CN-116337492-B - Vehicle-mounted vibration-based wheel non-circularization anti-interference detection method

Abstract

The invention discloses a vehicle-mounted vibration-based wheel non-circularization anti-interference detection method which comprises the following steps of obtaining axle box vibration acceleration signals and wheel set rotating speed signals, carrying out equiangular resampling on the axle box vibration acceleration signals according to the wheel set rotating speed signals to obtain sampling data, calculating and calculating a weighted moving average value according to weight vectors of the sampling data, carrying out frequency domain analysis processing on the weighted moving average value to obtain and judge whether a peak value exceeds a set value according to a frequency spectrum or a time spectrum, if so, a periodic wheel non-circularization defect exists, otherwise, entering the next step, judging whether the peak value of the frequency spectrum exceeds the set value when an envelope spectrum or an envelope exists, if so, carrying out local wheel non-circularization defect, otherwise, judging that the wheel state is normal. The method solves the problem of low-order resolution caused by cycle number loss in the traditional time synchronization average method, reduces the influence of abnormal impact, and avoids amplitude loss caused by accumulated phase error.

Inventors

• XU WENTIAN
• LIANG SHULIN
• CHI MAORU
• CAI WUBIN
• Tao Gongquan

Assignees

• 西南交通大学

Dates

Publication Date

20260512

Application Date

20230331

Claims (3)

1. 1. The vehicle-mounted vibration-based wheel non-rounding anti-interference detection method is characterized by comprising the following steps of: s1, acquiring an axle box vibration acceleration signal And wheel set rotational speed signal ; S2, according to the rotating speed signal of the wheel set Vibration acceleration signal for axle box Carrying out equal-angle resampling to obtain sampling data; S3, calculating a weight vector of the sampling data; S4, calculating a weighted moving average value according to the weight vector; S5, carrying out frequency domain analysis processing on the weighted moving average to obtain a frequency spectrum or a time frequency spectrum containing the periodical non-circular order and amplitude characteristics of the wheel and an envelope spectrum or an envelope time frequency spectrum containing the local non-circular abrasion information of the wheel; S6, judging whether the peak value of the frequency spectrum containing the periodical non-circular order and amplitude characteristics of the wheel exceeds a set value, if so, the periodical non-circular defect of the wheel, namely the polygonal defect of the wheel exists, otherwise, entering a step S7; S7, judging whether an envelope spectrum containing the local non-circular abrasion information of the wheel or a peak value of the spectrum during envelope exceeds a set value, if so, the local non-circular abrasion information of the wheel is in flat or scratch, otherwise, the wheel is in a normal state; The specific implementation manner of the step S3 is as follows: S3-1, according to the formula: Obtaining the elements of the ith row and jth column of the bias matrix B Wherein, the method comprises the steps of, For the ith data point in the moving window of equiangular resampling for one sample interval time, ; Is in combination with Kth data point within the same window, wherein ; Pi represents the circumference ratio; The axle box vibration acceleration after equal-angle resampling is obtained; S3-2, according to the formula: Obtaining the element of the c row and the d column of the weight matrix W ; S3-3, obtaining a corresponding weight vector according to the weight matrix W ; The specific implementation manner of the step S4 is as follows: According to the formula: Obtaining a weighted moving average of sampling points within a sampling interval time Wherein Q is the number of turns of the wheel rotation, The number of turns of the wheel rotation of the movable window which is resampled at equal angles in the time of one sampling interval; Is the mth data point in the moving window of equiangular resampling for one sampling interval time.
2. 2. The vehicle-mounted vibration-based wheel non-circular anti-interference detection method according to claim 1, wherein the specific mode of the step S2 is as follows: According to the formula: Obtaining axle box vibration acceleration after equal-angle resampling Wherein, the method comprises the steps of, L is the total number of angle samples; The vibration acceleration signal is an axle box vibration acceleration signal; N equal time sampling points in total; for equal sampling time intervals; interpolate (·) is the interpolation function; As a function of phase.
3. 3. The vehicle-mounted vibration-based wheel non-circular anti-interference detection method according to claim 2, wherein the specific implementation manner of the step S5 is as follows: S5-1, according to the formula: Obtaining a spectrum comprising cyclic non-rounded order and amplitude features of a wheel Either (or) or (b) According to the formula: obtaining a time spectrum comprising cyclic non-rounded order and amplitude features of a wheel Wherein e is a natural constant; A weighted moving average of the data in the angular domain; = ; j represents a unit imaginary number; s5-2, according to the formula: Obtaining an analytic signal Wherein, the method comprises the steps of, Representing a Hilbert transform; s5-3, according to the formula: Obtaining the envelope ; S5-4, according to the formula: Obtaining an envelope spectrum containing localized non-circular wear information for a wheel ; S5-5, according to the formula: Obtaining an envelope time spectrum containing localized non-circular wear information for a wheel 。

