Method and system for evaluating risk of continuously welded rail based on vehicle-mounted inspection

Abstract

The invention provides a seamless track steel rail risk assessment method and system based on vehicle-mounted inspection. The method comprises the following steps: dynamically acquiring image data including a steel rail displacement observation pile and a scale by adopting a camera device arranged at the bottom of a vehicle carrier; after the image is divided to determine foreground and background areas, contour features of a displacement observation pile and a scale are determined based on an edge detection algorithm; normalizing the displacement observation pile, the scale and the template object, and determining the central positions of the displacement observation pile and the scale and the template object; after determining the detection linear distance between the centers of the two, calculating the longitudinal displacement of the corresponding seamless steel rail by combining the installation distance between the centers of the two, and further analyzing the temperature stress index of the line steel rail based on the longitudinal displacement to realize rail breakage risk evaluation. According to the scheme, the problems of long detection time consumption, high error probability and high cost in the prior art can be solved, accurate and efficient detection of the longitudinal displacement of the steel rail in the running state is realized, and accurate temperature stress data is quickly obtained to reliably evaluate the risk of the steel rail.

Description

Technical Field

[0001] The present invention relates to the technical field of seamless line rail inspection and maintenance, and in particular to a seamless line rail risk assessment method and system based on vehicle-mounted inspection. Background Art

[0002] Since the seamless line design eliminates rail joints, it helps to improve the track structure and reduce vibration, and is being used more and more widely in newly built lines. But on the other hand, due to the disappearance of the rail joints, the rails cannot expand and contract freely when the temperature changes, which will cause great temperature stress inside the rails. When the fastener pressure is not enough to offset the temperature stress, the rails will undergo longitudinal displacement, causing rail expansion and track runaway when the temperature rises, and rail breakage is likely to occur when the temperature is very low, seriously affecting driving safety.

[0003] In order to ensure operational safety and understand rail stress, it is necessary to conduct periodic inspections or real-time monitoring of seamless line rail displacement observation piles. However, the main monitoring technologies currently used include two categories. One is to use manual visual observation one by one, which is a contact measurement method. For example, through manual on-site inspection, a square ruler is used to compare and measure the relative position changes between the displacement observation pile and the rail bottom scale. However, this method requires a huge amount of human resources and time resources when used. In addition to consuming a large amount of manpower, manual inspection is difficult to accurately control the selection of the geometric center of the observation pile, resulting in large measurement errors and low detection efficiency. It also requires high professionalism of the inspection personnel, has great limitations, has many influencing factors, and has a high error rate.

[0004] The other type is to install monitoring devices at the rail joint position, which is a non-contact measurement method. For example, Liu Liping et al. proposed a railway rail creep displacement detection method based on monocular vision measurement technology. The image containing circular markers is collected by an image acquisition device installed on the inherent facilities next to the rail. The position of the marker in the reconstructed image is preliminarily located by the template matching method. The center of mass coordinates and radius of the marker are accurately calculated by combining edge detection and circle fitting methods. The physical displacement of the rail creep is detected based on the image imaging principle. In addition, rail temperature monitoring sensors can be set at the joint position to realize real-time monitoring of rail temperature changes and calculate the longitudinal displacement change of the rail. When this method is used, a large number of monitoring device components or sensor equipment need to be set up along the line. There are shortcomings such as large equipment investment and high installation cost. The later maintenance workload is complicated. It is also difficult to meet the objective needs of high-speed dynamic measurement and cannot achieve timely and accurate rail break risk assessment.

[0005] The information disclosed in the background technology section of the present invention is only intended to deepen the understanding of the general background technology of the present invention, and should not be regarded as acknowledging or suggesting in any form that the information constitutes the prior art known to those skilled in the art. Summary of the invention

[0006] To solve the above problems, the present invention provides a seamless line rail risk assessment method and system based on vehicle-mounted inspection, which obtains high-resolution image information of seamless line observation piles through a vehicle-mounted camera, applies the Otsu algorithm to divide the foreground and background, applies edge detection to realize shape feature extraction of observation piles and rulers, applies the template matching algorithm to realize the matching of target rectangles, target circles and shapes in the picture, calculates the displacement of the rails through center extraction, and calculates the temperature stress, temperature change, temperature force and other indicators of the rails of the entire line in real time, thereby realizing the risk assessment of broken rails; preferably, in one embodiment, the method includes:

[0007] Dynamic collection step: when the vehicle carrier moves relative to the measured rail, a camera device arranged at the bottom of the vehicle carrier is used to collect image data including the rail displacement observation pile and the scale;

[0008] Image segmentation step: Use the Otsu algorithm to perform binary processing on the collected image data to determine the foreground area and background area;

[0009] Feature extraction step: Determine the position and contour features of the displacement observation pile and ruler in the image based on the Canny edge detection algorithm;

[0010] The calculation center determination step: normalize the contours of the displacement observation pile and the ruler in the image and the template object respectively, and then match and determine the center positions of the displacement observation pile and the ruler;

[0011] Longitudinal displacement calculation steps: determine the detection straight line distance between the center of the displacement observation pile and the center of the scale, and calculate the longitudinal displacement of the corresponding seamless rail in combination with the installation distance between the center of the displacement observation pile and the center of the scale;

[0012] Real-time rail temperature analysis step: based on the longitudinal displacement combined with the expansion and contraction properties of the rail, calculating the change of the actual rail temperature at the current position of the seamless rail relative to the construction locked rail temperature, and then determining the temperature stress index at different positions of the rail based on the change, the temperature stress index includes temperature stress and full-section temperature force;

[0013] Risk assessment and processing steps: determine the matching rail risk assessment results based on the distribution of the temperature stress index, provide corresponding processing solutions in combination with the set decision-making principles, and output them visually to the user.

[0014] In an optional embodiment, the camera device uses a high-definition linear array laser camera that is installed on the vehicle relative to the rails to dynamically collect a grayscale image including the rail displacement observation piles and scales.

[0015] Preferably, in one embodiment, in the image division step, according to the inter-class variance calculation principle of image pixel classification, a target grayscale level that maximizes the inter-class variance is obtained, and all pixels of the image are divided into two categories according to the target grayscale level to achieve division.

[0016] Furthermore, in one embodiment, the inter-class variance of image pixel classification is calculated according to the following formula:

[0017] σ 2 =p1(l1-l G ) 2 +p2(l2-l G ) 2

[0018] Among them, p1l1+p2l2=l G ,but:

[0019] σ 2 =p1p2(l1-l2) 2

[0020] In the formula, σ 2 represents the inter-class variance of image pixel classification, l1 and l2 represent the means of the two types of pixels respectively, and l G represents the global pixel mean of the image, p1 and p2 represent the probability of pixels being classified into the first and second categories respectively.

