Wheel tread inspection device

Abstract

To provide a wheel tread inspection device which reduces the data processing load for detecting damages on wheel treads of railroad vehicles.SOLUTION: A wheel tread inspection device is provided, comprising an acquisition unit for acquiring event data from event sensors configured to sense a wheel tread of a railroad vehicle and output event data indicative of brightness variation of each pixel, and a detection unit configured to detect damages on the wheel tread based on the event data acquired by the acquisition unit.SELECTED DRAWING: Figure 8

Description

[Technical field]

[0001] The present invention relates to a wheel tread inspection device. [Background technology]

[0002] Conventionally, there is known a wheel tread inspection device that captures images of the wheel tread of a railway vehicle using a camera and detects damage to the tread based on the image data output from the camera. For example, the technology of Patent Document 1 captures images of the wheel tread of a railway vehicle running on a track continuously using an image capture device installed beside the track to obtain multiple image data, and detects damage to the wheel from these multiple image data by image processing. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2020-148744 A Summary of the Invention [Problem to be solved by the invention]

[0004] In the above-mentioned conventional technology, image data captured by a camera is used to detect scratches on the wheel tread, and therefore a large amount of data needs to be processed for the scratch detection, resulting in a large load for the scratch detection data processing.

[0005] One of the problems that the embodiments of the present invention attempt to solve is to provide a wheel tread inspection device that can reduce the load of data processing in detecting scratches on the tread of railway vehicle wheels. [Means for solving the problem]

[0006] A wheel tread inspection device according to an embodiment of the present invention comprises an acquisition unit that acquires event data from an event sensor that senses the tread of a railway vehicle wheel and outputs event data indicating brightness changes for each pixel, and a detection unit that detects scratches on the tread based on the event data acquired by the acquisition unit. Effect of the Invention

[0007] A wheel tread inspection device according to an embodiment of the present invention can reduce the load of data processing in detecting defects on the wheel tread of a railway vehicle. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram illustrating a schematic installation example of a wheel tread inspection system according to an embodiment. [Diagram 2] FIG. 2 is a perspective view that illustrates a part of the wheel tread inspection system according to the embodiment and a wheel of a railway vehicle. [Diagram 3] FIG. 3 is a block diagram illustrating an example of a hardware configuration of the data processing device according to the embodiment. [Figure 4] FIG. 4 is a block diagram illustrating an example of a functional configuration of the data processing device according to the embodiment. [Diagram 5] FIG. 5 is a perspective view that illustrates a part of the wheel tread inspection system according to the embodiment and a reference wheel. [Figure 6] FIG. 6 is a diagram showing an example of a detection result of the event sensor according to the embodiment, in which a scratch is present on the wheel tread. [Figure 7] FIG. 7 is a flowchart illustrating an example of a threshold setting process executed by the data processing device of the embodiment. [Figure 8] FIG. 8 is a flowchart illustrating an example of a flaw detection process executed by the data processing device of the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Exemplary embodiments of the present invention are disclosed below. The configurations of the embodiments shown below, and the actions, results, and effects brought about by the configurations are merely examples. The present invention can be realized with configurations other than those disclosed in the following embodiments, and it is possible to obtain at least one of various effects and derivative effects based on the basic configurations.

[0010] <Wheel tread inspection system 1> Fig. 1 is a diagram showing a schematic diagram of an installation example of a wheel tread inspection system 1 according to this embodiment. As shown in Fig. 1, the wheel tread inspection system 1 according to this embodiment includes a plurality of event sensors 5A, 5B, lighting devices 6A, 6B, a detection device 7, and a data processing device 10. Hereinafter, the event sensor 5 will be used as a general term for the plurality of event sensors 5A, 5B. Furthermore, the lighting device 6 will be used as a general term for the plurality of lighting devices 6A, 6B.

[0011] The multiple event sensors 5A, 5B and the detection device 7 are set beside the track 300 and are arranged at intervals from each other. In the moving direction of the railway vehicle 200 traveling on the track 300 (the direction of the arrow F1 in FIG. 1), the event sensor 5A is arranged in front of the detection device 7, and the event sensor 5B is arranged above the event sensor 5A. In addition, a lighting device 6 is provided for each event sensor 5 and fixed to the corresponding event sensor 5. The data processing device 10 is installed at an arbitrary location. The data processing device 10 is connected to the event sensor 5 and the detection device 7 so as to be able to communicate with them. The data processing device 10 is an example of a wheel tread inspection device.

