Non-contact automatic measurement method for wear of subway contact rail

Abstract

The invention belongs to the technical field of railway detection, and particularly relates to a non-contact automatic measurement method for metro contact rail wear, which adopts a non-contact detection method to continuously scan and measure a contact rail surface through a high-precision 3D camera, quickly and accurately calculates a contact rail wear value, a current receiving surface included angle and an eccentric wear value, and is accurate in measurement, high in efficiency and high in accuracy. The whole-line uninterrupted non-power-failure measurement can be realized; according to the method, a method of combining 3D line scanning of the contour of the contact rail and image processing after photographing of an industrial camera is adopted, and the contour line processing of the 3D camera has certain inaccuracy when the contact rail is only loaded with a place with eccentric wear on one side of the contact rail, the section integrity is calculated by adopting the contour line of the 3D camera, and the position where the eccentric wear possibly exists is quickly marked; and then analyzing and calculating by an image method, more accurately analyzing the eccentric wear caused by the fact that the collector shoe only carries one side of the contact rail, greatly reducing the picture processing amount, improving the real-time performance of measurement, and providing a rapid and effective basis for the adjustment of the contact rail.

Description

Technical Field

[0001] The invention belongs to the technical field of railway detection, and in particular relates to a non-contact automatic measurement method for subway contact rail wear. Background Art

[0002] In recent years, my country's subway rail transit has achieved rapid development. The subway contact rail is an important way to power supply for subway rail transit, and the contact rail surface condition is an important guarantee for the safe operation of trains. During the long-term operation of subway lines, high temperature, corrosion, poor overlap of power supply shoes and other reasons will cause eccentric wear of the contact rail surface, thereby reducing the diversion effect. At present, the subway operation and maintenance department mostly uses manual observation to detect contact rail wear, but there is no automated detection method, which is time-consuming and labor-intensive, and there are too many human factors in observation, and the accuracy is low. Therefore, a new automated contact rail wear detection method is needed to improve the efficiency and accuracy of contact rail wear detection. Summary of the invention

[0003] The purpose of the present invention is to provide a non-contact automatic measurement method for subway contact rail wear, so as to solve the technical problems mentioned in the background technology.

[0004] In order to achieve the above technical objectives, the technical solution adopted by the present invention is as follows:

[0005] A non-contact automatic measurement method for subway contact rail wear, characterized in that it specifically comprises the following steps:

[0006] Step 1: Use a high-precision 3D camera to continuously scan the subway contact rail and extract each frame of contact rail profile scanning data; use a high-definition industrial camera to continuously take pictures of the contact rail surface and save each contact rail surface picture; synchronously mark the mileage for each frame of contact rail profile scanning data and each contact rail surface picture;

[0007] Step 2: Filter and segment each frame of contact rail profile scanning data, extract the contact rail surface contour line, and calculate the wear value, current receiving surface angle and eccentric wear value;

[0008] Step 3: Perform signal strength filtering on the 3D camera data of the contact rail surface contour collected by the camera, perform edge detection on the contact rail image at the corresponding position where the contour data after filtering is severely missing, calculate the friction surface width of the power supply shoe on the contact rail, and infer the carrying capacity of the power supply shoe on the contact rail surface based on the friction surface width.

[0009] The specific steps of step 1 are:

[0010] S1: The contact rail automatic detection device moves along the subway track, and the high-precision 3D camera mounted on the contact rail measurement robot continuously scans the contact rail surface including the third rail and return rail;

[0011] S2: The high-definition industrial camera mounted on the contact rail measuring robot automatically takes pictures of the contact rail surface at regular intervals according to the mileage changes of the contact rail automatic detection device;

[0012] S3: Each frame of contact rail surface data scanned by the 3D camera and each rail surface picture taken by the high-definition industrial camera are associated with the mileage sensor data obtained in real time when the contact rail automatic detection device moves.