[0021] Preferably, in one embodiment, in the feature extraction step, the process of determining the position and contour features of the displacement observation pile and the scale in the image includes:

[0022] Smooth the image using a Gaussian filter;

[0023] Use finite differences of first-order partial derivatives to compute the magnitude and direction of the gradient;

[0024] Perform non-maximum suppression on the gradient amplitude;

[0025] Then, a dual threshold algorithm is used to detect and connect the edges to obtain the contour features and positions of the rectangular scale and circular observation pile.

[0026] Optionally, in one embodiment, in the longitudinal displacement calculation step, the longitudinal displacement a of the corresponding seamless rail is calculated according to the following formula:

[0027]

[0028] Where b is the installation distance between the geometric center of the displacement observation pile and the geometric center of the scale, and c is the detection straight-line distance between the center of the displacement observation pile and the center of the scale.

[0029] Furthermore, in one embodiment, in the real-time rail temperature analysis step, the change in the actual rail temperature at the current position of the seamless rail relative to the construction locked rail temperature is calculated based on the longitudinal displacement according to the following formula:

[0030] Δt=ΔL / (a·L)

[0031] Where: Δt is the temperature change between the actual rail temperature and the construction locked rail temperature measurement result at the end of construction (°C); ΔL is the expansion of the rail (mm); a is the expansion coefficient of the rail, a = 0.0118mm / (m·°C), that is, when the rail temperature changes by 1°C per meter of rail, the rail will expand and contract by 0.0118mm; L is the length of the rail (m).

[0032] In a preferred embodiment, the full-section temperature force of the rail is calculated according to the following formula:

[0033] P t =σ t ·S 钢轨 =247.8·Δt·S 钢轨

[0034] Among them, σ t =247.8·Δt

[0035] Where: Pt is the temperature force (N); S 钢轨 is the cross-sectional area of ​​the rail (cm2); σt is the temperature stress (N); Δt is the rail temperature variation between the actual rail temperature and the construction locked rail temperature measurement result at the end of construction (℃).

[0036] Based on other aspects of the method described in one or more embodiments above, the present invention further provides a storage medium storing program codes that can implement the method described in one or more embodiments above.

[0037] Based on the application of the method described in one or more embodiments above, the present invention further provides a seamless line rail risk assessment system based on vehicle-mounted inspection, which executes the method described in one or more embodiments above. Preferably, in one embodiment, the system includes:

[0038] A dynamic acquisition module, comprising a camera device and a communication unit disposed at the bottom of the vehicle carrier, which is configured to dynamically acquire image data including rail displacement observation piles and scales and transmit the image data to the data processing module during the movement of the vehicle carrier relative to the measured rail;

[0039] The data processing module is fixedly arranged at the bottom or inside of the vehicle carrier, and is connected to the dynamic acquisition module and the detection control center for communication. It is used to process the acquired image data and determine the centroid of the displacement observation pile and the scale in the image, and then calculate the longitudinal displacement of the corresponding seamless rail according to the distribution of the centroid combined with the installation distance between the displacement observation pile and the scale, and determine the rail temperature stress index based on the calculation, so as to realize the risk assessment of broken rails and perform visual output of relevant data;

[0040] The data processing module comprises:

[0041] An image division unit is configured to perform binarization processing on the collected image data using the Otsu algorithm to divide and determine the foreground area and the background area;

[0042] A feature extraction unit configured to determine the position and contour features of the displacement observation pile and the scale in the image based on a Canny edge detection algorithm;

[0043] A calculation center determination unit is configured to normalize the contours of the displacement observation pile and the scale in the image and the template object, and then match and determine the center positions of the displacement observation pile and the scale;

[0044] A longitudinal displacement calculation unit, configured to determine a detection straight-line distance between the center of the displacement observation pile and the center of the scale, and calculate the longitudinal displacement of the corresponding seamless rail in combination with the installation distance between the center of the displacement observation pile and the center of the scale;

[0045] A real-time rail temperature analysis unit is configured to calculate the change of the actual rail temperature at the current position of the seamless rail relative to the construction locked rail temperature based on the longitudinal displacement combined with the expansion and contraction properties of the rail, and then determine the temperature stress index at different positions of the rail based on the change, wherein the temperature stress index includes temperature stress and full-section temperature force;

[0046] The risk assessment and processing unit is configured to determine the matching rail risk assessment result based on the distribution of the temperature stress index, provide a corresponding processing solution in combination with the set decision principle, and output it visually to the user.

[0047] Compared with the closest prior art, the present invention also has the following beneficial effects:

[0048] The present invention provides a seamless line rail risk assessment method and system based on vehicle-mounted inspection. The method adopts a camera device arranged at the bottom of a vehicle carrier to dynamically collect image data including rail displacement observation piles and rulers. The method can effectively obtain images of the rail observation piles and rulers based on the camera device in motion, which are used to analyze and calculate the longitudinal creeping displacement of the rails. There is no need to set up a large number of laser cameras or temperature sensors along the line, nor is there any need to calculate and determine the displacement based on temperature data, which saves equipment installation costs and reduces the probability of errors caused by intermediate calculations.

[0049] Furthermore, after the Otsu algorithm is used to divide the collected image data to determine the foreground area and the background area, the contour features of the displacement observation piles and rulers in the image are determined based on the edge detection algorithm; then the displacement observation piles and rulers in the image are normalized with the template object, and the center positions of the two are determined by matching; based on rigorous image processing methods, the precise displacement observation points and the center positions of the rulers are identified, which facilitates the acquisition of the distance between the center positions as the detection straight-line distance involved in the displacement calculation, thereby ensuring the reliability and authenticity of the calculation results from the aspect of calculation data.

[0050] After determining the detection straight-line distance between the center of the displacement observation pile and the center of the scale, the longitudinal displacement of the corresponding seamless rail is calculated in combination with the installation distance between the center of the displacement observation pile and the center of the scale; through computer vision technology, combined with image recognition algorithm, the rapid detection of the longitudinal displacement of the rail in the running state is realized, the efficiency of the longitudinal displacement detection of the seamless line rail is improved, the labor cost is reduced, and the line operation and maintenance efficiency is improved. The professional requirements for operators are not high, and it is suitable for lines of different mileages and a variety of transport vehicles, and it is very practical;

[0051] After calculating the longitudinal displacement of the corresponding seamless rail, the temperature stress, temperature change, temperature force and other data of the line rail are analyzed based on the longitudinal displacement to realize the risk assessment of rail breakage. This solution is based on computer vision technology and image recognition algorithm to achieve accurate and efficient detection of the longitudinal creep displacement of the rail under operation, quickly obtain accurate temperature stress data to realize reliable assessment of rail risks, and provide support for rail operation and maintenance.

