CN108982507A - A kind of Rail Surface hurt detection device and method - Google Patents
A kind of Rail Surface hurt detection device and method Download PDFInfo
- Publication number
- CN108982507A CN108982507A CN201710401859.7A CN201710401859A CN108982507A CN 108982507 A CN108982507 A CN 108982507A CN 201710401859 A CN201710401859 A CN 201710401859A CN 108982507 A CN108982507 A CN 108982507A
- Authority
- CN
- China
- Prior art keywords
- rail
- hurt
- data
- point
- sensor
- 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
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8854—Grading and classifying of flaws
- G01N2021/8861—Determining coordinates of flaws
Abstract
The present invention is a kind of Rail Surface hurt detection device and method.The device mainly includes four laser sensors to acquire data, and conveying roller transmits rail, and encoder records location of rail.This method is by collected data anticipatory staging and removes lap.A small range data are first analyzed when calculating, the adjacent small range data of post analysis are so progressive until completing entire rail profile.This method represents small-scale desired profile using Parabolic Fit or straight line fitting, then judges hurt according to the distance of respective point in data point to desired profile.When finding hurt, which will stop rail movement, real-time sound-light alarm, and record current and subsequent several scanning section profiles and current location of rail.User can inquire these profiles 3 D stereo view and other data.The present invention provides the key devices and method of a kind of rail automatic detection, can significantly improve the accuracy rate of hurt detection, greatly reduce the security risk in subsequent use.
Description
Technical field
The present invention relates to Rail Surface hurt detection fields, and in particular to a kind of Rail Surface hurt detection device and side
Method.
Technical background
Background technique
As the speed raising and large area of railway cover, railway traffic is increasingly becoming the mainstream tool of people's trip, every year
The volume of the flow of passengers increase swift and violent, the surface hurt of rail can be accumulated gradually in train load and under the influence of greatly improving of density
It causes rail internal injury and then the accident of broken rail occurs, therefore find and check Rail Surface situation as early as possible to the operational safety of train
It plays a crucial role.It is main at present at home that surface hurt is detected using artificial range estimation, the degree of automation, labour it is strong
Big, low efficiency, poor reliability are spent, detection is carried out using detection device indirect labor and is greatly improved detection efficiency and detection accurately
Rate.
Summary of the invention
The technical problem to be solved by the present invention is in view of the above, provide a kind of Rail Surface hurt detection device and
Method is easy to implement automation to improve the efficiency and accuracy of hurt detection.
The present invention solves its technical problem, and the following technical solution is employed:
Rail Surface hurt detection device provided by the invention, it is characterized in that including pedestal, sensor stand, encoder branch
Frame, orbit transports roller, encoder, rail, four sensors, wherein encoder bracket is fixed on the base and encoder center is inclined
The length of center at least code wheel radius from sensor stand, encoder is the incremental encoder that diameter is 120mm, installation
On encoder bracket, sensor stand and orbit transports roller are fixed on the base, the support of orbit transports roller and conveying rail, steel
Rail is tangent with code wheel, and four sensors are mounted at four interior angles of sensor stand.
Rail Surface hurt detection method provided by the invention, specifically: principle of triangulation is utilized, it is symmetrical using four
Be mounted on be in central axes 45 ° of angles sensor stand interior angle at line structure optical sensor scanning Rail Surface, lead to
The discussion that hurt judgement is carried out to data point is crossed, sequence takes three point P1, P2, P3 quasi- using parabola at data starting position first
It closes, matched curve S, N number of point is spaced between adjacent two o'clock, due to a possibility that there are random errors for the two ends of data
Greatly, therefore can suitably remove each 5 points of head and the tail, take adjacent 3 points Q1, Q2, Q3, Q1 and M, the interval P1 point, ask respectively Q1, Q2,
The distance of Q3 to S, and be compared with standard, judge whether there is hurt.
In the above method, using the data carry out rail defects and failures detection before, first according to rail profile the characteristics of, Mei Gechuan
There are blind areas for the data of sensor acquisition, therefore it is two segment datas that a sensor is collected, and due to sensor installation position
Feature is set, there are overlapping region, at the midpoint of overlapping region search overlay region, removes that there are the overlappings of two segment data of overlay region
End, to improve detection efficiency and accuracy.Before carrying out hurt Data Detection, 5 points are removed to the head and the tail of every segment data, with
The random disturbances at endpoint are avoided, and are turning point at the radius of curvature mutation reduction of scanning rail profile by data segment
Place is segmented, and detection efficiency and accuracy are improved.
