TWI624651B - Train wheel anomaly detection system and method thereof - Google Patents

Abstract

The invention provides a train wheel abnormality detecting system, comprising: a sensor disposed on a track, the sensor picking up a wheel of the train to vibrate an acceleration signal through one of the tracks; and an analysis controller, the sensor and the sensor The signal is connected and receives the vibration acceleration signal. The analysis controller has an initial analysis module, a feature analysis module and a determination module. The initial analysis module analyzes the vibration acceleration signal according to the vibration acceleration signal, and the characteristic analysis module vibrates. The speed signal obtains a time interval and a vibration source signal through one of the tracks passing through the track, and the characteristic analysis module analyzes the vibration amount according to the time interval and the vibration source signal, and determines the vibration amount of the module to be compared with a basic ratio. Compare the values and determine if the amount of vibration exceeds the base comparison value.

Description

Train wheel abnormality detecting system and method

The present invention relates to a detection system and method, and more particularly to a train wheel anomaly detection system and method.

In the transportation system, trains including trams, MRT and high-speed rails play an important role. When the train is on the track, the wheels of the train will frequently rub against the track and cause damage and deformation of the wheels.

As the contour of the wheel changes, additional impact vibrations are generated on the moving train body and the track, which seriously affects traffic safety and the service life of the train and track.

Generally, the way to detect whether the wheel is abnormal is mostly by manual detection. The manual detection method needs to be based on the experience of the maintenance personnel. As a benchmark for judging whether the wheel is abnormal, the detection method cannot effectively find the abnormal state of the wheel. If the number of inspectors is insufficient, it is impossible to detect frequently, or when the experience is insufficient, the abnormal problem cannot be effectively judged, and there is a big worry about the safety of the train.

Furthermore, the Republic of China Patent Publication No. 201533439 discloses a system and method for detecting abnormality of a train wheel during driving, which can only detect whether an abnormality occurs in each wheel, but cannot detect the corresponding car or wheel set, and whether it also occurs. For abnormal problems, it is necessary to manually detect whether the wheel or the car is abnormal when the wheel is abnormal. It is relatively complicated and time consuming to detect.

Therefore, how to improve the shortcomings and uncertainties of the aforementioned detections and provide a more effective train wheel detection method is to be solved by relevant industry players.

In order to solve the above problems, the present invention provides a train wheel abnormality detecting system and method, which can take a signal generated by a sensor when a wheel passes through a track, and analyze the signal through an analysis controller to determine whether the train is abnormal, thereby enabling Effectively and quickly detect the problem of train wheel anomalies and increase the driving safety of the train.

An embodiment of the present invention provides a train wheel abnormality detecting system, comprising: a sensor disposed on a track, the sensor picking up a wheel of the train to vibrate an acceleration signal through one of the tracks; and an analysis controller, The signal is connected to the sensor signal and receives the vibration acceleration signal. The analysis controller has an initial analysis module, a feature analysis module and a determination module. The initial analysis module analyzes the vibration acceleration signal into a vibration speed signal according to the vibration acceleration signal. The analysis module analyzes the vibration speed signal by a frequency distribution to obtain the vibration source signal of one of the train passing track and the corresponding passage time interval, and the characteristic analysis module analyzes the vibration according to the passage time interval and the vibration source signal. The quantity judgment module compares the vibration amount with a basic comparison value to determine whether the vibration amount exceeds the basic comparison value.

An embodiment of the present invention provides a method for detecting an abnormality of a train wheel, comprising the steps of: arranging a sensor on a track, and extracting a vibration acceleration signal by a sensor when the wheel of the train passes the track; and accelerating the vibration The signal is transmitted to an analysis controller; the initial analysis module of the analysis controller is analyzed as a vibration speed signal according to the vibration acceleration signal conversion; and the characteristic analysis module of the analysis controller obtains the train by a frequency distribution analysis according to the vibration speed signal. The wheel passes through one of the tracks through the time interval and the corresponding time interval of the vibration source signal; the feature analysis module analyzes a vibration amount according to the time interval and the vibration source signal; and one of the analysis controllers determines the vibration amount of the module Compared with a basic comparison value, it is judged whether the vibration amount exceeds the basic comparison value, and when the vibration amount is greater than the basic comparison value, an abnormal signal is issued.

