CN108827665A - Wheel flat fault detection method based on empirical mode decomposition and multi-scale entropy - Google Patents
Wheel flat fault detection method based on empirical mode decomposition and multi-scale entropy Download PDFInfo
- Publication number
- CN108827665A CN108827665A CN201810530132.3A CN201810530132A CN108827665A CN 108827665 A CN108827665 A CN 108827665A CN 201810530132 A CN201810530132 A CN 201810530132A CN 108827665 A CN108827665 A CN 108827665A
- Authority
- CN
- China
- Prior art keywords
- entropy
- sensor
- multiple dimensioned
- vibration
- mode decomposition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M17/00—Testing of vehicles
- G01M17/08—Railway vehicles
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61K—AUXILIARY EQUIPMENT SPECIALLY ADAPTED FOR RAILWAYS, NOT OTHERWISE PROVIDED FOR
- B61K9/00—Railway vehicle profile gauges; Detecting or indicating overheating of components; Apparatus on locomotives or cars to indicate bad track sections; General design of track recording vehicles
- B61K9/12—Measuring or surveying wheel-rims
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Electric Propulsion And Braking For Vehicles (AREA)
- Vehicle Body Suspensions (AREA)
Abstract
The invention discloses a kind of wheel flat fault detection method based on empirical mode decomposition and multi-scale entropy.Method is：Rail vibration signal when 4 vibration acceleration sensors are installed in track unilateral side, and acquisition train passes through；Collected rail vibration signal is filtered, signal related with unsteadiness of wheels is retained；Empirical mode decomposition is carried out to filtered signal, extracts first three natural mode of vibration component of signal；The multiple dimensioned entropy for calculating first three natural mode of vibration component sum obtains the multiple dimensioned entropy of each sensor；The average multiple dimensioned entropy for calculating each sensor judges wheel with the presence or absence of flat scar failure by comparing the multiple dimensioned entropy curve of same train normal wheels.The present invention utilizes the collected rail vibration signal of vibration acceleration sensor, by carrying out the analysis of wavelet-packet energy entropy to vibration signal, realizes the detection to rail Corrugation, has the advantages that real-time is good, easy to detect and speed is fast, applied widely.
Description
Technical field
The present invention relates to city rail vehicle wheel flat detection technique fields, especially a kind of based on empirical mode decomposition and more
The wheel flat fault detection method of Scale Entropy.
Background technique
With the quickening of urbanization process, the increase of urban population quantity, the platform distance of urban track traffic is increasingly
Short, departure frequency is more and more intensive, and passenger capacity is more and more, and vehicle operational safety problem is caused also to become more and more prominent.Wheel
To the critical component as locomotive operation, vehicle weight is not only carry, also carries the various impact forces between wheel track.Work as wheel
When tyre tread is damaged there are wheel wear and flat scar, additional shock loading can be caused to wheel, wheel shaft.With adding for flat scar degree
Deep and car speed increase, bring shock loading can reach several times that wheel bears dead load, and due to the wheel period
Property rotation, flat scar periodic shock track will cause the failures such as track wave mill, and then exacerbates the degree of impairment of track.If no
Wheel flat failure can be found in time, and maintenance replacement is carried out to wheel, gently then vehicle vibration is aggravated, it is steady to influence train operation
It is qualitative, reduce passenger's riding comfort, it is heavy then cause wheel shaft break, the major accidents such as vehicle rollover, derailing.
Currently, most of wheel flat failures are found during the scheduled maintenance inspection to vehicle.Due to car test
The car test personnel of department are detected by railway special measuring tool, detect large labor intensity and inefficiency, exist to find in time
A possibility that wheel flat failure causes missing inspection, and artificial detection have it is stronger it is empirical, can not real-time tracking, be not easy to
The disadvantages of information system management.
Summary of the invention
Good, easy to detect and speed that the purpose of the present invention is to provide a kind of real-times is fast, applied widely based on warp
Test the wheel flat fault detection method of mode decomposition and multi-scale entropy.