[0052] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0054] Figure 1 It is a flow chart of a seamless line rail risk assessment method based on vehicle-mounted inspection provided by one embodiment of the present invention;

[0055] Figure 2 It is a detailed schematic diagram of the operating principle of the longitudinal displacement detection process in the seamless line rail risk assessment method based on vehicle-mounted inspection provided by an embodiment of the present invention;

[0056] Figure 3 It is a schematic diagram of the dynamic collection steps of a seamless line rail risk assessment method based on vehicle-mounted inspection provided by an embodiment of the present invention;

[0057] Figure 4 It is a schematic diagram of the camera device setting method of the seamless railway rail risk assessment method based on vehicle-mounted inspection provided by an embodiment of the present invention;

[0058] Figure 5 (a) and Figure 5 (b) are respectively example images captured by the left and right camera devices of the seamless railway rail risk assessment method based on vehicle-mounted inspection provided by an embodiment of the present invention;

[0059] Figure 6 It is a schematic diagram of the binarization result of the image collected by the right camera setting of the seamless railway rail risk assessment method based on vehicle-mounted inspection provided by one embodiment of the present invention;

[0060] Figure 7 It is a schematic diagram of contour feature recognition of an image collected by a right-side camera setting in a seamless line rail risk assessment method based on vehicle-mounted inspection provided in an embodiment of the present invention;

[0061] Figure 8 It is a schematic diagram of the principle of calculating the longitudinal displacement based on the displacement observation pile and the scale position of the seamless line rail risk assessment method based on vehicle-mounted inspection provided by an embodiment of the present invention;

[0062] Fig. 9 This is an example diagram of the distribution of the temperature, longitudinal displacement, and displacement difference between the left and right rails and the previous observation pile at different observation points of the seamless line rail risk assessment method based on vehicle-mounted inspection provided by one embodiment of the present invention;

[0063] Fig.10 This is an example diagram of rail temperature changes of a single unit rail section detected in multiple phases of the seamless line rail risk assessment method based on vehicle-mounted inspection provided by an embodiment of the present invention;

[0064] FIG11( a ) is a schematic diagram showing the distribution of rail temperature, temperature force and risk assessment results of a seamless railway rail risk assessment method based on vehicle-mounted inspection provided in an embodiment of the present invention;

[0065] FIG11( b ) is a schematic diagram of the statistical results of risk points along the entire line of the seamless line rail risk assessment method based on vehicle-mounted inspection provided by an embodiment of the present invention;

[0066] Fig.12 It is a structural schematic diagram of a seamless line rail risk assessment system based on vehicle-mounted inspection provided by an embodiment of the present invention;

[0067] Fig.13It is an example diagram of the setting principle of the seamless line rail risk assessment system based on vehicle-mounted inspection provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0068] The following will describe the implementation methods of the present invention in detail in conjunction with the accompanying drawings and embodiments, so that the implementers of the present invention can fully understand how the present invention applies technical means to solve technical problems and achieve the implementation process of technical effects and implement the present invention specifically according to the above implementation process. It should be noted that as long as there is no conflict, the various embodiments and various features of the embodiments in the present invention can be combined with each other, and the technical solutions formed are all within the protection scope of the present invention.

[0069] Although the flowcharts describe the operations as sequential processes, many of the operations may be performed in parallel, concurrently, or simultaneously. The order of the operations may be rearranged. A process may be terminated when its operations are completed, but may also have additional steps not included in the accompanying drawings. A process may correspond to a method, function, procedure, subroutine, subprogram, etc.

[0070] Computer devices include user devices and network devices. Among them, user devices or clients include but are not limited to computers, smart phones, PDAs (Personal Digital Assistants), etc.; network devices include but are not limited to a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers. Computer devices can be run alone to implement the present invention, or they can be connected to the network and implement the present invention through interactive operations with other computer devices in the network. The network where the computer device is located includes but is not limited to the Internet, wide area network, metropolitan area network, local area network, VPN network, etc.

[0071] The terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments. Unless the context clearly indicates otherwise, the singular forms "one", "one" and "item" used herein are also intended to include plural numbers. It should also be understood that the terms "include" and / or "comprise" used herein specify the existence of stated features, integers, steps, operations, units and / or components, without excluding the existence or addition of one or more other features, integers, steps, operations, units, components and / or combinations thereof.

[0072] In newly built urban rail lines, the seamless line design eliminates rail joints, helps improve track structure and reduce vibration, and is used more and more widely in newly built lines. But on the other hand, due to the disappearance of rail joints, the rails cannot expand and contract freely when the temperature changes, which will cause great temperature stress inside the rails. When the fastener pressure is not enough to offset the temperature stress, the rails will undergo longitudinal displacement, causing rail expansion and track runaway when the temperature rises, and rail breakage is likely to occur when the temperature is very low, seriously affecting driving safety.

[0073] In order to ensure operational safety and understand rail stress, it is necessary to conduct periodic inspections or real-time monitoring of seamless line rail displacement observation piles. However, the main monitoring technologies currently used include two categories. One is to use manual visual observation one by one, which is a contact measurement method. For example, through manual on-site inspection, a square ruler is used to compare and measure the relative position changes between the displacement observation pile and the rail bottom scale. However, this method requires a huge amount of human resources and time resources when used. In addition to consuming a large amount of manpower, manual inspection is difficult to accurately control the selection of the geometric center of the observation pile, resulting in large measurement errors and low detection efficiency. It also requires high professionalism of the inspection personnel, has great limitations, has many influencing factors, and has a high error rate.

