CN111487189A - Tread damage automatic detection system - Google Patents
Tread damage automatic detection system Download PDFInfo
- Publication number
- CN111487189A CN111487189A CN202010259481.3A CN202010259481A CN111487189A CN 111487189 A CN111487189 A CN 111487189A CN 202010259481 A CN202010259481 A CN 202010259481A CN 111487189 A CN111487189 A CN 111487189A
- Authority
- CN
- China
- Prior art keywords
- wheel
- tread
- detection
- driving
- rim
- 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
Images
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
- G01N21/01—Arrangements or apparatus for facilitating the optical investigation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
-
- 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
- G01M17/007—Wheeled or endless-tracked vehicles
- G01M17/013—Wheels
-
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- 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
- G01N21/01—Arrangements or apparatus for facilitating the optical investigation
- G01N2021/0106—General arrangement of respective parts
- G01N2021/0112—Apparatus in one mechanical, optical or electronic block
-
- 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
- G01N2021/8887—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 based on image processing techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Abstract
Tread damage automatic check out system relates to railway vehicle and detects technical field. The invention aims to solve the problems of large workload and low manual detection accuracy of wheel drop detection in the damage detection of the wheel tread in the prior art. According to the automatic tread damage detection system, high-definition images are obtained by continuously shooting treads, the visible parts of the wheel treads of the motor train unit are detected by using the technologies such as image processing and the like, detection items comprise scratches, bulges, sunken shapes, sizes and the like, and the phenomenon of missed detection is avoided to the greatest extent; meanwhile, the collected images are put into a processing device for storage and filing so as to be checked and examined manually.
Description
Technical Field
The invention belongs to the technical field of railway vehicle detection.
Background
A wheel is a rotating member that bears the load between the tire and the axle, and is typically composed of a rim and spokes. The rim, commonly known as a rim or hub, is an intermediate component that holds the tire and connects the tire to the axle. The wheel tread is the contact part of the wheel and the top surface of the steel rail. The wheels continuously rub the steel rail in the running process, so that the diameters of the wheels are continuously reduced, and the wheel tread is abraded; in addition, when the wheel passes through a curve or a turnout, the rim part of the wheel rubs against the inner side surface of the rail to cause rim abrasion. Tread wear and rim wear cause the wheel dimensions to change, greatly affecting ride comfort and ride stability. The brake shoe is locked when the vehicle is braked emergently or is not released when the vehicle is started, so that the tread is scratched. When the vehicle runs, the peak value of the wheel-rail load caused by tread scratch is related to the running speed of the vehicle, the scratch depth of the tread and the scratch wavelength and can exceed 400KN sometimes, so that the generation of fatigue cracks of the vehicle and the track parts can be accelerated by the operation of large impact load on the wheels and the track parts, accidents are caused, and the direct threat to the driving safety is caused.
With the high-speed development of China railways and the continuous improvement of the speed of the train, the timeliness and controllability of the detection of the shapes of the wheel rim and the wheel tread of a train wheel are particularly important. The detection to the wheel generally all mainly relies on the check out test set to detect, need separate wheel and box moreover and just can detect (wheel fall detects) to cause the big problem of detection achievement volume. Or manual detection is adopted, which results in longer time consumption and lower efficiency of the whole detection process, and the low accuracy of the manual detection easily causes missed detection.
Disclosure of Invention
The invention provides an automatic tread damage detection system for solving the problems of large workload and low manual detection accuracy of wheel drop detection in the damage detection of the existing wheel tread.
