Research On Identification Of Groove Rail Joint Status Based On Tramcar Vibration Signal

In recent years,the tram as a new mode of travel for urban rail transit is developing rapidly.The tram line is a seamless line.Welding joints may have joint defects due to geometric non-uniformity and material non-uniformity.With the continuous growth of line operation mileage,the continuous increase in tram traffic,and the continuous increase in operating speed,huge impact loads will be generated between the wheels and rails,leading to the continuous development of joint diseases,which in turn will exacerbate the effect of wheel and rail forces.Seriously affect the safety and stability of tram operation.Therefore,it is necessary to grasp the service status of the rail joints in real time to ensure the safe operation of the tram.This article focuses on the research of grooved rail high joint diseases caused by irregular welds.The vertical acceleration signal of the tram at the high joint is collected by the rail condition detection system to construct a sample set of high joint disease signals to ensure subsequent quantitative diagnosis and characteristics The effectiveness of extraction,using zero-averaging,wavelet soft threshold denoising and other preprocessing methods to optimize the signal,and propose joint impact indicators,based on the Gaussian pyramid filtered sample set to develop high joint damage evaluation criteria,and then apply time domain analysis,Frequency domain analysis,short-time Fourier transform,wavelet packet analysis,Hilbert yellow transform and edge detection analyze the high joint signal and extract multi-dimensional features,and finally perform online diagnosis of the rail joint status on the line to achieve The identification of the joint type,location and damage degree is proved,which proves the feasibility and reliability of identifying the state of the grooved rail joint based on the vibration signal of the tram.The specific research work of this article is as follows:(1)Establishment of rail joint condition detection system.Based on the detection principle,the design of the detection system program is designed.The system is composed of sensors,data acquisition devices,host computers,and other structures.The detection parameters are determined according to the characteristics of the groove rail joint disease,which provides a data basis for subsequent signal analysis.(2)The establishment of the evaluation standard for the degree of damage to the rail joint disease.The vertical acceleration signal of the tram at the high junction of the whole line is used as a sample set,and the pre-processing methods of eliminating trend terms,zero averaging and wavelet soft threshold denoising are used to make the signal easy to analyze,and the randomness of the signal is reduced by Gaussian pyramid bandpass filtering,Statistical analysis of the joint impact indicators of each high-joint signal,and the evaluation criteria for the degree of damage were formulated based on the results,and quantitative diagnosis of high-joint diseases was realized.(3)Feature extraction of rail joint disease signal.Signal processing methods such as time domain analysis,frequency domain analysis,short-time Fourier transform,wavelet packet analysis,Hilbert-Huang Transform,etc.are used to analyze and compare the normal signal of the rail and the high joint signal,and the high joint is extracted The multi-dimensional features of the disease not only improve the accuracy of diagnosis,but also enhance the robustness of identification.(4)Online identification of line connector status.Analyzed the field detection data of a section in a city’s tram line,and based on the high joint damage assessment criteria and disease characteristics,systematically identified the status of the joints on the line and determined the type,location and damage degree of the joints.The recognition result can also guide the line maintenance and repair work.