[0074] The other type is to install a monitoring device at the rail joint position, which is a non-contact measurement method. For example, Chong Aixin et al. proposed a method for measuring the longitudinal displacement of seamless rails based on binocular vision technology. By printing the coded mark to be measured on the waist of the rail for distance measurement and spraying the reference target on the fixed reference facility at the side shoulder of the line, a binocular camera is set on the ground, and the image three-dimensional ranging algorithm is used to achieve accurate measurement of the longitudinal displacement of the rail. Liu Liping et al. proposed a railway rail creep displacement detection method based on monocular vision measurement technology. The image containing circular markers is collected by an image acquisition device installed on the inherent facilities next to the rail. The position of the marker in the reconstructed image is preliminarily located by the template matching method. The center of mass coordinates and radius of the marker are accurately calculated by combining edge detection and circle fitting methods. The physical displacement of the rail creep is detected based on the image imaging principle. In addition, rail temperature monitoring sensors can be installed at the joints to achieve real-time monitoring of rail temperature changes and calculate the longitudinal displacement changes of the rails. This method requires a large number of monitoring device components or sensor equipment to be installed along the line. This method has the disadvantages of large equipment investment and high installation cost. The subsequent maintenance workload is complicated and it is difficult to meet the objective needs of high-speed dynamic measurement.

[0075] To solve the above problems, the present invention provides a seamless line rail risk assessment method and system based on vehicle-mounted inspection. The method uses computer vision technology in conjunction with an image recognition algorithm to achieve rapid detection of the longitudinal displacement of the seamless line rail under normal operating conditions, obtains high-resolution image information of the seamless line observation pile through a vehicle-mounted camera, applies the Otsu algorithm to divide the foreground and background, applies edge detection to achieve shape feature extraction of the observation pile and the ruler, applies a template matching algorithm to achieve matching of the target rectangle, the target circle and the shape in the picture, and calculates the rail displacement through center extraction. The method uses machine vision technology to achieve efficient periodic detection of the longitudinal displacement of the seamless line rail on the vehicle, with a detection error of less than 1 mm and significantly improved accuracy. Based on this, the temperature stress, temperature change, temperature force and other indicators of the rails on the entire line are calculated in real time, and the risk of broken rails on the entire line is assessed.

[0076] Next, the detailed process of the method of the embodiment of the present invention is described in detail based on the accompanying drawings. The steps shown in the flowchart of the accompanying drawings can be executed in a computer system including a set of computer executable instructions. Although the logical order of each step is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0077] Embodiment 1:

[0078] Figure 1 The schematic diagram of the process of the seamless railway rail risk assessment method based on vehicle-mounted inspection provided by the first embodiment of the present invention is shown. Figure 1 It can be seen that the method includes the following steps.

[0079] Dynamic collection step: when the vehicle carrier moves relative to the measured rail, a camera device arranged at the bottom of the vehicle carrier is used to collect image data including the rail displacement observation pile and the scale;

[0080] Image segmentation step: Use the Otsu algorithm to perform binary processing on the collected image data to determine the foreground area and background area;

[0081] Feature extraction step: Determine the position and contour features of the displacement observation pile and ruler in the image based on the Canny edge detection algorithm;

[0082] Operation center determination step: normalizing the contours of the displacement observation pile and the ruler in the image with the template object, and then matching and determining the center positions of the displacement observation pile and the ruler;

[0083] Longitudinal displacement calculation steps: determine the detection straight line distance between the center of the displacement observation pile and the center of the scale, and calculate the longitudinal displacement of the corresponding seamless rail in combination with the installation distance between the center of the displacement observation pile and the center of the scale;

[0084] Real-time rail temperature analysis step: based on the longitudinal displacement combined with the expansion and contraction properties of the rail, calculating the change of the actual rail temperature at the current position of the seamless rail relative to the construction locked rail temperature, and then determining the temperature stress index at different positions of the rail based on the change, the temperature stress index includes temperature stress and full-section temperature force;

[0085] Risk assessment and processing steps: Based on the distribution of the temperature stress index, the matching rail risk assessment result is determined, and a corresponding processing solution is provided in combination with the set decision-making principle, and the output is visualized for the user.

[0086] Although some studies have adopted non-contact technology to realize the longitudinal creep displacement detection of seamless line rails in the past, they realize real-time monitoring of rail temperature changes and calculate the change in longitudinal displacement of the rails by setting rail temperature monitoring sensors at the joint positions, or rely on labels arranged on the rails and the roadbed or the ground. In addition, it is necessary to arrange a batch of image acquisition equipment beside the rails of the entire line or on the inherent facilities. The investment cost, application cost and subsequent maintenance cost are all high, which cannot meet the practical requirements in line operation and maintenance operations; and in the existing methods of measuring rail displacement by machine vision technology, binocular cameras are mostly fixedly installed relative to the ground, which is difficult to meet the objective needs of high-speed dynamic measurement, and cannot achieve timely and accurate rail risk assessment, which is easy to miss the best processing time or fail to achieve the best maintenance effect after processing.

[0087] The present invention uses a high-definition linear array camera located under the vehicle body and intelligent means such as image recognition to inspect the displacement of the seamless line rail; obtains high-resolution image information of the seamless line observation pile through the on-board camera, applies the Otsu algorithm to divide the foreground and background, applies edge detection to realize the shape feature extraction of the observation pile and the ruler, applies the template matching algorithm to realize the matching of the target rectangle, the target circle and the shape in the picture, and calculates the rail displacement through center extraction, so as to realize the rail displacement detection with an accuracy better than 1mm, and can overcome the problems or shortcomings in the above-mentioned prior art, realize high-speed dynamic measurement of the seamless line rail, based on which, the temperature change data and temperature stress parameters at the corresponding position of the rail are calculated in real time, and then the real-time evaluation of the risk status of the rail of the whole line is realized, so as to provide timely and reliable data support for the staff to realize the maintenance of the line rail, and the requirements for the professional knowledge of the technical personnel are relatively low, and there is no need to rely too much on professional experience, and it is suitable for a variety of wireless rail lines with evaluation needs, and it is more practical.

[0088] Figure 2 The following is a detailed flow chart showing the operating principle of the longitudinal displacement detection of the rail in the seamless line rail risk assessment method based on vehicle-mounted inspection provided by the embodiment of the present invention. Figure 2As shown, preferably, in one embodiment, in the dynamic acquisition step, the camera device uses a high-definition linear array laser camera arranged on the vehicle relative to the rail, and dynamically acquires a grayscale image including the rail displacement observation pile and the scale, such as Figure 3 shown.

[0089] In actual application, high-definition linear array camera equipment is used to collect displacement observation piles and scale grayscale images during the operation of the vehicle carrier. Usually, camera devices are set above the rails on both sides, such as Figure 4 As shown in the figure, the grayscale images including rail displacement observation piles and scales collected by the cameras on the left and right sides are shown in the figure. Figure 5 (a) and Figure 5 (b) is shown. The vehicle carrier used to set the camera device can be an operation and maintenance inspection vehicle or a normal operation vehicle.