The technical solution for realizing the aim of the invention is as follows：A kind of wheel based on empirical mode decomposition and multi-scale entropy is flat
Scar fault detection method, includes the following steps：
Step 1, the rail vibration signal when 4 vibration acceleration sensors are installed in track unilateral side, and acquisition train passes through；
Step 2, collected Vertical Acceleration signal is filtered using low-pass filter, removes high frequency
Noise jamming；
Step 3, empirical mode decomposition is carried out to the Vertical Acceleration signal after filtering processing, it is intrinsic extracts first three
Modal components, and calculate itself and value；
Step 4, multi-scale entropy analysis is carried out in step 3 and value, obtains the multiple dimensioned entropy of each sensor；
Step 5, the average multiple dimensioned entropy of each sensor is calculated；
Step 6, the average multiple dimensioned entropy of the average multiple dimensioned entropy of each sensor and normal wheels is compared,
If the average multi-scale entropy curve of collected each sensor judges above the average multi-scale entropy curve of normal wheels
Wheel is on the contrary then flat scar failure is not present there are flat scar failure.
Further, low-pass filter described in step 2, specially Chebyshev's bandpass filter, turn-on frequency are
[80Hz,700Hz]。
Further, empirical mode decomposition is carried out to the Vertical Acceleration signal after filtering processing described in step 3,
First three natural mode of vibration component is extracted, and calculates itself and value, it is specific as follows：
Xi=IMF1+IMF2+IMF3 (1)
In formula, XiFor i-th of sensor first three items natural mode of vibration component and value, i=1,2,3,4, IMF1、IMF2、
IMF3For the first three items natural mode of vibration component of i-th of sensor.
Further, multi-scale entropy analysis is carried out in step 3 and value described in step 4, obtains each sensor
Multiple dimensioned entropy, it is specific as follows：
To the first three items natural mode of vibration component of i-th of sensor and value XiMulti-scale entropy analysis is carried out, is obtained multiple dimensioned
Entropy curve MSEi, i=1,2,3,4.
Further, the average multiple dimensioned entropy of each sensor of calculating described in step 5, it is specific as follows：
Wherein, MSEavgFor the average multiple dimensioned entropy of sensor.
Compared with prior art, the present invention its remarkable advantage is：(1) on-line checking of the detection mode based on vehicle in use,
Real-time is good, easy to detect, at low cost；(2) vibrating sensor installation is simple and convenient, and adapts to the severe building ring in scene
Border；(3) have the advantages that detection speed is fast, applied widely, subway maintenance department is detected in time, efficiently and accurately flat
Scar failure ensures train operation safety, promotes passenger comfort, has very important significance.
Detailed description of the invention
Fig. 1 is that the present invention is based on the flow charts of empirical mode decomposition and the wheel flat fault detection method of multi-scale entropy.
Fig. 2 is Chebyshev's bandpass filter passband curve graph in the present invention.
Fig. 3 is collected vertical vibration signal graph in the embodiment of the present invention.
Fig. 4 is the vertical vibration signal graph in the embodiment of the present invention after frequency reducing.
Fig. 5 is average multi-scale entropy curve graph in the embodiment of the present invention.
Fig. 6 is live wheel flat figure in the embodiment of the present invention.
Specific embodiment
The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
In conjunction with Fig. 1, the wheel flat fault detection method based on empirical mode decomposition and multi-scale entropy in the present invention, packet
Include following steps：
Step 1, the rail vibration signal when 4 vibration acceleration sensors are installed in track unilateral side, and acquisition train passes through.
Step 2, it is filtered using vibration signal of the low-pass filter to acquisition, retaining has with raceway surface recess
The information of pass, removal high-frequency noise interference.The vibration frequency range as caused by car body itself, flat scar etc. is [50Hz, 500Hz],
And the speed of service of municipal rail train is no more than 80km/h, therefore the passband section of bandpass filter is set as [80Hz, 700Hz], such as
Shown in Fig. 2.