[0012] 2 is a perspective view showing a part of a wheel tread inspection system 1 according to an embodiment and a wheel 21 of a railway vehicle 200. The wheel tread inspection system 1 detects a flaw 210 on a tread 201a of a wheel 201 of the railway vehicle 200 running on a track 300.

[0013] Returning to FIG. 1 , the detection device 7 is capable of detecting the railway vehicle 200 traveling on the track 300. For example, the railway vehicle 200 travels at a commercial speed. That is, the detection device 7 is capable of detecting the railway vehicle 200 traveling on the track 300 at the commercial speed. The detection device 7 detects the railway vehicle 200 passing the detection device 7. The detection device 7 is, for example, an optical detection sensor capable of detecting the railway vehicle 200 traveling on the track 300 in a non-contact manner. When the detection device 7 detects the railway vehicle 200, it transmits vehicle passing information to the data processing device 10.

[0014] Each event sensor 5 is installed so as to be able to sense the tread 201a of the wheel 201. In this embodiment, the event sensor 5A can sense a part (half) of one circumference of the tread 201a of the wheel 201 of the railway vehicle 200 traveling at commercial speed, and the event sensor 5B can sense the other part (half) of one circumference of the tread 201a. Note that a part of the sensing area of ​​the event sensor 5A and a part of the sensing area of ​​the event sensor 5B may overlap with respect to the tread 201a. Note that the number, installation position, sensing range, etc. of the event sensors 5 are not limited to those described above.

[0015] The event sensor 5 is also called an event-based vision sensor. The event sensor 5 detects a change in luminance (change over time) of each pixel, and outputs only the changed luminance in combination with its coordinates and the time of occurrence. That is, each event sensor 5 detects the luminance change of each pixel as an event, and outputs luminance distribution data indicating the timing at which the event was detected, the pixel position at which the event occurred, and the like. The luminance distribution data indicates the luminance change of each pixel. The luminance distribution data is luminance information regarding the luminance according to the state of the tread surface 201a of the wheel 201 of the railway vehicle 200. The luminance distribution data is an example of event data. The luminance distribution data includes the timing (time) at which the luminance change occurred, the pixel position (coordinates) which is the position of the pixel at which the luminance change occurred, and the luminance change amount which is the amount of change in luminance. The luminance change amount has two values, an increase (+) or a decrease (-). An increase in the luminance change amount represents "+1", and a decrease in the luminance change amount represents "-1". The luminance change amount is also called a luminance change value. The luminance distribution data does not include data such as the amount of luminance change for pixels where no luminance change, which is an event, has occurred. Such luminance distribution data output by the event sensor 5 tends to have a smaller amount of data than imaging data output by a camera such as an RGB camera. This is because imaging data includes imaging data for all pixels, whereas the luminance distribution data includes only data for pixels where the luminance has changed, and does not include data for pixels where the luminance has not changed.

[0016] The lighting device 6 illuminates the sensing range of the event sensor 5. That is, the lighting device 6 illuminates the tread surface 201a of the wheel 201 that is sensed by the event sensor 5.

[0017] 3 is a block diagram showing an example of a hardware configuration of the data processing device 10 according to the embodiment. As shown in FIG. 3, the data processing device 10 includes a control device 101, a communication unit 102, and a storage device 103.

[0018] The control device 101 includes a CPU 111, a ROM 112, a RAM 113, etc. The CPU 111 executes a program stored in the ROM 112, for example, to perform various controls of the data processing device 10 and various calculations.

[0019] The communication unit 102 is an element that realizes a communication function for communicating with an external device. In the data processing device 10 of this embodiment, the communication unit 102 executes communication with the detection device 7, the event sensor 5, and the like.

[0020] The storage device 103 is configured with a HDD (Hard Disk Drive) etc. The storage device 103 stores various types of information and data.

[0021] FIG. 4 is a block diagram showing an example of a functional configuration of the control device 101 of the data processing device 10 according to the present embodiment. As shown in FIG. 4, the control device 101 has, as functional units, an acquisition unit 121, a detection unit 122, an output unit 123, and a setting unit 124. These functional units (acquisition unit 121, detection unit 122, output unit 123, setting unit 124) can be realized by cooperation of hardware elements (e.g., a CPU, a memory, etc.) and software elements (e.g., a program stored in a memory, etc.) of the data processing device 10. That is, the software elements (programs) can include, as an example, modules corresponding to each block. In addition, at least a part of these functional units (acquisition unit 121, detection unit 122, output unit 123, setting unit 124) may be configured by dedicated hardware (circuits, etc.).