[0090] Furthermore, a data processing module fixedly arranged at the bottom or inside of the vehicle carrier and communicatively connected with the dynamic acquisition module and the detection control center is used to process the collected image data and determine the centroid of the displacement observation pile and the scale in the image. Then, the longitudinal displacement of the corresponding seamless rail is calculated according to the distribution of the centroid combined with the installation distance between the displacement observation pile and the scale. Based on the calculation, the temperature stress index of the rail is determined to realize the broken rail risk assessment and visualize the output of relevant data.

[0091] Specifically, the image segmentation step is first performed, and the collected image data is binarized using the Otsu algorithm to determine the foreground area and the background area.

[0092] In an optional embodiment, in the image division step, according to the inter-class variance calculation principle of image pixel classification, the target grayscale level that maximizes the inter-class variance is obtained, and all pixels of the image are divided into two categories according to the target grayscale level to achieve division.

[0093] In practical applications, the process of using the Otsu algorithm to determine the image binary segmentation threshold includes the following operations:

[0094] In the grayscale range of 0-255, suppose there is a grayscale n, and all pixels in the image are divided into two categories I1 (grayscale < n) and I2 (grayscale ≥ n). The mean values ​​of these two categories of pixels are l1 and l2, respectively, and the global mean value of the image is l. G , the probability of a pixel being classified as I1 and I2 is p1 and p2, so:

[0095] p1l1+p2l2=l G (1)

[0096] p1+p2=1 (2)

[0097] According to the concept of variance, the expression of between-class variance is:

[0098] σ 2 =p1(l1-l G ) 2 +p2(l2-l G ) 2 (3)

[0099] Substituting (1) into (3) and simplifying it, we can get

[0100] σ 2 =p1p2(l1-l2) 2 (4)

[0101] Traverse the grayscale levels from 0 to 255, find the grayscale level n that maximizes equation (4), and use the grayscale level that maximizes the inter-class variance to divide the image.

[0102] In the formula, σ 2 represents the inter-class variance of image pixel classification, l1 and l2 represent the means of the two types of pixels respectively, and l G represents the global pixel mean of the image, p1 and p2 represent the probability of pixels being classified into the first and second categories respectively; the binary division result of the image captured by the camera device on the right is as follows Figure 6 shown.

[0103] Next, the feature extraction step is performed to determine the position and contour features of the displacement observation pile and ruler in the image based on the Canny edge detection algorithm.

[0104] Preferably, in one embodiment, in the feature extraction step, the process of determining the position and contour features of the displacement observation pile and the scale in the image includes:

[0105] Step a1: Use a Gaussian filter to smooth the image to reduce noise interference in the image;

[0106] Step a2: Use the finite difference of the first-order partial derivative to calculate the magnitude and direction of the gradient; wherein the edge is calculated in both the horizontal and vertical directions, and then the square root of the sum of the squares is calculated;

[0107] Step a3: Perform non-maximum suppression on the gradient amplitude and retain the pixel with the largest local gradient to improve the edge positioning accuracy;

[0108] Step a4: Then use the dual threshold algorithm to detect and connect edges; perform calculations based on the set high threshold and low threshold, where the high threshold is twice the low threshold, and traverse all pixel points: when the pixel value of a pixel point is less than the low threshold, set its pixel value to 0; when its pixel value is greater than the high threshold, the pixel point is a strong edge point, and its pixel value is set to 255; when its pixel value is between the low threshold and the high threshold, the pixel point is a weak edge point, and if there is a strong edge point in its 8-neighborhood domain, set the pixel value of this point to 255, and if there is no strong edge point in its 8-neighborhood domain, set the pixel value of this point to 0. Connect the strong edge points to obtain the contour features and positions of the rectangular ruler and the circular observation pile; the edge feature recognition results of the image captured by the camera device on the right are shown in Figure 1. Figure 7 shown.

[0109] Furthermore, the operation center determination step is executed, and after the outlines of the displacement observation pile and the ruler in the image are normalized with the template object, the center positions of the displacement observation pile and the ruler are determined by matching.

[0110] In actual application, the objects to be matched and the template objects are normalized, wherein the objects to be matched are displacement detection piles and scale objects in the image, and the template objects are pre-set displacement detection piles and scale template graphics that are consistent with the model and structural parameters of the objects to be matched;

[0111] Then, the center positions of the rectangular scale and the circular observation pile are determined by matching the object to be matched with the template object.

[0112] Next, the longitudinal displacement calculation step is used to determine the detection straight-line distance between the center of the displacement observation pile and the center of the scale, and the longitudinal displacement of the corresponding seamless rail is calculated in combination with the installation distance between the center of the displacement observation pile and the center of the scale.

[0113] The longitudinal displacement of the rail can be calculated based on the installation distance between the center of the displacement observation pile and the center of the scale, and the detection distance between the center of the displacement observation pile and the center of the scale.

[0114] like Figure 8 As shown, A is the center of the displacement observation pile, B is the center of the scale, b is the installation distance between the center of the displacement observation pile and the center of the scale, c is the detection distance between the center of the displacement observation pile and the center of the scale, and a is the longitudinal displacement of the rail.

[0115] In the longitudinal displacement calculation step, the longitudinal displacement a of the corresponding seamless rail is calculated according to the following formula:

[0116]

[0117] Where b is the installation distance between the geometric center of the displacement observation pile and the geometric center of the scale, and c is the detection straight-line distance between the center of the displacement observation pile and the center of the scale.

[0118] Furthermore, a real-time rail temperature analysis step is performed to calculate the change in the actual rail temperature at the current position of the seamless rail relative to the construction locked rail temperature based on the longitudinal displacement combined with the expansion and contraction properties of the rail, and then determine the temperature stress index at different positions of the rail based on it, and the temperature stress index includes temperature stress and full-section temperature force.

[0119] In an optional embodiment, the change in the actual rail temperature at the current position of the seamless rail relative to the construction locked rail temperature is calculated based on the longitudinal displacement according to the following formula:

[0120] The change of the actual rail temperature relative to the construction locking rail temperature is:

[0121] Δt=ΔL / (a·L)

[0122] Where: ΔL is the expansion and contraction of the rail (mm); a is the expansion coefficient of the rail, a = 0.0118mm / (m·℃), that is, for every meter of rail, when the rail temperature changes by 1℃, the rail will expand and contract by 0.0118mm; L is the length of the rail (m); Δt is the change in actual rail temperature relative to the construction locked rail temperature (℃).

[0123] The actual rail temperature is further calculated based on the rail temperature change and the measurement result of the rail temperature locked in the entire line at the end of the construction:

[0124] Tr=T0-Δt

[0125] Where: Tr. is the actual rail temperature (℃); T0. is the construction locked rail temperature (℃).