According to the above, the present invention can achieve the effect, and the train wheel abnormality detecting system and method can detect and analyze the wheel state of the train, and can accurately determine which wheel of the train is abnormal, and the degree of abnormality.

Furthermore, the present invention can be more accurate than the conventional human detection method, and it can be analyzed by the analysis controller that each train includes several cars, wheels, and left and right wheels of each wheel set, and It is possible to use the vibration amount to compare with the reference value to accurately determine whether an abnormality has occurred in each car, each wheel group and each wheel. If an abnormality occurs, the abnormality can be quickly found to effectively improve the shortcomings and defects of the manual detection. Certainty achieves traffic safety and extends the service life of trains and tracks.

For the convenience of the description, the central idea expressed by the present invention in the column of the above summary of the invention is expressed by the specific embodiments. Various items in the embodiments are depicted in terms of ratios, dimensions, amounts of deformation, or displacements that are suitable for illustration, and are not drawn to the proportions of actual elements, as set forth above.

Referring to FIG. 1 to FIG. 9, the present invention provides a train wheel abnormality detecting system 100, which includes:

A sensor 10 is disposed on the track 1 for providing the train 2 to travel. The train 2 is of a normal type, for example, a MRT, a train, a high-speed train, etc., and the train 2 includes a plurality of cars 3, each A carriage 3 is provided with a wheel set 4 corresponding to each of the front and rear ends, and each wheel set 4 is provided with a wheel 5 on each side of the track 1 for the left wheel 5a and the right wheel 5b.

When the wheel 5 of the train 2 passes the sensor 10 along the track 1, the sensor 10 draws the wheel 5 of the train 2 to generate a vibration acceleration signal through the track 1. In the embodiment of the present invention, the sensor 10 The acceleration gauge is one, and the number of the sensors 10 is one. The sensor 10 is vertically disposed on one side 6 of the track 1, as shown in FIG.

Furthermore, in the embodiment of the present invention, the sampling frequency of the sensor 10 is greater than 1000 Hz, indicating that the sensor 10 samples at least 1000 times per second, wherein the sampling frequency can be adjusted according to the type or demand of the train 2 The vibration acceleration signal captured includes a plurality of vibration sources because of the frequency generated by the vibration, wherein the vibration source is a track 1, a train 2, a car 3, a wheel 5, or a vibration generated by the train 2, and each vibration source is vibrated. The frequency of generation is different.

An analysis controller 20 is connected to the sensor 10 signal, and the signal connection method is a wired electrical connection or a wireless signal connection. The analysis controller 20 is configured to receive the vibration acceleration signal, and the analysis controller 20 has an initial analysis module 21, and the initial analysis module 21 analyzes the vibration acceleration signal into a vibration speed signal, wherein the initial analysis module 21 has an integral. The unit 211 converts the vibration acceleration signal integral into a vibration speed signal by the integration unit 211. In the embodiment of the present invention, the initial analysis module 21 converts the vibration acceleration signal generated by various vibration sources into a vibration speed signal.

Furthermore, the analysis controller 20 has a feature analysis module 22, which analyzes the vibration velocity signal generated by the wheel 5 by the vibration velocity signal by a frequency distribution analysis of the structural characteristics, and the vibration in the non-frequency distribution. The speed signal is omitted, wherein since the wheel 5 is a rigid body, the vibration frequency generated by the wheel 5 is a low frequency, and therefore, the frequency distribution due to the structural characteristics is 0.1 Hz to 5 Hz.

According to the frequency distribution of the structural characteristics, the vibration speed signal is distinguished, and based on the frequency domain harmonic wavelet transform, the time waveform of the wheel 5 of the train 2 passing through the track 1 is calculated, and a passage time interval T is obtained by using a threshold value analysis for Confirming the exact time when the wheel 5 passes, since the time of each wheel 5 of the train 2 passing through the sensor 10 is short, the frequency interval harmonic wavelet transform can be used to analyze and obtain the passage time interval T of each wheel 5, wherein the valve The value is a threshold effect, which can clearly reflect the vibration speed signal generated when the wheel 5 passes, and the threshold can be adjusted according to the type of the different trains 2.