Description

Vehicle-mounted vibration-based wheel non-circularization anti-interference detection method Technical Field The invention relates to the technical field of out-of-roundness detection, in particular to a vehicle-mounted vibration-based wheel non-roundness anti-interference detection method. Background The non-circularization problem of the wheel is the non-uniform abrasion and damage problem along the circumference and the longitudinal direction of the wheel. Including periodic wear of the wheel circumference, i.e. the wheel polygon, and also localized non-circularization problems represented by wheel flats, scratches, etc. Severe wheel non-rounding can lead to severe vibration, noise, and structural safety problems, and the initial polygon, once formed, develops relatively quickly, so wheel non-rounding is a form of failure of high concern for railway operator doors. For this reason, various means have been developed to examine it, which can be largely classified into trackside and car inspection technologies. The vehicle-mounted technology has been widely used in the bogie on-line monitoring system with the advantages of direct path, high signal-to-noise ratio, long-term tracking and the like. However, the current vehicle-mounted detection mainly aims at threshold clamping control under the condition of severe middle and late vibration. Considering that vibration and noise of non-circularization of the middle and late wheels can cause damage, the method has very important significance for on-line monitoring of the non-circularization of the wheels in an initial state. In recent years, many scholars have studied various vibration-based methods of diagnosing non-circular wear of wheels. In terms of signal processing methods, empirical mode decomposition and its modification methods are widely used to consider the modulated nonlinear characteristics of the wheel response, including EMD, ACMD, VMD methods, HHT, and the like. In the aspect of quantitative identification of the non-circular wear of the wheel, the inertial integration method researches quantitative diagnosis of the non-circular wear of the wheel, but structural resonance which has great influence on the quantitative diagnosis is not considered. As a direct acting object of wheel rail excitation, the diagnosis method based on axle box vibration acceleration is easy to be interfered by factors such as rail shortwave irregularity, wheel rotation speed fluctuation, structure self resonance and the like. In particular to rail wave grinding, which is a harmonic excitation on rails, the vibration characteristics of the rail wave grinding are very similar to the abrasion of the wheel polygon, so that misjudgment is very easy to occur with the wheel polygon. For the diagnosis of early wheel non-circular wear, the influence of these disturbing factors is more troublesome, which makes it difficult to accurately detect early wheel non-circular wear by the axle box vibration acceleration. Disclosure of Invention Aiming at the defects in the prior art, the vehicle-mounted vibration-based wheel non-circular anti-interference detection method provided by the invention solves the problems that the early wheel non-circular characteristic is weak, random noise is easy to submerge due to rail roughness, structural resonance and the like, interference frequency generated by track equal interval impact or other rotating equipment is easy to be mistakenly considered as wheel non-circular, fluctuation of rotating speed can cause non-linear modulation of wheel non-circular response, and frequency spectrum leakage and blurring in frequency spectrum characteristics are caused. In order to achieve the aim of the invention, the technical scheme adopted by the invention is that the vehicle-mounted vibration-based wheel non-rounding anti-interference detection method comprises the following steps: s1, acquiring an axle box vibration acceleration signal x (t n) and a wheel set rotating speed signal S2, according to the rotating speed signal of the wheel setThe axle box vibration acceleration signal x (t n) is subjected to equal-angle resampling to obtain sampling data; S3, calculating a weight vector of the sampling data; S4, calculating a weighted moving average value according to the weight vector; S5, carrying out frequency domain analysis processing on the weighted moving average to obtain a frequency spectrum or a time frequency spectrum containing the periodical non-circular order and amplitude characteristics of the wheel and an envelope spectrum or an envelope time frequency spectrum containing the local non-circular abrasion information of the wheel; S6, judging whether the peak value of the frequency spectrum containing the periodical non-circular order and amplitude characteristics of the wheel exceeds a set value, if so, the periodical non-circular defect of the wheel, namely the polygonal defect of the wheel exists, otherwise, entering a step S7; S7, judging wheth