Quantitative Diagnosis Method of Locomotive Wheelset Bearing Based on Adaptive Filter Demodulation

Abstract

The disclosure discloses a locomotive wheelset bearing quantitative diagnosis method based on adaptive filtering demodulation. The locomotive wheelset bearing quantitative diagnosis method based on adaptive filtering demodulation comprises two stages of parameter training and real-time diagnosis, wherein in the parameter training stage, three filter parameters of a length, a step length and a frequency band of a filter are traversed, optimal extraction of fault impact features is implemented by iterative loop, optical filter parameters and an optimal envelope spectrum are adaptively determinedso as to calculate a fault degree index FSI, a corresponding relationship between a fault degree and the fault symptom index FSI is established and a threshold is determined; and in the real-time diagnosis stage, the optimal filter parameters and the threshold which are obtained in the parameter training stage are adopted, a wheelset bearing signal acquired in real time is filtered, the optimal envelope spectrum is solved, and a size relationship between the current fault degree index and the threshold is compared, so as to implement fault alarm and quantitative diagnosis for a wheelset bearing.

Description

technical field [0001] The present disclosure relates to a method for quantitative diagnosis of locomotive wheel set bearings, in particular to a method for quantitative diagnosis of wheel set bearings based on adaptive filter demodulation. Background technique [0002] The locomotive wheel set bearing is one of the core components of the locomotive, and its performance status directly affects the reliable operation of the locomotive. It is of great significance to accurately and timely identify the initiation and evolution of faults during the operation of bearings, and quantitatively diagnose the fault degree of bearings to ensure the safe operation of locomotives and avoid economic losses and catastrophic accidents. [0003] However, the vibration acceleration signal of the locomotive wheel set bearing cannot be obtained directly from the locomotive wheel set bearing, usually it is obtained from the wheel set bearing housing or the axle box. During the acquisition process...

AI Technical Summary

Problems solved by technology

[0003] However, the vibration acceleration signal of the locomotive wheel set bearing cannot be obtained directly from the locomotive wheel set bearing, usually it is obtained from the wheel set bearing housing or the axle box. During the acquisition process, the vibration acceleration signal needs to go through a certain transmission path. Therefore, The obtained vibration acceleration signal is composed of the vibration of the wheel set bearing and the vibration of other matching parts, which is bound to be mixed with a lot of noise and unnecessary vibration signals

Although conventional filtering methods and modern signal processing methods can reduce the influence of white noise to a certain extent, they cannot remove the vibration interference of parts adjacent to the locomotive wheel pair bearings, thereby reducing the accuracy of diagnosis

In addition, because the existing failure degree indicators are easily affected by external working conditions, they cannot effectively reflect the failure degree of the actual wheel set bearings, resulting in the inability to carry out quantitative diagnosis of wheel set bearing failures in the early stage, resulting in great safety hazards

Images