Step 3, empirical mode decomposition is carried out to the signal of 4 sensors respectively, extracts first three natural mode of vibration component, and
Itself and value are calculated, the intrinsic mode function and value of four sensors are respectively obtained, it is specific as follows：
Xi=IMF1+IMF2+IMF3 (1)
X in formulaiFor the first three items intrinsic mode function and value of i-th (i=1,2,3,4) a sensor, IMF1、IMF2、IMF3
For first three intrinsic mode function of i-th of sensor.
Step 4, to XiMulti-scale entropy analysis is carried out, multi-scale entropy curve MSE is obtainedi(i=1,2,3,4).
Step 5, the average multiple dimensioned entropy of each sensor is calculated, specially：
Step 6, the average multiple dimensioned entropy of the average multiple dimensioned entropy of each sensor and normal wheels is compared,
If the average multi-scale entropy curve of collected each sensor judges above the average multi-scale entropy curve of normal wheels
Wheel is on the contrary then flat scar failure is not present there are flat scar failure.
Combined with specific embodiments below, invention is further described in detail.
Embodiment 1
The present embodiment passes through rail vibration signal when certain route, fortune using certain type train of certain collected metro company
Row speed per hour is 40km/h, and wherein No.1 sensor acquisition vibration signal is as shown in Figure 3.For convenience of description, with one of sensing
The data of device do calculation specifications.This section of vibration signal is filtered using chebyshev low-pass filter first, is filtered out
Useless clutter information.Vibration signal after filtering processing is as shown in Figure 4.Since the vehicle is the marshalling of 6 sections, 12 steerings are shared
Frame.Measured signal is collected to the sensor to be segmented according to compartment.As shown in Figure 4, go out in the position vibration acceleration value of 4-8s
Show more apparent amplitude transition, therefore most possibly occurs flat scar failure at this.
Front of the car is named as 1 end, rear end is named as 2 ends, and the B vehicle and C vehicle of the odd number license number in selection figure are divided
Analysis, is denoted as 1B, 2B, 1C, 2C for B vehicle and C vehicle bogie respectively.Experience is carried out to the vbs1 vibration sensor signal of B1 bogie
Mode decomposition, and corresponding IMF1+IMF2+IMF3 value is acquired, the feature by the intrinsic mode function and value as 2B bogie
Signal carries out multi-scale entropy analysis to this feature signal；Above step is repeated, the multi-scale entropy for acquiring its excess-three respectively is bent
Line finds out the average value of multi-scale entropy, and the average multi-scale entropy curve for obtaining 1B, 2B, 1C, 2C is as shown in Figure 5.As shown in Figure 5,
The corresponding time of vibration sequence complexity of bogie 2C is maximum, therefore determines comprising fault message in the segment data, i.e., in bogie
There are flat scar failures on 2C.
For check algorithm correctness, the vehicle wheel is checked in rolling stock section after detection, is found in odd number vehicle
Number wheel of C vehicle rear end 6 has peeling, and on-site actual situations find the flat scar length about 4cm as shown in fig. 6, measuring the position
Wide 2cm depth 0.3mm, this and analysis result are coincide substantially, so that wheel flat failure can preferably be realized by demonstrating the method
Identification and the positioning of abort situation.
Claims (5)
1. a kind of wheel flat fault detection method based on empirical mode decomposition and multi-scale entropy, which is characterized in that including with
Lower step：
Step 1, the rail vibration signal when 4 vibration acceleration sensors are installed in track unilateral side, and acquisition train passes through；
Step 2, collected Vertical Acceleration signal is filtered using low-pass filter, removes high-frequency noise
Interference；
Step 3, empirical mode decomposition is carried out to the Vertical Acceleration signal after filtering processing, extracts first three natural mode of vibration
Component, and calculate itself and value；
Step 4, multi-scale entropy analysis is carried out in step 3 and value, obtains the multiple dimensioned entropy of each sensor；
Step 5, the average multiple dimensioned entropy of each sensor is calculated；
Step 6, the average multiple dimensioned entropy of the average multiple dimensioned entropy of each sensor and normal wheels is compared, if adopting
The average multi-scale entropy curve of each sensor collected then judges wheel above the average multi-scale entropy curve of normal wheels
There are flat scar failure, it is on the contrary then be not present flat scar failure.