The automatic tread damage detection system comprises a walking device, an image acquisition device and a processing device, wherein the image acquisition device is fixed on the walking device, and the image acquisition device and the processing device realize data interaction in a wireless transmission mode;
the walking device can be adsorbed on the tread of the wheel to be tested in a magnetic adsorption mode and drives the image acquisition device to move along the outer circumference of the wheel to be tested;
the image acquisition device includes: the device comprises a plurality of endoscopes and reflectors, wherein the endoscopes are arranged in parallel in a straight line, the central axes of the endoscopes are parallel to each other, the included angle between the mirror surface of each reflector and the central axes of the endoscopes is 45 degrees, the endoscopes are all used for collecting images in the reflectors, and the included angle between each reflector and the tread of a wheel to be detected is 45 degrees;
the processing device comprises the following software-implemented units:
a training data set establishing unit: collecting wheel pictures of different types of railway trucks under different time, places and environments to establish a sample library, wherein the wheel pictures comprise wheel pictures under a normal state and wheel pictures under a fault state, dividing each picture into 8-10 areas, marking whether a fault exists in each area, generating a corresponding label file, and establishing a training data set of a Faster rcnn target detection network by using all the wheel pictures and the corresponding label files;
a weight training unit: substituting the training data set into a Faster rcnn target detection network model, training the Faster rcnn target detection network model to obtain a weight value of the Faster rcnn target detection network model, and substituting the weight value into the Faster rcnn target detection network model to finish model training;
the picture acquisition unit: collecting a wheel picture to be identified as a detection picture;
a fault recognition unit: and inputting the detection picture into a trained Faster rcnn target detection network model to obtain a recognition result.
The image acquisition device also comprises a plurality of L ED lamps, and a plurality of L ED lamps are uniformly distributed around the reflector.
The above-mentioned running gear includes: the adsorption equipment, carry on main part and drive device, drive device and adsorption equipment all fix on carrying on the main part, carry on the main part and can adsorb on the tread of wheel under test through adsorption equipment, drive device can drive and carry on the main part and follow the tread motion of wheel under test.
Above-mentioned adsorption equipment is the permanent magnet, the carrying main part is the casing, drive arrangement includes the battery, drive wheelset and supplementary wheelset, the battery is used for supplying power for drive wheelset, drive wheelset installs on one side inner wall of casing, and drive wheelset can rotate along the tread of being surveyed the wheel under the drive of battery, supplementary wheelset is installed on the casing, form between the rim of supplementary wheelset and being surveyed the wheel and the tread and roll, the permanent magnet is installed on the casing, make the casing adsorb on being surveyed the wheel.
The shell comprises two wheel side end face frame bodies, a wheel rim frame body and a tread frame body, wherein the wheel rim frame body and the tread frame body are sequentially connected and horizontally arranged, the left end of the wheel rim frame body and the right end of the tread frame body are respectively connected through a vertically arranged wheel side end face frame body, and the two wheel side end face frame bodies, the wheel rim frame body and the tread frame body are of an integrated structure.
The driving wheel set comprises a motor and a rubber wheel, the motor is installed in the wheel side end face frame body, the motor is powered by a storage battery, and the rubber wheel is connected with the output end of the motor.
According to the automatic tread damage detection system, high-definition images are obtained by continuously shooting treads, the visible parts of the wheel treads of the motor train unit are detected by using the technologies such as image processing and the like, detection items comprise scratches, bulges, sunken shapes, sizes and the like, and the phenomenon of missed detection is avoided to the greatest extent; meanwhile, the collected images are put into a processing device for storage and filing so as to be checked and examined manually.
Drawings
FIG. 1 is a state diagram of an automatic tread damage detection system installed on a wheel to be detected;
FIG. 2 is a schematic view of the internal structure of the automatic tread damage detection system according to the present invention;
FIG. 3 is a flow chart of the internal processing of the processing apparatus;
FIG. 4 is a side view of the automatic tread damage detection system of the present invention.
Detailed Description
The first embodiment is as follows: the tread damage automatic detection system comprises a walking device, an image acquisition device and a processing device, wherein the image acquisition device is fixed on the walking device, and the image acquisition device and the processing device realize data interaction in a wireless transmission mode.
The walking device comprises: the device comprises adsorption equipment, a carrying main body and driving equipment, wherein the driving equipment and the adsorption equipment are both fixed on the carrying main body, the carrying main body can be adsorbed on a tread of a wheel 9 to be tested through the adsorption equipment, and the driving equipment can drive the carrying main body to move along the tread of the wheel 9 to be tested.