Before carrying out hurt detection, according to the installation site of sensor, the web of the rail that sensor 2 and sensor 4 are scanned
Segment data carries out character zone judgement, protrudes character depth D=1mm, and recess character depth is 0.2mm, since character is located at half
On the circular arc of diameter R=400mm, provided according to TB/T3276-2011, TB/T2344-2012 hurt, rail tread hurt threshold value T1
=0.3mm, other hurts threshold value T2=0.5mm, therefore the character that is recessed judges no influence to hurt, according to the circular arc at the web of the rail
Two o'clock find center of circle O, intercept L point (x of the middle section of the circular arc0, y0), (x1, y1) ... (xl, yl), sentence point by point
Break each point and O point distance S0, S1... Sl, compare SiWith R, i=0,1......l, if having more than 60% R-Si≥
0.6, then the region is judged for character area, which is carried out to the region of every scan line.When being character zone at the web of the rail,
Hurt threshold value at this is set as T3=1.5mm.
In the above method, hurt judgement is carried out, may comprise steps of:
Step 1: tactic three point P1, P2, P3 at equal intervals in the data segment after removing overlay region, to this 3 points into
It is used as ideal curve after horizontal parabola fitting, the centre between P2, P3 adjacent 3 points Q1, Q2, Q3 is then taken, Q1, Q2, Q3 is asked to arrive
The distance of respective point is compared with iron mark on ideal curve, can judge whether there is hurt;
Step 2: if 3 points of P1, P2, P3 point-blank or it is approximate point-blank, take by P2, P3 point it
Between line segment as ideal curve, remaining is calculated with step 1;
Step 3: if the slope of the connected straight line of P2, P3 is smaller, the centre between P2, P3 adjacent 3 points Q1, Q2, Q3 are taken,
The vertical distance and iron mark for asking Q1, Q2, Q3 and ideal curve respectively compare, to judge whether there is hurt;
This method is in application, when judgement is lost, and encoder records the position of the hurt, according to iron mark TB/
The rail profile scanned is removed overlay region, is sentenced using the above method by T3276-2011, TB/T2344-2012 hurt regulation
Disconnected Rail Surface whether there is hurt, and acquired results repeatability omission factor is low, detect high stability.
The present invention has the advantages that below main compared with prior art:
1. the method using Parabolic Fit is fitted Rail Surface scan data point, more with the actual profile shape of rail
It is close, be conducive to improve hurt detection accuracy;
2. can determine hurt position, targetedly direction is provided for rail subsequent maintenance and maintenance;
3. replacing artificial detection using high-precision sensor, the labor intensity of artificial detection is alleviated, is advantageously implemented certainly
Dynamicization measurement.
Detailed description of the invention
Fig. 1 is the structural schematic diagram of Rail Surface hurt detection device of the present invention.
Fig. 2 is Rail Surface hurt of the present invention detection acquisition schematic diagram data.
Fig. 3 is the effect picture of Rail Surface hurt of the present invention detection.
In figure: 1. first sensors, 2. second sensors, 3. 3rd sensors, 4. the 4th sensors, 5. sensor branch
Frame, 6. pedestals, 7. orbit transports rollers, 8. rail, 9. encoder brackets, 10. encoders.
Specific embodiment
Below with reference to examples and drawings, the invention will be further described, but does not limit the present invention.
The present invention provides a kind of rail defects and failures detection device and method, and device is realized according to Fig. 1, which utilizes line
Structured light sensor acquires the surface data of rail (8), and encoder (10) is mounted at the center 65mm away from sensor stand (5)
Position, avoid encoder from blocking sensor and the data of rail foot surface acquired.
The device include 4 laser sensors, sensor stand (5), pedestal (6), orbit transports roller (7), rail (8) and
Encoder bracket (9), encoder (10), in which: four sensors are line structure optical sensor, and use same model, they
It is separately mounted at four interior angles of sensor stand (5), and the Z axis of each sensor coordinate system intersects at sensor stand
Center, sensor stand is fixed on the pedestal (6) of the detection device, while orbit transports roller (7) are fixed on pedestal, use
It supports and conveys rail (8), encoder bracket (9) are fixed on pedestal, for installing encoder (10).Due to four sensings
The installation site relationship of device, four collected data of sensor as shown in Fig. 2, there are repeated acquisition regions, therefore need to first by
The repeat region removes.