[0013] The specific steps of step 2 are:

[0014] S1: Filter the contact rail surface data, remove invalid data, perform straight line segmentation, and separate all points of the contact rail surface contour line;

[0015] S2: Perform straight line fitting on the rail surface contour points, and make the points basically on one side of the fitted virtual straight line OA;

[0016] S3: Taking the virtual straight line OA as the baseline, calculate the deviation value and the number of deviation points from the OA point to determine whether there is wear;

[0017] S4: Find the point D with the farthest deviation along the straight line OA along the worn contour line, make DC perpendicular to OA, calculate CD value as the maximum wear value and OD value as the most eccentric wear position value;

[0018] S5: Draw straight lines OB and OA parallel to the rail surface and intersect at O, and calculate the angle AOB as the angle of the current receiving surface;

[0019] S6: Repeat S2 to S5 to continuously calculate multiple contact rail surface contour lines, confirm the repeatability of the wear value, the current receiving surface angle and the eccentric wear value, and output the measurement results.

[0020] The specific steps of step 3 are:

[0021] S1: Using the 3D camera data intensity of the normal unworn contact rail surface as a reference, perform intensity filtering on the contact rail surface contour points, calculate the effective point distance of the track surface contour line for the filtered points, and mark the position where the effective point distance is insufficient as a reference for possible eccentric wear caused by the power supply shoe overlapping the contact rail surface;

[0022] S2: Find the contact rail surface image at the corresponding position, and perform filtering and edge detection;

[0023] S3: Follow the 3D camera laser red line upward to find a point on the left edge of the contact rail surface as the starting point, horizontally to the right to find a point on the friction edge of the contact rail as the end point, and calculate the pixel values ​​from the starting point to the end point;

[0024] S4: repeating step S3 along the direction in which the contact rail measuring device is moving, calculating the pixel values ​​of multiple positions from the left edge of the rail surface to the friction edge of the contact rail, and finding the average value of the multiple pixel values;

[0025] S5: Comparing the pixels occupied by the complete rail surface of the contact rail with the actual rail surface width, calculating the rail surface friction surface width value corresponding to the pixel average value;

[0026] S6: Output the ratio of the rail friction surface width value calculated by image analysis to the actual rail surface width, and warn if the ratio is insufficient that there is eccentric wear on the contact rail surface of the output power shoe.

[0027] This method adopts a non-contact detection method to continuously scan and measure the contact rail surface through a high-precision 3D camera, and quickly and accurately calculates the contact rail wear value, current receiving surface angle, and eccentric wear value. The measurement is accurate and efficient, and the whole line can be measured uninterruptedly without power outages. This method adopts a method that combines 3D line scanning of the contact rail contour and image processing after industrial camera photography. There is a certain inaccuracy in the 3D camera contour line processing for the place where the power supply shoe is only carried on one side of the contact rail and eccentric wear occurs. The 3D camera contour line is used to calculate the cross-sectional integrity, and the position where eccentric wear may exist is quickly marked. Then, the image method is used for analysis and calculation, and the eccentric wear caused by the power supply shoe being only carried on one side of the contact rail is more accurately analyzed, and the amount of image processing is greatly reduced, thereby improving the real-time measurement and providing a fast and effective basis for contact rail adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The present invention is further illustrated by means of the non-limiting examples given in the accompanying drawings.

[0029] Figure 1 It is a schematic diagram of the method steps of the present invention.

[0030] Figure 2 It is a schematic diagram of a contact rail cross section and a 3D point cloud scan of an unworn contact rail of the present invention.

[0031] Figure 3 It is a schematic diagram of the contact rail that produces wear and eccentric wear, the scanned 3D point cloud and the calculation model diagram of the present invention.

[0032] Figure 4 It is a real-life picture of the contact rail surface with insufficient carrying capacity, a point cloud analysis diagram and a schematic diagram after edge conversion of the present invention. DETAILED DESCRIPTION

[0033] In order to enable those skilled in the art to better understand the present invention, the technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0034] like Figure 1-4 As shown, a non-contact automatic measurement method for subway contact rail wear specifically includes the following steps:

[0035] Step 1: Use a high-precision 3D camera to continuously scan the subway contact rail and extract each frame of contact rail profile scanning data; use a high-definition industrial camera to continuously take pictures of the contact rail surface and save each contact rail surface picture; synchronously mark the mileage for each frame of contact rail profile scanning data and each contact rail surface picture;

[0036] Step 2: Filter and segment each frame of contact rail profile scanning data, extract the contact rail surface contour line, and calculate the wear value, current receiving surface angle and eccentric wear value;

[0037] Step 3: Perform signal strength filtering on the 3D camera data of the contact rail surface contour collected by the camera, perform edge detection on the contact rail image at the corresponding position where the contour data after filtering is severely missing, calculate the friction surface width of the power supply shoe on the contact rail, and infer the carrying capacity of the power supply shoe on the contact rail surface based on the friction surface width.