The image acquisition device comprises a plurality of L ED lamps 8, a plurality of endoscopes 7 and a reflector 6, wherein the plurality of endoscopes 7 are arranged in parallel in a straight line, the central axes of the plurality of endoscopes 7 are parallel to each other, the included angle between the mirror surface of the reflector 6 and the central axes of the plurality of endoscopes 7 is 45 degrees, the plurality of endoscopes 7 are all used for acquiring images in the reflector 6, the included angle between the reflector 6 and the tread of a wheel 9 to be detected is 45 degrees, and the plurality of L ED lamps 8 are uniformly distributed around the reflector 6 and used for light supplement.
The processing device comprises the following software-implemented units:
a training data set establishing unit:
the method comprises the steps of collecting wheel pictures of different types of railway wagons in different time, places and environments from a network big database or an actual application environment, then respectively carrying out stretching, rotation and mirror image transformation on each collected picture, and establishing a sample library by utilizing all pictures before and after transformation, wherein the wheel pictures comprise wheel pictures in a normal state and wheel pictures in a fault state. For example: the method comprises the steps of collecting a wheel picture, copying three same pictures, stretching, rotating and carrying out mirror image transformation on the three copied pictures respectively to obtain a stretched picture, a rotated picture and a mirrored picture, and storing an original picture and the three transformed pictures into a sample library together. The operation aims at amplifying the sample, collecting the car coupler images under different conditions is beneficial to enriching the sample data, and the robustness and the adaptability of the subsequent training result are improved.
And respectively dividing each picture into 8-10 areas, marking whether a fault exists in each area, generating a corresponding label file, and establishing a training data set of the Faster rcnn target detection network by using all the wheel pictures and the corresponding label files.
A weight training unit:
the Inception v2 model has higher running speed relative to a feature extraction network such as renet. Therefore, in this embodiment, the Faster rcnn target detection network is trained using the inclusion v2 pre-training network model.
Specifically, the training data set is substituted into an increment v2 pre-training network model, the whole image in the training data set is input into an increment v2 for feature extraction, and an RPN is used for generating a suggestion window. In order to increase the operation speed of the network, the embodiment reduces the number of the recommended windows to 100 for training.
And obtaining a weight value of the Faster rcnn target detection network model through the training, and substituting the weight value into the Faster rcnn target detection network model to finish model training.
After model training is completed, the trained model is optimized by using tensorRT (inference optimizer), and the prediction speed of the network is improved. Data type accuracy in the prediction process is reduced by combining and replacing some network layers by the tensorRT, and the FP32 is changed into the FP 16.
The picture acquisition unit:
high-definition imaging devices are arranged on the two sides and the center of a rail of the truck, when the truck passes through a device installation position, a wheel picture to be identified can be obtained as a detection picture, and the sizes of all the detection pictures are unified;
a fault recognition unit:
and inputting the detection picture into a trained Faster rcnn target detection network model to obtain a recognition result.
In particular, in practical application, each truck has 8 wheels. Therefore, in the identification process, tens of wheel images of the whole train are all read into the memory in parallel. The wheel images of each truck are then fused into a matrix of size (8,512, 3) and then input into the network for identification.
Because of the large image size of the wheel and the small failure of the wheel, the image size cannot be reduced to preserve the wheel rim and tread damage characteristics, and the number of wheels in a train typically exceeds 450. Therefore, the rim and tread damage faults are detected in the wheel image by adopting the deep learning Faster rcnn. Compared with the target detection networks such as SSD, yolo and the like, Faster rcnn has higher accuracy and recognition precision. Because the efficiency of the Faster rcnn network in the process of predicting images is low, the tensorRT is adopted to accelerate the prediction process of the Faster rcnn network, and the prediction speed of the network is improved.
In practical application, the automatic tread damage detection system is firstly placed on the wheel 9 to be detected, and the power switch is turned on to power on the system. After the detection system starts normal connection, a user can press a start key of the image acquisition device, and high-definition images are obtained by continuously shooting the tread. The image collected by the image collecting device is transmitted to the processing device in real time, and when the image is collected, the image collecting device stops transmitting the image and enters a standby mode. The processing device carries out fault identification on the received image data, displays the fault, gives an alarm and outputs a prompt, and can quickly find out an alarm area if the received image data is abnormal.