Rail defects and failures detection method provided by the invention, before carrying out rail defects and failures detection using the data, first according to steel
The characteristics of rail profile, there are blind areas for the data of each sensor acquisition, therefore it is two segment datas that a sensor is collected, and
And due to sensor mounting location feature, there are overlapping regions to remove at the midpoint of overlapping region search overlay region in the presence of overlapping
The overlapping ends of two segment data of area, to improve detection efficiency and accuracy.Before carrying out hurt Data Detection, to the head of every segment data
Tail respectively removes 5 points, and to avoid the random disturbances at endpoint, there are turning point, two end datas of turning point to be hurt for rail profile
It easily judges by accident, therefore need to be segmented data at turning point when damage detection, be to turn by looking for radius of curvature smallest point
Break.
Rail defects and failures detection method provided by the invention, specific practice the following steps are included:
Three point P1, P2, P3 of step 1 takes that treated one piece of data are spaced 18 points between two o'clock, carry out to this 3 points
Parabolic Fit takes adjacent 3 points Q1, Q2, Q3 between P2, P3, the midpoint of Q2 P2, P3, Q1 if fit line is curve
With Q3 respectively with Q2 close to asking Q1, Q2, Q3 to the distance S of curve P1P3 respectively1, S2, S3, in tyre tread data segment, compare
S1, S2, S3Compare S in other data segments with T11, S2, S3Compare S in character zone data segment with T21, S2, S3
With T3, hurt can be judged whether there is;
If the fit line that 3 points of step 2 is straight line, but when the straight line differs larger with horizontal line slope, Q1, Q2, Q3 is asked to arrive
The distance S of straight line P1P31, S2, S3, in tyre tread data segment, compare S1, S2, S3Compare with T1 in other data segments
S1, S2, S3Compare S in character zone data segment with T21, S2, S3With T3, hurt can be judged whether there is;If step 3
3 points of fit line is parabola, but the slope between P2, P3 is smaller, asks tangent with Q1, Q2, Q3 respectively and is parallel to P2P3's
Straight line, in tyre tread data segment, compares S at a distance from straight line P2P31, S2, S3Compare with T1 in other data segments
S1, S2, S3Compare S in character zone data segment with T21, S2, S3With T3, hurt can be judged whether there is;
If step 4 P1, P2, P3 are point-blank, and slope is smaller, then judges Q1, Q2, Q3 and fitting a straight line at 3 points
Vertical distance S1, S2, S3, judge whether to be hurt;
As shown in figure 3, being Rail Surface hurt detection effect schematic diagram, detect to have at the right base angle at the position scarce
Mouthful, rail to be detected is placed on rail carrying roller, and is opened with the speed uniform motion of 300mm/s when rail is located at detection zone
Beginning surface sweeping Rail Surface data, and make hurt detection, when detecting hurt, data are automatically saved, and save as a result, result figure
As shown in Figure 3.
Claims (7)
1. a kind of Rail Surface hurt detection device and method, it is characterized in that including four sensors, sensor stand (5), bottom
Seat (6), orbit transports roller (7), rail (8) and encoder bracket (9), encoder (10), in which: four sensors are cable architecture
Optical sensor, they are separately mounted at four interior angles of sensor stand (5), and the Z axis of each sensor coordinate system is homogeneous
The center of sensor stand is met at, sensor stand is fixed on the pedestal (6) of the detection device, while rail is fixed on pedestal
Road conveying roller (7) is fixed with encoder bracket (9) for supporting and conveying rail (8) on pedestal, is installed encoder (10), is compiled
The length of the center at least code wheel radius of code device center deviation sensor stand.
2. Rail Surface hurt detection device according to claim 1, it is characterized in that four sensors, are all made of
The linear laser distance measuring sensor of same model.The specifically configured laser parameter of sensor, so that sensor is not mutually dry
It disturbs.Encoder uses incremental encoder, and diameter 120mm, when rail movement, rail foot surface is tangent with code wheel, and band
Dynamic code wheel rotation, the distance of rail operation is obtained by the reading of encoder.
3. a kind of Rail Surface hurt detection method, it is characterized in that acquiring rail profile using four linear laser distance measuring sensors
Outline data simultaneously carries out data preparation before analysis.The characteristics of due to rail profile, there are blind for the data of each sensor acquisition
Area, but the data of all the sensors acquisition can constitute complete rail profile profile altogether.And because sensor mounting location causes
Data overlap, it is necessary to remove the data of overlapping.Furthermore in order to avoid the ideal surfaced curve that is fitted in subsequent calculating across
Data are less than at threshold value at i.e. turning point in radius of curvature and are segmented by rail corner angle.