[0038] The specific steps of step 1 are:

[0039] S1: The contact rail automatic detection device moves along the subway track, and the high-precision 3D camera mounted on the contact rail measurement robot continuously scans the contact rail surface including the third rail and return rail;

[0040] S2: The high-definition industrial camera mounted on the contact rail measuring robot automatically takes pictures of the contact rail surface at regular intervals according to the mileage changes of the contact rail automatic detection device;

[0041] S3: Each frame of contact rail surface data scanned by the 3D camera and each rail surface picture taken by the high-definition industrial camera are associated with the mileage sensor data obtained in real time when the contact rail automatic detection device moves.

[0042] The specific steps of step 2 are:

[0043] S1: Filter the contact rail surface data, remove invalid data, perform straight line segmentation, and use the RANSAC straight line segmentation method to separate all 3D camera scanning point data on the contact rail surface contour line;

[0044] S2: Perform straight line fitting on the rail surface contour points, and make the points basically on one side of the fitted virtual straight line OA;

[0045] S3: Taking the virtual straight line OA as the baseline, calculate the maximum deviation value and the number of deviation points from the straight line OA. When the deviation value is not less than 2 mm and the number of points is greater than 3, it is determined that there is rail surface wear and continue to the next step of calculation;

[0046] S4: Find the point D with the farthest deviation along the straight line OA along the worn contour line, and multiple points with similar deviation values ​​near point D, make DC perpendicular to OA, calculate CD value as the maximum wear value and OD value as the most eccentric wear position value;

[0047] S5: Draw straight lines OB and OA parallel to the rail surface and intersect at O, and calculate the angle AOB as the angle of the current receiving surface;

[0048] S6: Repeat S2 to S5 to continuously calculate multiple contact rail surface contour lines, confirm the repeatability of the wear value, the current receiving surface angle and the eccentric wear value, and output the measurement results.

[0049] The specific steps of step 3 are:

[0050] S1: Using the 3D camera data intensity of the normal unworn contact rail surface as a reference, perform intensity filtering on the contact rail surface contour points, calculate the effective point distance x1x3 of the track surface contour line for the filtered points, and mark the position where the effective point distance x1x3 is insufficient as a reference for possible eccentric wear caused by the power supply shoe overlapping the contact rail surface;

[0051] S2: Find the contact rail surface image at the corresponding position, perform filtering and edge detection, first use multiple images with different wear ratios and insufficient loading of the power supply shoe on the contact rail surface offline, use OpenCV to perform edge calculation, and iterate the Canny operator;

[0052] Take out multiple point cloud-annotated contact rail surface images of the corresponding positions where the power supply shoe is not loaded on the contact rail surface in real time, convert them into grayscale images using OpenCV, blur them with the blur method to reduce noise, and use the Canny operator calculated offline for edge detection;

[0053] S3: Follow the 3D camera laser red line upward to find a point on the left edge of the contact rail surface as the starting point, horizontally to the right to find a point on the friction edge of the contact rail as the end point, and calculate the pixel values ​​from the starting point to the end point;

[0054] S4: along the walking direction of the contact rail measuring device, repeat more than 3 consecutive contact rail surface pictures, calculate the pixel values ​​of multiple positions from the left edge of the rail surface to the friction edge of the contact rail, and calculate the average value of the multiple pixel values;

[0055] S5: Comparing the pixels occupied by the complete rail surface of the contact rail with the actual rail surface width, calculating the rail surface friction surface width value corresponding to the pixel average value;

[0056] S6: Output the ratio of the rail friction surface width value calculated by image analysis to the actual rail surface width, and warn if the ratio is insufficient that there is eccentric wear on the contact rail surface of the output power shoe.

