optimal matching of signal components, and effectively improves the denoising effect and the accuracy of fault diagnosis

Images