In addition, the vibration velocity signal is distinguished by the frequency distribution, and based on the frequency domain harmonic wavelet transform analysis, the vibration source signal of the wheel 5 passing through the track 1 corresponding to the time interval T is calculated by the frequency domain harmonic wavelet transform analysis, and the feature analysis module 22 according to the passage time interval. T and the vibration source signal use a root mean square analysis to obtain a vibration amount. Wherein, the number of time intervals T and the number of vibration source signals are plural, each time interval T and each vibration source signal correspond to one wheel 5, and each wheel 5 corresponds to a vibration amount A, whereby the feature The analysis module 22 can effectively analyze that a train 2 includes several compartments 3, wheel sets 4 and wheels 5, and in the embodiment of the present invention, as shown in FIG. 1, FIG. 3, FIG. 6 and FIG. The module 22 analyzes and obtains 24 vibration quantities A according to the vibration acceleration signal captured by the sensor 10, and is obtained by each time interval T. The train 2 is provided with 6 compartments 3, and each compartment 3 is provided. There are two wheel sets 4 at the front end and the back end, for a total of 12 wheel sets 4, each wheel set 4 is provided with one wheel 5 on each side of the track 1, and the wheel 5 is divided into a left wheel 5a and a right wheel. 5b, a total of 24 wheels 5, therefore, the present invention can clearly analyze each of the cars 3, the wheel sets 4 and the wheels 5 of each train 2 by the analysis controller 20, and the corresponding vibration amount A Follow up for anomaly detection.

The analysis controller 20 has a determination module 23, and the determination module 23 compares the vibration amount A with a basic comparison value B, and the determination module 23 determines whether the vibration amount A exceeds the basic comparison value B, and if the vibration amount A is greater than The base comparison value B sends an exception signal. The basic comparison value B is obtained by using the train 2 to travel through the track 1 for the first time, and the vibration acceleration signal captured by the sensor 10 is analyzed and obtained by the initial analysis module 21 and the feature analysis module 22 of the analysis controller 20. The amount A', the vibration amount A' calculated by the first traveling train 2 through the track 1 is established as the basic comparison value B, and stored in the judging module 23, and then the vibration amount A calculated by the train 2 through the track 1 Then, it is possible to make a comparison judgment with the base comparison value B.

Therefore, the basic comparison value B of each train 2 is different, and the comparison value of the basic comparison value B established by the characteristics of each train 2 can be compared, and the comparison value is directly set. Accurate, more able to meet the judgment criteria of each train 2, and effectively find the anomaly.

One embodiment of the present invention provides a method for detecting an abnormality of a train wheel by the foregoing train wheel abnormality detecting system 100, which comprises the following steps:

Step S1: The sensor 10 is vertically disposed on one side 6 of the track 1 , and the sensor 10 picks up the vibration acceleration signal when the wheel 5 of the train 2 passes the track 1 , and the sensor 10 vibrates The acceleration signal is transmitted to the analysis controller 20 in a signal transmission manner.

The first analysis step S2: the initial analysis module 21 converts and analyzes the vibration velocity signal by the integration unit 211 according to the vibration acceleration signal.

The second analysis step S3: the feature analysis module 22 distinguishes the vibration source from the vibration velocity signal of the wheel 5 according to the vibration frequency signal, and the vibration velocity signal differentiated by the frequency distribution is based on the frequency domain harmonic. Wavelet transform, calculate the time waveform, and use the threshold to obtain the passage time interval T; while calculating the passage time interval T, the vibration velocity signal differentiated by the frequency distribution is obtained based on the frequency domain harmonic wavelet transform calculation and analysis to obtain the vibration source signal.

Third analysis step S4: The feature analysis module 22 obtains the vibration amount A by using the root mean square based on the time interval T and the vibration source signal.

Establishing a benchmark step S5: When each train 2 is traveling for the first time, the vibration acceleration signal when the wheel 5 of the train 2 first travels through the track 1 can be obtained by the step S1; the first time, the analysis controller 20 is used according to the first time. After the vibration acceleration signal passes through the first analysis step S1, the second analysis step S2, and the third analysis step S3, the vibration amount A′ when the wheel 5 of the first train 2 passes through the track 1 is calculated and analyzed, and the initial vibration amount A′ is established. The value B is compared and stored in the determination module 23.