[0057] This method adopts a non-contact detection method to continuously scan and measure the contact rail surface through a high-precision 3D camera, and quickly and accurately calculates the contact rail wear value, current receiving surface angle, and eccentric wear value. The measurement is accurate and efficient, and the whole line can be measured uninterruptedly without power outages. This method adopts a method that combines 3D line scanning of the contact rail contour and image processing after industrial camera photography. There is a certain inaccuracy in the 3D camera contour line processing for the place where the power supply shoe is only carried on one side of the contact rail and eccentric wear occurs. The 3D camera contour line is used to calculate the cross-sectional integrity, and the position where eccentric wear may exist is quickly marked. Then, the image method is used for analysis and calculation, and the eccentric wear caused by the power supply shoe being only carried on one side of the contact rail is more accurately analyzed, and the amount of image processing is greatly reduced, thereby improving the real-time measurement and providing a fast and effective basis for contact rail adjustment.

[0058] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present invention. Therefore, all equivalent modifications or changes made by those with ordinary knowledge in the technical field without departing from the spirit and technical ideas disclosed by the present invention should still be covered by the claims of the present invention.

Claims

1. A non-contact automatic measurement method for subway contact rail wear, characterized in that: The specific steps include: Step 1: Use a high-precision 3D camera to continuously scan the subway contact rail and extract each frame of contact rail profile scanning data; use a high-definition industrial camera to continuously take pictures of the contact rail surface and save each contact rail surface picture; synchronously mark the mileage for each frame of contact rail profile scanning data and each contact rail surface picture; Step 2: Filter and segment each frame of contact rail profile scanning data, extract the contact rail surface contour line, and calculate the wear value, current receiving surface angle and eccentric wear value; Step 3: Perform signal strength filtering on the 3D camera data of the contact rail surface contour collected by the camera, perform edge detection on the contact rail image at the corresponding position where the contour data after filtering is severely missing, calculate the friction surface width of the power supply shoe on the contact rail, and infer the carrying capacity of the power supply shoe on the contact rail surface based on the friction surface width.

2. The non-contact automatic measurement method for subway contact rail wear according to claim 1 is characterized in that: The specific steps of step 2 are: S1: Filter the contact rail surface data, remove invalid data, perform straight line segmentation, and separate all points of the contact rail surface contour line; S2: Perform straight line fitting on the rail surface contour points, and make the points basically on one side of the fitted virtual straight line OA; S3: Taking the virtual straight line OA as the baseline, calculate the deviation value and the number of deviation points from the OA point to determine whether there is wear; S4: Find the point D with the farthest deviation along the straight line OA along the worn contour line, make DC perpendicular to OA, calculate CD value as the maximum wear value and OD value as the most eccentric wear position value; S5: Draw straight lines OB and OA parallel to the rail surface and intersect at O, and calculate the angle AOB as the angle of the current receiving surface; S6: Repeat S2 to S5 to continuously calculate multiple contact rail surface contour lines, confirm the repeatability of the wear value, the current receiving surface angle and the eccentric wear value, and output the measurement results.

3. The non-contact automatic measurement method for subway contact rail wear according to claim 2 is characterized in that: The specific steps of step 3 are: S1: Using the 3D camera data intensity of the normal unworn contact rail surface as a reference, perform intensity filtering on the contact rail surface contour points, calculate the effective point distance x1x3 of the track surface contour line for the filtered points, and mark the position where the effective point distance x1x3 is insufficient as a reference for possible eccentric wear caused by the power supply shoe overlapping the contact rail surface; S2: Find the contact rail surface image at the corresponding position, and perform filtering and edge detection; S3: Follow the 3D camera laser red line upward to find a point on the left edge of the contact rail surface as the starting point, horizontally to the right to find a point on the friction edge of the contact rail as the end point, and calculate the pixel values ​​from the starting point to the end point; S4: repeating step S3 along the direction in which the contact rail measuring device is moving, calculating the pixel values ​​of multiple positions from the left edge of the rail surface to the friction edge of the contact rail, and finding the average value of the multiple pixel values; S5: Comparing the pixels occupied by the complete rail surface of the contact rail with the actual rail surface width, calculating the rail surface friction surface width value corresponding to the pixel average value; S6: Output the ratio of the rail friction surface width value calculated by image analysis to the actual rail surface width, and warn if the ratio is insufficient that there is eccentric wear on the contact rail surface of the output power shoe.

Images