[0126] Based on this, the rail temperature change parameter of the actual rail temperature relative to the designed locked rail temperature is calculated according to the following formula based on the designed locked rail temperature parameters:

[0127] △t d =Tr-Td

[0128] Where: Td is the measured design locking rail temperature (℃); △t d It is the rail temperature variation range (℃) of the actual rail temperature relative to the designed locked rail temperature.

[0129] The following is an example of the distribution of the left and right rail temperature, longitudinal displacement, and displacement difference between the previous observation pile and the previous observation pile at different observation points of the seamless line rail risk assessment method based on vehicle-mounted inspection provided by the embodiment of the present invention: Fig. 9 shown.

[0130] Furthermore, considering that there is a correlation between the rail temperature change data and the compressive stress or tensile stress generated on the rail section per square centimeter, when the rail temperature rises or falls by 1°C, the compressive stress or tensile stress generated on the rail section per square centimeter of the locked rail is 247.8N. Therefore, the calculation formula for determining the rail temperature stress is as follows:

[0131] σ t =247.8·Δt

[0132] The temperature stress on the whole section of the seamless railway rail is called temperature force. Optionally, in one embodiment, the whole section temperature force of the rail is calculated according to the following formula:

[0133] P t =σ t ·S 钢轨 =247.8·Δt·S 钢轨

[0134] Where: σt is the temperature stress (N); Δt is the rail temperature variation (℃); Srail is the rail cross-sectional area (cm 2 ); Pt is temperature force (N); an example diagram of rail temperature change of a single unit rail section in a multi-stage detection method for seamless line rail risk assessment based on vehicle-mounted inspection provided by an embodiment of the present invention is shown in FIG. Fig.10 As shown in the figure, the horizontal axis is the line mileage.

[0135] According to the railway standards in the field, the actual rail temperature of high-speed railway rails should be within the specified range of the designed locked rail temperature for that section of the line. Generally, the actual rail temperature is allowed to fluctuate by ±5°C relative to the designed locked rail temperature. The allowable fluctuation value for seamless lines on bridges or in cold areas is ±3°C.

[0136] In addition, in order to accurately ensure the safe operation of trains, preferably, the line rails should also meet the following requirements: for adjacent unit rail sections, the difference in actual rail temperature should be less than 5°C; for the left rail and the right rail of the same unit rail section, the difference in actual rail temperature should be kept within 3°C; for the rails of the unit rail sections in the same section, the peak-to-peak value of the rail temperature should be less than 10°C.

[0137] In summary, in a preferred embodiment, the present invention uses the displacement difference of adjacent points, the displacement of a single point, the variation parameter of the actual rail temperature relative to the designed locked rail temperature, the actual rail temperature difference between adjacent unit rail sections, the actual rail temperature difference between the left and right rails of the same unit rail section, the highest and lowest actual rail temperature difference in the same interval, and the recent two-period rail temperature difference of a single point as independent risk assessment indicators. The recent two-period rail temperature difference of a single point represents the difference between the actual rail temperature results obtained by two period detections. In the actual identification process, the risk status of the current rail is determined according to the following principles:

[0138] (1) If the current rail has any of the following conditions, it is determined to be Level I overrun, and an overrun reminder is generated and output simultaneously:

[0139] Displacement difference between adjacent points ≥±5mm,

[0140] Single point displacement ≥ ±5mm,

[0141] The actual rail temperature difference between adjacent unit rail sections is ≥±3℃,

[0142] The actual rail temperature difference between the left and right rails of the same rail section is ≥±2℃,

[0143] The highest and lowest actual rail temperature difference within the same section is ≥±6℃,

[0144] The actual rail temperature varies with the designed locked rail temperature by ≥±3℃;

[0145] (2) If the current rail has any of the following conditions, it is determined to be over-limit level II; while giving an over-limit reminder, the stress relief construction process should be arranged immediately:

[0146] Displacement difference between adjacent points ≥ ±10mm,

[0147] Single point displacement ≥ ±10mm,

[0148] The actual rail temperature difference between adjacent unit rail sections is ≥±5℃,

[0149] The actual rail temperature difference between the left and right rails of the same rail section is ≥±3℃,

[0150] The difference between the highest and lowest actual rail temperature in the same section is ≥±10℃,

[0151] The actual rail temperature varies with the designed locked rail temperature by ≥±5℃.

[0152] A schematic diagram of the distribution of rail temperature, temperature force and risk assessment results of the seamless line rail risk assessment method based on vehicle-mounted inspection provided by an embodiment of the present invention is shown in FIG11( a ), and a schematic diagram of the statistical results of risk points along the entire line is shown in FIG11( b ).

[0153] Next, risk assessment and processing steps are performed based on the calculated indicators. Specifically, the matching rail risk assessment results are determined based on the distribution of the temperature stress indicators, and a corresponding processing solution is provided in combination with the set decision-making principles.

[0154] Compared with the prior art, the present invention can realize the rapid detection of the longitudinal displacement of the seamless line rail at the conventional detection running speed through computer vision technology and image recognition algorithm, obtain the high-resolution image information of the seamless line observation pile through the laser camera, divide the foreground and background by the Otsu algorithm, realize the shape feature extraction of the observation pile and the ruler by edge detection, realize the matching of the target rectangle, the target circle and the shape in the picture by the template matching algorithm, obtain the centroid of the displacement observation pile and the rail bottom ruler by the target graphic matching algorithm, calculate the longitudinal creep of the rail in combination with the Pythagorean theorem, then calculate the temperature stress, temperature change and temperature force of the rail, finally realize the real-time landing of data through the vehicle network, and judge the risk of rail breakage of the whole line in real time through the data management and analysis module, and provide corresponding treatment plan guidance, improve the efficiency of risk assessment of seamless line rail, reduce labor costs, and effectively improve the efficiency of subway operation and maintenance.

[0155] Preferably, in one embodiment, in order to facilitate technical personnel to intuitively understand the real-time data of the rails of the entire line, the present invention also displays indicator data such as the longitudinal creep of the rails and the change in rail temperature in real time through a visual interface, and outputs risk assessment results and corresponding maintenance treatment plan strategies.

[0156] For the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the order of the actions described, because according to the present invention, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0157] It should be pointed out that in other embodiments of the present invention, the method can also obtain a new seamless line rail risk assessment method based on vehicle-mounted inspection by combining one or several of the above-mentioned embodiments, so as to realize accurate monitoring and maintenance of the operating status of seamless line rails.