4. a kind of Rail Surface hurt detection method according to claim 3, it is characterized in that may be present to web of the rail section
Convex or recessed character area is especially handled.It is located on the circular arc of radius R=400mm due to character area, takes the circle at the web of the rail
Two o'clock can find center of circle O on arc, then intercept L point (x of the circular arc middle section0, y0), (x1, y1) ... ... (xl, yl), by
Point calculates each point and O point distance S0, S1... ... SlIf having more than 60% R-Si(standard character is a height of by >=0.7mm
1.0mm, but the high error of character is larger), then judge that scanning position enters character area.When there is no characters in continuous 40 scan lines
When then judge scan position gone out character area.When scanning position in character zone, it is biggish to set a suitable opposite iron mark
Threshold value judges hurt.
5. a kind of Rail Surface hurt detection method according to claim 3, it is characterised in that: in view of the profile of rail
It is that circular arc and straight line link up smoothly, scan data point is fitted as its ideal surfaced curve, steel is then calculated
Data point is judged as hurt if the distance is more than iron mark permissible value to ideal surfaced curve corresponding points distance on track surface.
Specific step is as follows:
Step 1: tactic three point P1, P2, P3 at equal intervals in the data segment after removing overlay region throws at this 3 points
It is used as ideal curve after the fitting of object line, then takes the centre between P2, P3 adjacent 3 points Q1, Q2, Q3, asks Q1, Q2, Q3 to ideal
The distance of respective point is compared with iron mark on curve, can judge whether there is hurt;
Step 2: if 3 points of P1, P2, P3 point-blank or it is approximate point-blank, take by between P2, P3 point
As ideal curve, remaining is calculated with step 1 line segment;
Step 3: if the slope of the connected straight line of P2, P3 is smaller, taking the centre between P2, P3 adjacent 3 points Q1, Q2, Q3, respectively
The vertical distance and iron mark for asking Q1, Q2, Q3 and ideal curve compare, to judge whether there is hurt.
6. a kind of Rail Surface hurt detection method according to claim 5, it is characterised in that: use hurt zoning
P1, P2, P3 are indicated, judge centre adjacent three point Q1, Q2, Q3 between P2, P3 with the presence or absence of hurt.If not finding hurt,
Then by P1, P2, P3 three data points of synchronous forward movement, then repeatedly hurt is calculated, until data segment endpoint.If calculating
Middle discovery hurt, then on the scanning section hurt calculate stop, and export the scanning section rail length direction position (by
Encoder count determines).
7. a kind of Rail Surface hurt detection method according to claim 5, it is characterised in that:, should when finding hurt
Device will stop rail movement other than the position of rail length direction in addition to output Current Scan section, and issue sound-light alarm.
The device will also record Current Scan section profile and several subsequent section profiles.In user query, which can be shown
Show the 3 D stereo view of these profiles.It scans section profile number and 3 D stereo view viewpoint can be in a configuration text
It modifies and determines in part.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710401859.7A CN108982507A (en) | 2017-05-31 | 2017-05-31 | A kind of Rail Surface hurt detection device and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710401859.7A CN108982507A (en) | 2017-05-31 | 2017-05-31 | A kind of Rail Surface hurt detection device and method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108982507A true CN108982507A (en) | 2018-12-11 |
Family
ID=64502407
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710401859.7A Pending CN108982507A (en) | 2017-05-31 | 2017-05-31 | A kind of Rail Surface hurt detection device and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108982507A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110210496A (en) * | 2019-05-24 | 2019-09-06 | 武汉理工大学 | Rail pressed characters extraction element and method under a kind of working condition |
CN110609037A (en) * | 2019-07-12 | 2019-12-24 | 北京旷视科技有限公司 | Product defect detection system and method |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101041359A (en) * | 2007-03-08 | 2007-09-26 | 重庆大学 | Method for measurement of railcar stroke encoding wheel |
CN101055475A (en) * | 2006-11-10 | 2007-10-17 | 重庆大学 | Coding wheel positioning device for commuting of track vehicle and its controller |