The comparison determining step S6: after the train 2 has established the basic comparison value B, and then the corresponding train 2, then the step S1, the first analyzing step S2, the second analyzing step S3 and the third analyzing step can be performed. S4 obtains the vibration amount A of the wheel 5 during the current running, and compares the vibration amount A with the basic comparison value B by the determination module 23, and determines whether the vibration amount A exceeds the basic comparison value B.

The judging module 23 compares the vibration amount A of each of the wheels 5 with the basic comparison value B, and when one or more of the vibration amounts A of the wheels 5 are larger than the basic comparison value B, an abnormal signal is issued. In the embodiment of the present invention, as shown in FIG. 1 and FIG. 9, the vibration amount A of the right wheel 5b of the front end wheel group 4 of the third car 3 is greater than the basic comparison value B, and after the sixth car 3 If the vibration amount A of the left wheel 5a of the end wheel set 4 is greater than the basic comparison value B, the determination module 23 will send an abnormal signal, and the abnormal signal can be notified by the way, the front wheel of the third car 3 of the train 2 The right wheel 5b of the group 4 and the left wheel 5a of the rear wheel group 4 of the sixth car 3 are abnormal, and need to be repaired and processed.

When the judging module 23 compares the vibration amount A of each wheel 5 with the basic comparison value B, and the vibration amount A is smaller than the basic comparison value B, the judging module 23 judges that the wheels 5 are all in a normal state, as shown in the figure. As shown in Fig. 8, the vibration value A of each wheel is smaller than the basic comparison value B, and the train 2 can continue to drive safely.

In the embodiment of the present invention, the analysis controller 20 is capable of being connected to the smart mobile device, and the smart mobile device can display the determination module 23 to obtain the determination result, and the detecting personnel can use the monitoring personnel to monitor whether the wheel 5 is abnormal. The judging module 23 sends an abnormal signal, which is displayed in the smart mobile device. In the embodiment of the present invention, as shown in FIG. 1 and FIG. 9, the judging module 23 sends an abnormal signal, and the train is displayed on the smart mobile device. 2, the third car 3, the right wheel 5b of the front end wheel set 4, and the left wheel 5a of the rear end wheel set 4 of the sixth car 3, an abnormal problem occurs, thereby achieving the effect of being ready to use and monitoring at any time. .

Furthermore, the train wheel abnormality detecting system 100 of the present invention and the method thereof can analyze that each train 2 has several cars 3, wheel sets 4 and wheels 5, and effectively detects each wheel set 4 of each train 2. And the state of each of the wheels 5 of each of the wheel sets 4, and it is possible to accurately determine the abnormality of the car 3, the wheel set 4 or the wheels 5, and the abnormality.

In addition, the vibration acceleration signal is further converted, and the vibration analysis source is distinguished by the feature analysis module 22 and analyzed, thereby reducing analysis errors and improving measurement accuracy.

In addition, with the train wheel abnormality detecting system 100 of the present invention and the method thereof, the cost of manual detection and maintenance can be reduced, traffic safety can be promoted, and the service life of the train 2 and the track 1 can be extended.

The above embodiments are merely illustrative of the invention and are not intended to limit the scope of the invention. Any modifications or variations that are made without departing from the spirit of the invention are intended to be protected.

1‧‧‧ Track
22‧‧‧Characteristic Analysis Module
2‧‧‧ Train
23‧‧‧Judgement module
3‧‧‧ compartment
T‧‧‧ passage time interval
4‧‧‧ Wheel set
A‧‧‧vibration
5‧‧‧ Wheels
A'‧‧‧vibration
5a‧‧‧Left wheel
B‧‧‧Basic comparison value
5b‧‧‧Right wheel
S1‧‧‧ capture steps
6‧‧‧ side
S2‧‧‧First analysis step
100‧‧‧ Train wheel anomaly detection system
S3‧‧‧Second analysis step
10‧‧‧ Sensors
S4‧‧‧ third analysis step
20‧‧‧Analysis controller
S5‧‧‧ Establishing benchmark steps
21‧‧‧Initial analysis module
S6‧‧‧ comparison judgment step
211‧‧‧Integral unit