The second embodiment is as follows: referring to fig. 4 to explain this embodiment in detail, this embodiment is to further explain the tread damage automatic detection system described in the first embodiment, in this embodiment, the adsorption device is a permanent magnet 5, the carrying main body is a casing 1, the driving device includes a storage battery 2, a driving wheel set 3 and an auxiliary wheel set 4, the storage battery 2 is installed at one side of the casing 1, the driving wheel set 3 is installed on an inner wall of one side of the casing 1, the driving wheel set 3 rotates along a side end face of a train wheel under the driving of the storage battery 2, the auxiliary wheel set 4 is installed on the casing 1, the auxiliary wheel set 4 forms a pair of rollers with the rim and the tread of the train wheel, and the permanent magnet 5 is installed on the casing 1 to ensure the distance between the casing and the train wheel.
The embodiment adopts the technology of carrying out adsorption walking on the wheel, provides an installed carrier for the detection equipment of the wheel rim and the tread, and can ensure the detection of the train wheel to be detected without falling to the ground. The wheel rim and tread detection equipment is arranged on the wheel rim frame body 1-2 and the tread frame body 1-3, so that the wheel rim frame body, the tread frame body and the carried detection equipment can accurately move in a relatively narrow space, and the accurate static detection of the wheel rim and the tread in a wheel rotating state is ensured.
The case 1 of the present embodiment is a thin-walled case and is made of a resin material. This embodiment employs a lightweight structural design. The shell structure adopts a thin-wall design, so that the weight of the shell can be effectively reduced, and the weight of the carrying equipment is increased; the material adopts novel resin material, has realized light structure high rigidity. The shell 1 is an arc-shaped shell, and the shape of the shell 1 is matched with that of a train wheel. So set up, the casing appearance adopts the circular arc appearance similar with wheel rim and tread, make full use of effective space, reduces overall dimension.
Further, the casing 1 of the present embodiment includes two wheel side end face frame bodies 1-1, a wheel rim frame body 1-2 and a tread frame body 1-3, the wheel rim frame body 1-2 and the tread frame body 1-3 are connected in sequence and horizontally arranged, the left end of the wheel rim frame body 1-2 and the right end of the tread frame body 1-3 are respectively connected through one wheel side end face frame body 1-1 which is vertically arranged, and the two wheel side end face frame bodies 1-1, the wheel rim frame body 1-2 and the tread frame body 1-3 are integrated. So set up, simple structure is convenient for manufacture.
The wheel rim support body 1-2 is an arc support body, and the shape of the wheel rim support body 1-2 is matched with the shape of a wheel rim of a train wheel. So set up, be convenient for utilize the space between casing and the train wheel, improve the accuracy of detecting.
The number of the driving wheel sets 3 of the embodiment is preferably 2, the driving wheel sets are symmetrically arranged in the frame bodies 1-1 on the side end faces of the wheels, the permanent magnets 5 of the embodiment are arranged on the tread frame bodies 1-3, the arrangement is convenient for ensuring the distance between the shell and the wheels of the train for providing suction force, and particularly, the number of the permanent magnets 5 is 8, the arrangement is characterized in that 8 permanent magnets are firstly determined according to the shape, the mass distribution, the relative distance with the wheels, the movement characteristics and the like of the traveling device, the suction force of the magnets is about 3-4 kg according to the theoretical mass of the traveling device and the loaded equipment, and a simple calculation formula is adopted, the magnet suction force of the magnets is calculated by the magnet volume × and the magnet density ×, a neodymium boron magnet with the magnet suction force of about 3.6kg is selected, the shape size of 20mm × mm × mm, when the traveling device and the device are actually used, the distance between the lower end face of the shell 1 and the wheel rim and the tread of the wheel of the train wheel is 0.8-1, the walking device and the wheel, the gap between the wheel of the wheel and the wheel of the wheel can be successfully detected under the action of the walking device and the gap between the wheel of the wheel.