[0158] It should be noted that, based on the method in any one or more of the above-mentioned embodiments of the present invention, the present invention also provides a storage medium, on which is stored a program code that can implement the method as described in any one or more of the above-mentioned embodiments, and when the code is executed by the operating system, it can implement the seamless line rail risk assessment method based on vehicle-mounted inspection as described above.

[0159] Embodiment 2:

[0160] The method is described in detail in the embodiments disclosed in the above invention. The method of the present invention can be implemented by various forms of devices or systems. Therefore, based on other aspects of the method described in any one or more embodiments, the present invention also provides a seamless line rail risk assessment system based on vehicle-mounted inspection, which is used to execute the seamless line rail risk assessment method based on vehicle-mounted inspection described in any one or more embodiments. Specific embodiments are given below for detailed description.

[0161] Specifically, Fig.12 FIG. 4 shows a schematic diagram of the structure of a seamless railway rail risk assessment system based on vehicle-mounted inspection provided in an embodiment of the present invention. Fig.12 As shown, the system includes:

[0162] A dynamic acquisition module, comprising a camera device and a communication unit disposed at the bottom of the vehicle carrier, which is configured to dynamically acquire image data including rail displacement observation piles and scales and transmit the image data to the data processing module during the movement of the vehicle carrier relative to the measured rail;

[0163] The data processing module is arranged at the bottom or inside of the vehicle carrier, and is communicated with the dynamic acquisition module and the detection control center. It is used to process the collected image data and determine the centroid of the displacement observation pile and the ruler in the image, and then calculate the longitudinal displacement of the corresponding seamless rail according to the distribution of the centroid combined with the installation distance between the displacement observation pile and the ruler. Based on the calculation, the temperature stress index of the rail is determined to realize the broken rail risk assessment and visualize the relevant data.

[0164] Further, Fig.13 FIG. 4 shows an example diagram of the setting principle of the seamless line rail risk assessment system based on vehicle-mounted inspection provided in an embodiment of the present invention. Fig.13 As shown, the dynamic acquisition module adopts an image acquisition machine, and the data processing module is configured in the form of a data processing machine. In an optional embodiment, the data processing module adopts a data processing machine, including:

[0165] The image division unit is configured to use the Otsu algorithm to perform binarization processing on the collected image data to divide and determine the foreground area and the background area.

[0166] The feature extraction unit is configured to determine the position and contour features of the displacement observation pile and the ruler in the image based on the Canny edge detection algorithm.

[0167] A calculation center determination unit is configured to normalize the contours of the displacement observation pile and the scale in the image and the template object, and then match and determine the center positions of the displacement observation pile and the scale;

[0168] A longitudinal displacement calculation unit, configured to determine a detection straight-line distance between the center of the displacement observation pile and the center of the scale, and calculate the longitudinal displacement of the corresponding seamless rail in combination with the installation distance between the center of the displacement observation pile and the center of the scale;

[0169] A real-time rail temperature analysis unit is configured to calculate the change of the actual rail temperature at the current position of the seamless rail relative to the construction locked rail temperature based on the longitudinal displacement combined with the expansion and contraction properties of the rail, and then determine the temperature stress index at different positions of the rail based on the change, wherein the temperature stress index includes temperature stress and full-section temperature force;

[0170] The risk assessment and processing unit is configured to determine the matching rail risk assessment result based on the distribution of the temperature stress index, and provide a corresponding processing solution in combination with the set decision-making principle, and output it visually to the user.

[0171] Preferably, in one embodiment, the camera device of the dynamic acquisition module adopts a high-definition linear array laser camera installed on the vehicle relative to the rails, and dynamically acquires a grayscale image including the rail displacement observation piles and the scale.

[0172] Furthermore, in one embodiment, the image division unit is configured to obtain a target grayscale level that maximizes the inter-class variance based on an inter-class variance calculation principle for image pixel classification, and to divide all pixels of the image into two categories based on the target grayscale level to achieve division.

[0173] Specifically, in an optional embodiment, the image segmentation unit calculates the inter-class variance of image pixel classification according to the following formula:

[0174] σ 2 =p1(l1-l G ) 2 +p2(l2-l G ) 2

[0175] Among them, p1l1+p2l2=l G , then: 2 =p1p2(l1-l2) 2

[0176] In the formula, σ 2 represents the inter-class variance of image pixel classification, l1 and l2 represent the means of the two types of pixels respectively, and l G represents the global pixel mean of the image, p1 and p2 represent the probability of pixels being classified into the first and second categories respectively.

[0177] Furthermore, in one embodiment, the feature extraction unit determines the position and contour features of the displacement observation pile and the scale in the image according to the following operations:

[0178] Smooth the image using a Gaussian filter;

[0179] Use finite differences of first-order partial derivatives to calculate the magnitude and direction of the gradient; calculate the edge in both the horizontal and vertical directions, and then take the square root of the sum of the squares;

[0180] Perform non-maximum suppression on the gradient amplitude;

[0181] Then, a dual threshold algorithm is used to detect and connect the edges to obtain the contour features and positions of the rectangular scale and circular observation pile.

[0182] Preferably, in one embodiment, the longitudinal displacement calculation unit calculates the longitudinal displacement a of the seamless rail according to the following formula:

[0183]

[0184] Where b is the installation distance between the geometric center of the displacement observation pile and the geometric center of the scale, and c is the detection straight-line distance between the center of the displacement observation pile and the center of the scale.

[0185] In an optional embodiment, the real-time rail temperature analysis unit is configured to calculate the change in the actual rail temperature of the seamless rail at the current position relative to the construction locked rail temperature based on the longitudinal displacement according to the following formula:

[0186] Δt=ΔL / (a·L)

[0187] Where: ΔL is the expansion of the rail (mm); a is the expansion coefficient of the rail, a = 0.0118mm / (m·℃), that is, for every meter of rail, when the rail temperature changes by 1℃, the rail will expand and contract by 0.0118mm; L is the length of the rail (m); Δt is the change in the actual rail temperature relative to the construction locked rail temperature (℃);

[0188] Furthermore, considering that there is a correlation between the rail temperature change data and the compressive stress or tensile stress generated on the rail section per square centimeter, specifically when the rail temperature rises or falls by 1°C, the compressive stress or tensile stress generated on the rail section per square centimeter of the locked rail is 247.8N, it is determined that the real-time rail temperature analysis unit calculates the rail temperature stress according to the following formula:

[0189] σ t =247.8·Δt

[0190] Where: σt is the temperature stress (N); Δt is the variation of the actual rail temperature relative to the construction locking rail temperature (℃).