1 is a system architecture diagram of the present invention. 2 is a flow chart of the method of the present invention. 3 is a schematic view of the sensor of the present invention disposed on a track. Figure 4 is a schematic view of the vibration acceleration signal of the present invention. Figure 5 is a schematic view of the vibration velocity signal of the present invention. Figure 6 is a schematic illustration of the passage time interval of the present invention. Figure 7 is a schematic diagram of the vibration source signal of the present invention. Fig. 8 is a schematic view showing the vibration amount of the present invention, showing a vibration amount smaller than a basic comparison value. FIG. 9 is a schematic diagram showing the vibration amount larger than the basic comparison value in the embodiment of the present invention.

Claims (10)

A train wheel abnormality detecting system, comprising: a sensor disposed on a track, the sensor picking up a wheel of the train to vibrate an acceleration signal through one of the tracks; and an analysis controller, the sensor signal and the sensor signal Connecting and receiving the vibration acceleration signal, the analysis controller has an initial analysis module, a feature analysis module and a determination module, and the initial analysis module analyzes the vibration acceleration signal into a vibration speed signal, and the feature analysis The module analyzes the vibration speed signal by a frequency distribution to obtain a vibration source signal of one of the train passing track and a corresponding time interval, and the characteristic analysis module according to the passing time interval and the vibration source signal A vibration amount is analyzed, and the determining module compares the vibration amount with a basic comparison value to determine whether the vibration amount exceeds the basic comparison value. The train wheel abnormality detecting system of claim 1, wherein the initial analysis module has an integrating unit, and the integrating unit converts the vibration speed signal into the vibration speed signal according to the vibration acceleration signal. The train wheel abnormality detecting system according to claim 1 or 2, wherein the characteristic analyzing module converts the vibration speed signal by the frequency domain based on frequency domain harmonic wavelet transform, and obtains the transit time by using a threshold value analysis. Interval. The train wheel abnormality detecting system according to claim 3, wherein the characteristic analyzing module obtains the vibration source signal by using the frequency distribution based on the frequency domain harmonic wavelet transform analysis, and the characteristic analyzing module is configured according to the The passing time interval and the vibration source signal are obtained by using the root mean square. The train wheel abnormality detecting system according to claim 4, wherein the frequency distribution is 0.1 Hz to 5 Hz. The train wheel abnormality detecting system according to claim 1, wherein the number of the wheels of the train is plural, and the passing time interval and the number of the vibration source signals are plural, and each passing time interval and each vibration source signal are mutually Corresponds to each wheel. A method for detecting an abnormality of a train wheel includes the following steps: placing a sensor on a track, and extracting, by the sensor, a vibration acceleration signal of a train wheel passing through a track; transmitting the vibration acceleration signal to an analysis control The initial analysis module of the analysis controller analyzes the vibration acceleration signal as a vibration speed signal; and the characteristic analysis module of the analysis controller obtains the wheel of the train by using a frequency distribution analysis according to the vibration speed signal. Passing one of the tracks through the time interval and corresponding to the vibration source signal of one of the time intervals; the feature analysis module analyzes a vibration amount according to the passage time interval and the vibration source signal; and the determination module of the analysis controller The vibration amount is compared with a basic comparison value to determine whether the vibration amount exceeds the basic comparison value, and when the vibration amount is greater than the basic comparison value, an abnormal signal is emitted. The train wheel abnormality detecting method according to claim 7, wherein the vibration acceleration signal of the first train passing the track is captured by the sensor, and the analysis controller is used to establish the basic ratio according to the initial vibration acceleration signal analysis. For the value. The method for detecting an abnormality of a train wheel according to claim 7, wherein the characteristic analysis module converts the vibration speed signal based on the frequency domain harmonic wavelet transform, and obtains the passage time interval by using a threshold value analysis. The method for detecting an abnormality of a train wheel according to claim 7, wherein the characteristic analysis module obtains the vibration source signal based on the frequency domain harmonic wavelet transform analysis, and then the characteristic analysis mode The group obtains the vibration amount by using the root mean square according to the passage time interval and the vibration source signal.