[0022] The acquisition unit 121 acquires the luminance distribution data from each event sensor 5 .

[0023] The detection unit 122 detects the state of the tread 201a of the wheel 201 based on the luminance distribution data acquired by the acquisition unit 121. In other words, the detection unit 122 detects the state of the tread 201a of the wheel 201 based on a time-series change in luminance. In detail, the detection unit 122 detects the state of the tread 201a of the wheel 201 based on the amount of luminance change included in the luminance distribution data acquired by the acquisition unit 121 and a threshold value. The state detected by the detection unit 122 is a scratch as an appearance abnormality.

[0024] The output unit 123 outputs the state of the tread 201a of the wheel 201 detected by the detection unit 122. For example, the output unit 123 outputs the state of the tread 201a of the wheel 201 detected by the detection unit 122 to a management device such as a server that manages a plurality of railcars 200. The information output to the management device can be used as management information for the railcars 200, for example.

[0025] The setting unit 124 sets a threshold value used by the detection unit 122 to detect the state of the tread 201 a of the wheel 201 .

[0026] <Processing performed by the wheel tread inspection system 1> Next, a description will be given of various processes performed by the wheel tread inspection system 1. Fig. 5 is a perspective view showing a part of the wheel tread inspection system 1 according to the embodiment and a reference wheel 201A. Fig. 6 is a diagram showing an example of the detection result of the event sensor 5 of the embodiment, in which there is a scratch on the tread 201a of the wheel 201.

[0027] The data processing device 10 of the wheel tread inspection system 1 detects scratches 210 on the tread 201a of the inspection target railway vehicle 200 based on the tread 201a of the inspection target wheel 201 and a threshold value based on the tread 201aA of a reference wheel 201A (Figure 5).

[0028] When the tread surface 201a has a scratch 210, the amount of change in luminance increases compared to when the tread surface 201a does not have the scratch 210. Here, FIG. 6 is a diagram showing an example of the detection result of the event sensor 5 of the embodiment, and is a diagram showing a case where the tread surface 201a of the wheel 201 has a scratch. The horizontal axis of FIG. 6 is the position in the circumferential direction on the tread surface 201a of the wheel 201. The vertical axis of FIG. 6 is the amount of change in luminance. Line L1 in FIG. 6 is the amount of change in luminance detected by the event sensor 5. The amount of change in luminance shown by line L1 is, for example, the average value of the change in luminance in the area divided in the circumferential direction on the tread surface 201a of the wheel 201. Note that the division distance may be finely set at the mm level, or may be large at the cm or m level, as necessary. Also, line L2 in FIG. 6 indicates a threshold value based on the tread surface 201aA of the reference wheel 201A (FIG. 5). 6, when there is a scratch 210 on the tread 201a of the wheel 201, the amount of change in luminance in the area where the scratch 210 is present is greater than the amount of change in luminance in the tread 201aA of the reference wheel 201A, i.e., the threshold value. Also, when there is a scratch on the tread 201a of the wheel 201, the amount of change in luminance is greater only in a predetermined area of ​​the tread 201a that includes the scratch.

[0029] In this way, the amount of change in luminance differs between when there is a scratch on the tread 201a and when there is no scratch on the tread 201a of the wheel 201. Therefore, the data processing device 10 of the wheel tread inspection system 1 of this embodiment uses the difference in the amount of change in luminance to detect scratches on the tread 201a of the wheel 201. In other words, when a scratch occurs on the wheel 201, the change in luminance increases only in a part of the area of ​​a part of the wheel 201, so in this embodiment, the increase in the change in luminance is detected as a scratch. Also, as an example, the data processing device 10 uses the amount of change in luminance when there is no scratch on the tread 201a of the wheel 201 as a threshold value. The data processing device 10 detects scratches on the tread 201a of the wheel 201 based on the amount of change in luminance detected by the event sensor 5 and the threshold value.