The auxiliary wheel set 4 of the present embodiment is mounted on the rim frame body 1-2 and the tread frame body 1-3 of the housing 1. So set up, guarantee the balance of casing 1 when using, guarantee the equidistant arrangement of casing 1, the both ends at rim support body and tread support body are installed respectively to supplementary wheelset, correspond the rim and the tread and the wheel side position of wheel respectively, make even the exerting of equipment weight that rim support body and tread support body and carry on it on supplementary wheelset to realize quality evenly distributed, realize the balance.
The third concrete implementation mode: in the embodiment, the driving wheel set 3 comprises a motor 3-1 and a rubber wheel 3-2, the motor 3-1 is installed in a wheel side end face frame 1-1, the motor 3-1 is powered by a storage battery 2, and the rubber wheel 3-2 is connected with an output end of the motor 3-1. So set up, simple structure is convenient for realize.
When the tested wheel 9 walks on the embodiment, the specific working process is as follows:
firstly, the rim and tread detection device and the walking device are connected and fixed firmly through bolts and reserved connecting holes, connecting wires are placed in the wiring grooves, and vulnerable electrical elements of the detection device are protected in the shell through the cover plate, so that the detection device and the walking device are integrated. The whole structure is tightly attached to the wheel by pressing the whole structure on the wheel, and then a power switch on the shell of the device is turned on, so that the shell device sends out a wireless signal and is in butt joint with the shell device through the handheld device. The handheld equipment sends out signals to enable the wheel rim and the tread detection device to move along the wheel, when the wheel runs to the bottommost part of the wheel, the wheel rim and the tread detection device are controlled to run in the opposite direction through the proximity switch, the detection device is started to detect at the same time until the wheel rim and the tread detection device run to the contact position of the other side of the wheel and the steel rail, the movement is stopped, and detection data are sent, so that wheel rim and tread detection results of the train wheel are obtained.
Claims (6)
1. The automatic tread damage detection system is characterized by comprising a walking device, an image acquisition device and a processing device, wherein the image acquisition device is fixed on the walking device, and the image acquisition device and the processing device realize data interaction in a wireless transmission mode;
the walking device can be adsorbed on the tread of the wheel (9) to be tested in a magnetic adsorption mode and drives the image acquisition device to move along the outer circumference of the wheel (9) to be tested;
the image acquisition device includes: the device comprises a plurality of endoscopes (7) and a reflective mirror (6), wherein the plurality of endoscopes (7) are arranged in parallel in a straight line, the central axes of the plurality of endoscopes (7) are parallel to each other, the included angle between the mirror surface of the reflective mirror (6) and the central axes of the plurality of endoscopes (7) is 45 degrees, the plurality of endoscopes (7) are all used for collecting images in the reflective mirror (6), and the included angle between the reflective mirror (6) and the tread of a wheel (9) to be detected is 45 degrees;
the processing device comprises the following software-implemented units:
a training data set establishing unit: collecting wheel pictures of different types of railway trucks under different time, places and environments to establish a sample library, wherein the wheel pictures comprise wheel pictures under a normal state and wheel pictures under a fault state, dividing each picture into 8-10 areas, marking whether a fault exists in each area, generating a corresponding label file, and establishing a training data set of a Faster rcnn target detection network by using all the wheel pictures and the corresponding label files;
a weight training unit: substituting the training data set into a Faster rcnn target detection network model, training the Faster rcnn target detection network model to obtain a weight value of the Faster rcnn target detection network model, and substituting the weight value into the Faster rcnn target detection network model to finish model training;
the picture acquisition unit: collecting a wheel picture to be identified as a detection picture;
a fault recognition unit: and inputting the detection picture into a trained Faster rcnn target detection network model to obtain a recognition result.
2. The system according to claim 1, wherein the image capturing device further comprises a plurality of L ED lights (8), the plurality of L ED lights (8) being evenly distributed around the reflector (6).
3. The automated tread damage detection system of claim 1, wherein the walking device comprises: the device comprises adsorption equipment, a carrying main body and driving equipment, wherein the driving equipment and the adsorption equipment are fixed on the carrying main body, the carrying main body can be adsorbed on a tread of a wheel (9) to be tested through the adsorption equipment, and the driving equipment can drive the carrying main body to move along the tread of the wheel (9) to be tested.