2. the wheel flat fault detection method according to claim 1 based on empirical mode decomposition and multi-scale entropy,
Be characterized in that, low-pass filter described in step 2, specially Chebyshev's bandpass filter, turn-on frequency be [80Hz,
700Hz]。
3. the wheel flat fault detection method according to claim 1 based on empirical mode decomposition and multi-scale entropy,
It is characterized in that, empirical mode decomposition is carried out to the Vertical Acceleration signal after filtering processing described in step 3, extracts first three
A natural mode of vibration component, and itself and value are calculated, it is specific as follows：
Xi=IMF1+IMF2+IMF3 (1)
In formula, XiFor i-th of sensor first three items natural mode of vibration component and value, i=1,2,3,4, IMF1、IMF2、IMF3For
The first three items natural mode of vibration component of i-th of sensor.
4. the wheel flat fault detection method according to claim 3 based on empirical mode decomposition and multi-scale entropy,
It is characterized in that, multi-scale entropy analysis is carried out in step 3 and value described in step 4, obtains the multiple dimensioned of each sensor
Entropy, it is specific as follows：
To the first three items natural mode of vibration component of i-th of sensor and value XiMulti-scale entropy analysis is carried out, multi-scale entropy curve is obtained
MSEi, i=1,2,3,4.
5. the wheel flat fault detection method according to claim 4 based on empirical mode decomposition and multi-scale entropy,
It is characterized in that, the average multiple dimensioned entropy of each sensor of calculating described in step 5 is specific as follows：
Wherein, MSEavgFor the average multiple dimensioned entropy of sensor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810530132.3A CN108827665A (en) | 2018-05-29 | 2018-05-29 | Wheel flat fault detection method based on empirical mode decomposition and multi-scale entropy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810530132.3A CN108827665A (en) | 2018-05-29 | 2018-05-29 | Wheel flat fault detection method based on empirical mode decomposition and multi-scale entropy |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108827665A true CN108827665A (en) | 2018-11-16 |
Family
ID=64146094
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810530132.3A Pending CN108827665A (en) | 2018-05-29 | 2018-05-29 | Wheel flat fault detection method based on empirical mode decomposition and multi-scale entropy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108827665A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110171442A (en) * | 2019-06-12 | 2019-08-27 | 中国神华能源股份有限公司 | Detection system, the detection method of wheel flat |
| CN113111516A (en) * | 2021-04-14 | 2021-07-13 | 华东交通大学 | Track system infrastructure parameter identification method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103840742A (en) * | 2014-03-04 | 2014-06-04 | 常州恒大光伏科技有限公司 | Intelligent management method and system for photovoltaic module |
| CN105092467A (en) * | 2015-09-09 | 2015-11-25 | 南车工业研究院有限公司 | Rapid monitoring device and method for high-speed railway wheel tread abrasion |
| CN106156503A (en) * | 2016-07-05 | 2016-11-23 | 中国矿业大学 | A kind of multi-scale entropy characterizing method of anchor system internal flaw distribution |
| CN106250613A (en) * | 2016-07-28 | 2016-12-21 | 南京理工大学 | A kind of wheel service state security domain is estimated and method for diagnosing faults |
| CN107406090A (en) * | 2015-01-16 | 2017-11-28 | 国际电子机械公司 | abnormal vehicle dynamic detection |
-
2018
- 2018-05-29 CN CN201810530132.3A patent/CN108827665A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103840742A (en) * | 2014-03-04 | 2014-06-04 | 常州恒大光伏科技有限公司 | Intelligent management method and system for photovoltaic module |