CN200969115Y (en) * | 2006-11-10 | 2007-10-31 | 重庆大学 | Rail vehicle roundtrip walking multi-point positioning apparatus |
CN101639452A (en) * | 2009-09-11 | 2010-02-03 | 北京科技大学 | 3D detection method for rail surface defects |
CN101893580A (en) * | 2010-06-10 | 2010-11-24 | 北京交通大学 | Digital image based detection method of surface flaw of steel rail |
EP3138754A1 (en) * | 2015-09-03 | 2017-03-08 | Rail Vision Europe Ltd | Rail track asset survey system |
CN106546176A (en) * | 2015-09-21 | 2017-03-29 | 武汉汉宁科技有限公司 | A kind of rail outward appearance physical dimension and surface defect detection apparatus |
-
2017
- 2017-05-31 CN CN201710401859.7A patent/CN108982507A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101055475A (en) * | 2006-11-10 | 2007-10-17 | 重庆大学 | Coding wheel positioning device for commuting of track vehicle and its controller |
CN200969115Y (en) * | 2006-11-10 | 2007-10-31 | 重庆大学 | Rail vehicle roundtrip walking multi-point positioning apparatus |
CN101041359A (en) * | 2007-03-08 | 2007-09-26 | 重庆大学 | Method for measurement of railcar stroke encoding wheel |
CN101639452A (en) * | 2009-09-11 | 2010-02-03 | 北京科技大学 | 3D detection method for rail surface defects |
CN101893580A (en) * | 2010-06-10 | 2010-11-24 | 北京交通大学 | Digital image based detection method of surface flaw of steel rail |
EP3138754A1 (en) * | 2015-09-03 | 2017-03-08 | Rail Vision Europe Ltd | Rail track asset survey system |
CN106546176A (en) * | 2015-09-21 | 2017-03-29 | 武汉汉宁科技有限公司 | A kind of rail outward appearance physical dimension and surface defect detection apparatus |
Non-Patent Citations (1)
Title |
---|
徐科等: "基于激光线光源的钢轨表面缺陷三维检测方法", 《机械工程学报》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110210496A (en) * | 2019-05-24 | 2019-09-06 | 武汉理工大学 | Rail pressed characters extraction element and method under a kind of working condition |
CN110609037A (en) * | 2019-07-12 | 2019-12-24 | 北京旷视科技有限公司 | Product defect detection system and method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107632022A (en) | A kind of detection method of surface flaw of steel rail and device based on data processing | |
CN107200042B (en) | High-precision on-line detection method and detection device for diameter and roundness abrasion of train wheels | |
EP2872432B1 (en) | Guide rail straightness measuring system for elevator installations | |
CN109708615A (en) | A kind of subway tunnel limit dynamic testing method based on laser scanning | |
CN106091951B (en) | A kind of municipal rail train wheel rim parameter on-line detecting system and method | |
CN112172862A (en) | Multifunctional track detection system | |
CN107091616B (en) | Abnormal tunnel lining deformation monitoring method | |
CN108982507A (en) | A kind of Rail Surface hurt detection device and method | |
CN112883997A (en) | Rail transit fastener detection system and detection method | |
CN105261025A (en) | Line-scan camera fast and high-precision calibration device of high-speed rail detection system | |
CN109696134A (en) | Wheel diameters detection method and system based on two-dimensional laser location detecting technology | |
CN107117188B (en) | The vertically arranged tramcar wheel footpath online test method of linear sensor | |
CN108839674B (en) | Online dynamic measurement device and measurement method for geometric parameters of train wheels | |
CN106184284B (en) | Railway wheelset diameter automatic measurement method and system based on the scanning of line laser multi-section | |
CN214029539U (en) | Device for adjusting position of outer ring belt head of steel coil | |
CN111094104A (en) | Detecting and optimizing stop point accuracy for a vehicle | |
CN203601294U (en) | Urban rail vehicle wheel diameter detecting device with sensors installed linearly and obliquely | |
CN205140009U (en) | Quick high accuracy calibration device of linear array camera | |
CN108839676B (en) | Online dynamic measurement device and measurement method for geometric parameters of train wheels | |
CN103587552A (en) | Device and method for detecting urban rail vehicle wheel diameters through straight line slanting installation of sensors | |
RU78927U1 (en) | INSTALLATION FOR AUTOMATIC MEASUREMENT OF GEOMETRIC PARAMETERS OF RAILWAYS OF WHOLE-ROLLED WHEELS IN THE PRODUCTION FLOW | |
CN203605915U (en) | Urban rail vehicle wheel diameter detector with sensor installed along arc normal | |
CN207510444U (en) | A kind of train wheel diameter wears away high-precision on-line measuring device with circularity | |
CN205440401U (en) | Train wheel set size dynamic inspection system's basic platform | |
CN108891444B (en) | Device and method for dynamically measuring geometric parameters of train wheels on line |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181211 |