4. The automatic tread damage detection system according to claim 3, wherein the adsorption device is a permanent magnet (5), the carrying body is a casing (1), the driving device comprises a battery (2), a driving wheel set (3) and an auxiliary wheel set (4),
the storage battery (2) is used for supplying power to the driving wheel set (3), the driving wheel set (3) is installed on the inner wall of one side of the shell (1), the driving wheel set (3) can rotate along the tread of the detected wheel (9) under the driving of the storage battery (2), the auxiliary wheel set (4) is installed on the shell (1), the auxiliary wheel set (4) and the rim and the tread of the detected wheel (9) form a pair of rollers, the permanent magnet (5) is installed on the shell (1), and the shell (1) can be adsorbed on the detected wheel (9).
5. The automatic tread damage detection system according to claim 4, wherein the housing (1) comprises two wheel side end face frame bodies (1-1), a rim frame body (1-2) and a tread frame body (1-3),
the wheel rim support body (1-2) and the tread support body (1-3) are sequentially connected and horizontally arranged, the left end of the wheel rim support body (1-2) and the right end of the tread support body (1-3) are respectively connected through a wheel side end face support body (1-1) which is vertically arranged, and the two wheel side end face support bodies (1-1), the wheel rim support body (1-2) and the tread support body (1-3) are of an integrated structure.
6. Automatic tread damage detection system according to claim 4 or 5, characterized in that the driving wheel set (3) comprises a motor (3-1) and a rubber wheel (3-2),
the motor (3-1) is arranged in the end face frame body (1-1) on the side of the wheel, the motor (3-1) is powered by the storage battery (2), and the rubber wheel (3-2) is connected with the output end of the motor (3-1).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010259481.3A CN111487189A (en) | 2020-04-03 | 2020-04-03 | Tread damage automatic detection system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010259481.3A CN111487189A (en) | 2020-04-03 | 2020-04-03 | Tread damage automatic detection system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111487189A true CN111487189A (en) | 2020-08-04 |
Family
ID=71792453
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010259481.3A Pending CN111487189A (en) | 2020-04-03 | 2020-04-03 | Tread damage automatic detection system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111487189A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113643288A (en) * | 2021-10-13 | 2021-11-12 | 启东市德立神起重运输机械有限公司 | Crane wheel control method based on artificial intelligence |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102525397A (en) * | 2010-11-03 | 2012-07-04 | Tyco医疗健康集团 | Mirrored arthroscope |
| CN202625455U (en) * | 2012-05-29 | 2012-12-26 | 南京文采科技有限责任公司 | Imaging type margin detection and control device of bobbin paper |
| CN205507217U (en) * | 2016-03-25 | 2016-08-24 | 温州市鹿城区中津先进科技研究院 | A plug -in type mobile phone microscope for oral cavity is detected |
| CN109255805A (en) * | 2018-08-23 | 2019-01-22 | 苏州富鑫林光电科技有限公司 | The industrial intelligent data gathering system and method for machine learning |
| CN209191970U (en) * | 2018-12-17 | 2019-08-02 | 哈尔滨市科佳通用机电股份有限公司 | A kind of wheel running gear carrying train wheel detection device |
| CN110162649A (en) * | 2019-05-24 | 2019-08-23 | 北京百度网讯科技有限公司 | Sample data acquisition methods obtain system, server and computer-readable medium |
| CN110276297A (en) * | 2019-06-21 | 2019-09-24 | 中国神华能源股份有限公司 | Train tyre tread wear detecting system and method |
| US20200055531A1 (en) * | 2017-02-22 | 2020-02-20 | Tetra Tech, Inc. | Broken Wheel Detection System |
| CN110823920A (en) * | 2019-11-07 | 2020-02-21 | 深圳市智能机器人研究院 | Device, system and method for collecting surface defects of inner hole side wall |
-
2020