| CN107406090A (en) * | 2015-01-16 | 2017-11-28 | 国际电子机械公司 | abnormal vehicle dynamic detection |
| CN105092467A (en) * | 2015-09-09 | 2015-11-25 | 南车工业研究院有限公司 | Rapid monitoring device and method for high-speed railway wheel tread abrasion |
| CN106156503A (en) * | 2016-07-05 | 2016-11-23 | 中国矿业大学 | A kind of multi-scale entropy characterizing method of anchor system internal flaw distribution |
| CN106250613A (en) * | 2016-07-28 | 2016-12-21 | 南京理工大学 | A kind of wheel service state security domain is estimated and method for diagnosing faults |
Non-Patent Citations (3)
| Title |
|---|
| 周璇 等: "《基于经验模式分解和神经网络的车轮踏面擦伤检测方法》", 《润滑与密封》 * |
| 秦喜文 等: "《利用经验模态分解和多尺度熵的滚动轴承故障诊断研究》", 《制造业自动化》 * |
| 郭金龙: "《机车车辆轮对踏面故障检测系统研究》", 《中国优秀硕士学位论文数据库 工程科技II辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110171442A (en) * | 2019-06-12 | 2019-08-27 | 中国神华能源股份有限公司 | Detection system, the detection method of wheel flat |
| CN113111516A (en) * | 2021-04-14 | 2021-07-13 | 华东交通大学 | Track system infrastructure parameter identification method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106394606B (en) | A kind of railway car wheel loses circle detection method and detection device | |
| CN108845028B (en) | Method and device for dynamically detecting high-speed railway rail corrugation | |
| CN108515984B (en) | Wheel damage detection method and device | |
| CN106018557B (en) | A kind of raceway surface recess detection method based on wavelet analysis | |
| JP4388532B2 (en) | Railway vehicle abnormality detection device | |
| US20020056398A1 (en) | Method and device for monitoring bogies of multi-axle vehicles | |
| CN109278796B (en) | Vehicle-mounted wheel out-of-roundness detection system | |
| CN105923014B (en) | A kind of track transition Amplitude Estimation method based on evidential reasoning rule | |
| CN204055812U (en) | A kind of vehicle derailing monitoring system | |
| CN110606105B (en) | Safety monitoring system for walking part of unmanned subway vehicle | |
| CN108827665A (en) | Wheel flat fault detection method based on empirical mode decomposition and multi-scale entropy | |
| CN113624521A (en) | Method and system for monitoring snake-shaped instability of train based on axle box vibration | |
| CN105336176A (en) | Second-hand vehicle detection method and system | |
| CN113295137A (en) | Road unevenness identification method based on vehicle wheel speed fluctuation | |
| CN113032907A (en) | Method and system for correcting vehicle shaking disease data deviation based on waveform correlation | |
| CN110171442B (en) | Detection system and detection method for wheel flat scars | |
| CN109242117B (en) | Grading method for five-level overhaul of urban railway vehicle | |
| CN105608416B (en) | Vibrate bruise detection method | |
| CN113343919B (en) | Method and device for detecting continuous equidistant rubbing damage of steel rail and computer equipment | |
| CN217385548U (en) | Non-contact rail transit train speed direction mileage detection device | |
| JP4118780B2 (en) | Vehicle abnormality detection system and abnormality detection method | |
| CN117957158A (en) | Method for detecting damage to a transport system and control device therefor | |
| CN104504713A (en) | Automatic trouble identification method based on TEDS (trouble of moving EMU detection system) | |
| CN210591925U (en) | Axle box vibrating wheel diameter difference detection system | |
| CN108528475A (en) | A kind of track transition fault alarm method based on multi-level fusion |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181116 |
|
| RJ01 | Rejection of invention patent application after publication |