[0030] Next, a specific description will be given of the processing executed by the data processing device 10. First, a threshold value setting process will be described. Fig. 7 is a flowchart showing an example of the threshold value setting process executed by the data processing device 10 of the embodiment. In this threshold value setting process, a reference wheel 201A is mounted on the railway vehicle 200.

[0031] 7, the acquisition unit 121 waits to receive vehicle passage information from the detection device 7 (S101: No). When the acquisition unit 121 receives vehicle passage information from the detection device 7 (S101: Yes), the acquisition unit 121 turns on the event sensor 5 and the lighting device 6 (S102). In other words, the acquisition unit 121 activates the event sensor 5 and the lighting device 6.

[0032] Next, the acquisition unit 121 acquires luminance distribution data for one revolution of the tread surface 201aA of the reference wheel 201A from the event sensor 5 (S103). The acquisition unit 121 causes the storage device 103 to store the acquired luminance distribution data.

[0033] Next, the setting unit 124 sets a threshold value based on the luminance distribution data acquired by the acquisition unit 121 (S104). The threshold value is an average value of the luminance change amount included in the luminance distribution data. Specifically, the threshold value is set for each of a plurality of regions divided in the circumferential direction of the tread surface 201a. The threshold value is an average value of the luminance change amount in each region. The section distance may be set finely at the mm level or large at the cm or m level as necessary. The setting unit 124 stores the set threshold value for each region in the storage device 103. The luminance distribution data for the threshold value is also referred to as standard luminance distribution data. The threshold value may be set individually for each of the plurality of wheels 201 using a plurality of reference wheels 201A corresponding to each of the plurality of wheels 201 of the railcar 200, or one common threshold value may be set for the plurality of wheels 201 using one reference wheel 201A.

[0034] Next, a description will be given of a flaw detection process executed by the data processing device 10. Fig. 8 is a flowchart showing an example of the flaw detection process executed by the data processing device 10 according to the embodiment.

[0035] 8, the acquisition unit 121 waits to receive vehicle passage information from the detection device 7 (S201: No). When the acquisition unit 121 receives vehicle passage information from the detection device 7 (S201: Yes), the acquisition unit 121 turns on the event sensor 5 and the lighting device 6 (S202). In other words, the acquisition unit 121 activates the event sensor 5 and the lighting device 6.

[0036] Next, the acquisition unit 121 acquires luminance distribution data for one revolution of the tread surface 201a of the wheel 201 from the event sensor 5 (S203). The acquisition unit 121 stores the acquired luminance distribution data in the storage device 103.

[0037] Next, the detection unit 122 determines whether there is a region where the amount of change in luminance is greater than the threshold (increased region) (S204). At this time, the detection unit 122 determines for each region whether the average value of the amount of change in luminance in the region is greater than the threshold for the region.

[0038] When the detection unit 122 determines that there is an area where the amount of change in luminance is greater than the threshold (S204: Yes), it determines that there is a scratch on the tread 201a of the wheel 201 (S205). That is, the detection unit 122 detects a scratch on the tread 201a of the wheel 201. Note that the detection unit 122 may determine that there is a scratch in a part of the area when the part of the area is larger than the other areas by a predetermined amount or more.

[0039] On the other hand, if it is determined in S204 that there is no area in which the amount of change in luminance is greater than the threshold (S204: No), the detection unit 122 determines that there is no scratch on the tread surface 201a of the wheel 201 (S206). After S206, the processing from S203 onwards is repeatedly executed.

[0040] After the above processing, the output unit 123 outputs the detection result of the detection unit 122. At this time, the output unit 123 can reconstruct the portion of the brightness distribution data acquired by the acquisition unit 121 that has the scratch 210 as image data and output it.

[0041] <Summary> As described above, the data processing device 10 (wheel tread inspection device) of this embodiment includes the acquisition unit 121 and the detection unit 122. The acquisition unit 121 senses the tread 201a of the wheel 201 of the railway vehicle 200 and outputs event data indicating a luminance change for each pixel. The detection unit 122 detects scratches 210 on the tread 201a based on the event data acquired by the acquisition unit 121.