- 2020-04-03 CN CN202010259481.3A patent/CN111487189A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102525397A (en) * | 2010-11-03 | 2012-07-04 | Tyco医疗健康集团 | Mirrored arthroscope |
| CN202625455U (en) * | 2012-05-29 | 2012-12-26 | 南京文采科技有限责任公司 | Imaging type margin detection and control device of bobbin paper |
| CN205507217U (en) * | 2016-03-25 | 2016-08-24 | 温州市鹿城区中津先进科技研究院 | A plug -in type mobile phone microscope for oral cavity is detected |
| US20200055531A1 (en) * | 2017-02-22 | 2020-02-20 | Tetra Tech, Inc. | Broken Wheel Detection System |
| CN109255805A (en) * | 2018-08-23 | 2019-01-22 | 苏州富鑫林光电科技有限公司 | The industrial intelligent data gathering system and method for machine learning |
| CN209191970U (en) * | 2018-12-17 | 2019-08-02 | 哈尔滨市科佳通用机电股份有限公司 | A kind of wheel running gear carrying train wheel detection device |
| CN110162649A (en) * | 2019-05-24 | 2019-08-23 | 北京百度网讯科技有限公司 | Sample data acquisition methods obtain system, server and computer-readable medium |
| CN110276297A (en) * | 2019-06-21 | 2019-09-24 | 中国神华能源股份有限公司 | Train tyre tread wear detecting system and method |
| CN110823920A (en) * | 2019-11-07 | 2020-02-21 | 深圳市智能机器人研究院 | Device, system and method for collecting surface defects of inner hole side wall |
Non-Patent Citations (2)
| Title |
|---|
| 曾大为: "基于深度学习的动车关键部位故障图像检测", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 闵永智等: "基于图像灰度梯度特征的钢轨表面缺陷检测", 《仪器仪表学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113643288A (en) * | 2021-10-13 | 2021-11-12 | 启东市德立神起重运输机械有限公司 | Crane wheel control method based on artificial intelligence |
| CN113643288B (en) * | 2021-10-13 | 2022-02-18 | 启东市德立神起重运输机械有限公司 | Crane wheel control method based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107433952B (en) | A kind of intelligent inspection robot | |
| CN105644572B (en) | A kind of walking contact net intelligent patrol detection equipment certainly | |
| CN202011397U (en) | Multifunctional continuous inspection type track inspection vehicle | |
| CN111005272B (en) | Track maintenance equipment based on suspension type magnetic suspension track traffic system | |
| CN101913367A (en) | Railway wheel tread scratch image dynamic detection device | |
| CN211491564U (en) | Electric power piping lane patrols and examines robot | |
| CN106143530A (en) | Routing inspection trolley for suspension type monorail traffic system box track girder | |
| CN111487189A (en) | Tread damage automatic detection system | |
| CN108374366A (en) | Highway patrol robot | |
| KR20160115887A (en) | Testing device and Testing methods for two-wheeld vehicle | |
| CN109459128A (en) | A kind of fuel tank of vehicle oil liquid whine noise test macro | |
| CN208602482U (en) | A kind of track detection device actively walked along amusement facility track | |
| CN2820643Y (en) | Intelligent online detecting system for train tread bruise | |
| CN104792542A (en) | Automobile brake performance detecting system calibrating apparatus and automobile brake performance detecting system calibrating method based on stereoscopic vision | |
| CN208291192U (en) | A kind of straddle-type monorail intelligence inflation system tire auto-location device | |
| CN206388242U (en) | A kind of new tire lettering detection means | |
| CN208254813U (en) | Quotient multiplies convertible vehicle automatic emergency brake system property test platform | |
| CN115092205B (en) | Inspection equipment applied to vehicle warehouse-in and warehouse-out detection system | |
| CN207894521U (en) | Remote vehicle emissions detection device | |
| CN206208223U (en) | A kind of detection means for double track flatness | |
| CN115817570A (en) | Rail breakage inspection device and inspection method based on single-rail self-balancing | |
| CN107042452B (en) | A kind of steel rail grinding lathe | |
| CN205819030U (en) | Mobile automobile detection engineering truck | |
| CN114114247A (en) | Tunnel gas and lining quality detection device | |
| CN110626384A (en) | Regular online detection system for motor train unit |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200804 |