[0042] According to this configuration, the flaw 210 on the tread 201a of the wheel 201 is detected based on the event data from the event sensor 5, so the amount of data processing by the acquisition unit 121 and the detection unit 122 can be reduced compared to the case where the flaw 210 on the tread 201a of the wheel 201 is detected based on image data output from a camera that captures the tread 201a of the wheel 201. Therefore, according to the above configuration, the load of data processing for the flaw 210 detection can be reduced. Furthermore, according to the above configuration, it is possible to obtain a data processing device 10 that can detect the flaw 210 on the tread 201a of the wheel 201 without performing image processing. Therefore, it is possible to reduce running costs such as power consumption in the flaw detection process by the data processing device 10. Furthermore, since the flaw can be detected continuously in time using the event sensor 5, it is possible to suppress overlooking the flaw even when the railcar 200 is traveling at high speed.

[0043] Furthermore, when the luminance change in the event data is greater than a threshold value that is set in advance using the tread surface 201aA of a reference wheel 201A that does not have a scratch 210, the detection unit 122 determines that a scratch 210 exists.

[0044] According to this configuration, the damage 210 is detected based on a preset threshold value using the tread surface 201aA of the reference wheel 201A, so that the accuracy of the detection of the damage 210 can be improved.

[0045] Furthermore, the acquisition unit 121 acquires the event data when the detection device 7, which detects the railway vehicle 200 traveling on the track 300 at a commercial speed, detects the railway vehicle 200. The commercial speed is also referred to as the normal speed.

[0046] According to this configuration, scratches 210 on the tread 201a of the wheels 201 of the railway vehicle 200 traveling on the track 300 at commercial speed can be detected, so there is no need for the railway vehicle 200 to travel at an inspection speed that is different from the commercial speed, for example, in order to inspect the tread 201a of the wheels 201.

[0047] Further, a plurality of event sensors 5 are arranged along the track 300. When a detection device 7, which detects a railway vehicle 200 traveling on the track 300 at commercial speed, detects the railway vehicle 200, the acquisition unit 121 acquires event data for at least one revolution of the tread surface 201a from the plurality of event sensors 5 arranged at intervals along the track 300.

[0048] According to this configuration, the presence or absence of scratches 210 can be inspected for the entire circumference of the tread surface 201a of the wheel 201.

[0049] In the above embodiment, the luminance distribution data (event data) of the tread 201aA of the reference wheel 201A is acquired using the railway vehicle 200 traveling on the track 300, but the present invention is not limited to this. For example, the luminance distribution data (event data) of the tread 201aA of the reference wheel 201A may be acquired in a factory. In addition, in the above embodiment, the threshold value is calculated from the luminance distribution data (event data) of the tread 201aA of the reference wheel 201A, but the present invention is not limited to this. For example, it is assumed that the treads 201a of the wheels 201 of the majority of railway vehicles 200 in commercial operation are free of scratches 210, and the average of the luminance distribution data of the treads 201a of the wheels 201 of a predetermined number of railway vehicles 200 or more may be set as the threshold value.

[0050] In the above embodiment, an example in which a plurality of event sensors 5 and a plurality of lighting devices 6 are installed is shown, but the present invention is not limited to this. For example, there may be only one event sensor 5 and one lighting device 6. In this case, it is sufficient that one event sensor 5 acquires luminance distribution data for one revolution of the wheel 201.

[0051] Although some embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and spirit of the invention, and are included in the scope of the invention and its equivalents described in the claims. [Explanation of symbols]

[0052] 7...detection device, 10...data processing device (wheel tread inspection device), 121...acquisition unit, 122...detection unit, 200...railroad vehicle, 201...wheel, 201a...tread, 201A...reference wheel, 201aA...tread, 210...defect.

Claims

1. an acquisition unit that acquires event data from an event sensor that senses a wheel tread of a railway vehicle and outputs event data indicating a luminance change for each pixel; A detection unit that detects damage to the tread surface based on the event data acquired by the acquisition unit; A wheel tread inspection device equipped with

2. 2. The wheel tread inspection device according to claim 1, wherein the detection unit determines that the scratch is present when the luminance change in the event data is greater than a threshold value that is preset using the tread of a reference wheel without the scratch.

3. the acquisition unit acquires the event data when a detection device that detects the railway vehicle traveling on a track at a commercial speed detects the railway vehicle.

2. A wheel tread inspection device according to claim 1.

4. The event sensor is arranged along a railway line; the acquisition unit acquires the event data for at least one revolution of the tread from a plurality of the event sensors arranged at intervals along the track when a detection device that detects the railcar traveling on the track at a commercial speed detects the railcar.

2. A wheel tread inspection device according to claim 1.

Images