[0191] The temperature stress on the whole section of the seamless line rail is called temperature force. Optionally, in one embodiment, the real-time rail temperature analysis unit calculates the whole section temperature force of the rail according to the following formula:

[0192] P t =σt ·S 钢轨 =247.8·Δt·S 钢轨

[0193] Where: S rail is the cross-sectional area of ​​the rail (cm 2 );Pt is temperature force (N).

[0194] Furthermore, the risk assessment and processing unit is activated to determine the matching rail risk assessment result based on the distribution of the temperature stress index, and provide a corresponding processing solution in combination with the set decision-making principle.

[0195] The present invention uses a camera device arranged at the bottom of the vehicle carrier to dynamically collect image data including rail displacement observation piles and rulers; after dividing the collected image data to determine the foreground area and the background area, the contour features of the displacement observation piles and rulers in the image are determined based on the edge detection algorithm; then the displacement observation piles and rulers in the image are normalized with the template object, and the center positions of the two are matched to determine; after determining the detection straight-line distance between the center of the displacement observation pile and the center of the ruler, the longitudinal displacement of the corresponding seamless rail is calculated in combination with the installation distance between the center of the displacement observation pile and the center of the ruler. This solution can overcome the problems of long detection time, high error probability and high cost in the prior art. Computer vision technology cooperates with image recognition algorithm to realize accurate and efficient detection of the longitudinal displacement of the rails in the running state, so as to analyze and calculate the real-time temperature stress index data of the rails, conduct risk assessment of the rails of the entire line, provide technical personnel with more timely and accurate data support and operation guidance, and effectively reduce costs.

[0196] In the seamless line rail risk assessment system based on vehicle-mounted inspection provided by the embodiment of the present invention, each module or unit structure can operate independently or in combination according to actual information collection requirements and data calculation requirements to achieve corresponding technical effects.

[0197] It should be understood that the embodiments disclosed in the present invention are not limited to the specific structures, processing steps or materials disclosed herein, but should be extended to equivalent substitutions of these features understood by ordinary technicians in the relevant field. It should also be understood that the terms used herein are only used for the purpose of describing specific embodiments and are not meant to be limiting.

[0198] The "one embodiment" mentioned in the specification means that a particular feature, structure or characteristic described in conjunction with the embodiment is included in at least one embodiment of the present invention. Therefore, the phrase "one embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment.

[0199] Although the embodiments disclosed in the present invention are as above, the above contents are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present invention, but the patent protection scope of the present invention shall still be subject to the scope defined in the attached claims.

Claims

1. A seamless railway rail risk assessment method based on vehicle-mounted inspection, characterized in that: The method comprises: Dynamic collection step: when the vehicle carrier moves relative to the measured rail, a camera device arranged at the bottom of the vehicle carrier is used to collect image data including the rail displacement observation pile and the scale; Image segmentation step: Use the Otsu algorithm to perform binary processing on the collected image data to determine the foreground area and background area; Feature extraction step: Determine the position and contour features of the displacement observation pile and ruler in the image based on the Canny edge detection algorithm; The calculation center determination step: normalize the contours of the displacement observation pile and the ruler in the image and the template object respectively, and then match and determine the center positions of the displacement observation pile and the ruler; Longitudinal displacement calculation steps: determine the detection straight line distance between the center of the displacement observation pile and the center of the scale, and calculate the longitudinal displacement of the corresponding seamless rail in combination with the installation distance between the center of the displacement observation pile and the center of the scale; Real-time rail temperature analysis step: based on the longitudinal displacement combined with the expansion and contraction properties of the rail, calculating the change of the actual rail temperature at the current position of the seamless rail relative to the construction locked rail temperature, and then determining the temperature stress index at different positions of the rail based on the change, the temperature stress index includes temperature stress and full-section temperature force; Risk assessment and processing steps: determine the matching rail risk assessment results based on the distribution of the temperature stress index, provide corresponding processing solutions in combination with the set decision-making principles, and output them visually to the user.

2. The method according to claim 1, characterized in that In the dynamic acquisition step, the camera device uses a high-definition linear array laser camera that is installed on the vehicle relative to the rails to dynamically acquire a grayscale image including the rail displacement observation piles and the scale.

3. The method according to claim 1, characterized in that In the image division step, according to the inter-class variance calculation principle of image pixel classification, the target grayscale level that maximizes the inter-class variance is obtained, and all pixels of the image are divided into two categories according to the target grayscale level to achieve division.

4. The method according to claim 1, characterized in that: The inter-class variance of image pixel classification is calculated according to the following formula: σ 2 =p1(l1-l G ) 2 +p2(l2-l G ) 2 Among them, p1l1+p2l2=l G ,but: σ 2 =p1p2(l1-l2) 2 In the formula, σ 2 represents the inter-class variance of image pixel classification, l1 and l2 represent the means of the two types of pixels respectively, and l G represents the global pixel mean of the image, p1 and p2 represent the probability of pixels being classified into the first and second categories respectively.

5. The method according to claim 1, characterized in that In the feature extraction step, the process of determining the position and contour features of the displacement observation pile and the scale in the image includes: Smooth the image using a Gaussian filter; Use finite differences of first-order partial derivatives to compute the magnitude and direction of the gradient; Perform non-maximum suppression on the gradient amplitude; Then, a dual threshold algorithm is used to detect and connect the edges to obtain the contour features and positions of the rectangular scale and circular observation pile.

6. The method according to claim 1, characterized in that In the longitudinal displacement calculation step, the longitudinal displacement a of the corresponding seamless rail is calculated according to the following formula: Where b is the installation distance between the geometric center of the displacement observation pile and the geometric center of the scale, and c is the detection straight-line distance between the center of the displacement observation pile and the center of the scale.

7. The method according to claim 1, characterized in that In the real-time rail temperature analysis step, the change parameter of the actual rail temperature at the current position of the seamless rail relative to the construction locked rail temperature is calculated based on the longitudinal displacement according to the following formula: Δt=ΔL / (a·L) Wherein, Δt is the variation of actual rail temperature relative to the construction locked rail temperature (℃); ΔL is the expansion and contraction of the rail (mm); a is the expansion coefficient of the rail, a=0.0118mm / (m·℃), that is, for every meter of rail, when the rail temperature changes by 1℃, the rail will expand and contract by 0.0118mm; L is the length of the rail (m).

8. The method according to claim 1, characterized in that The full-section temperature force of the rail is calculated according to the following formula: P t =σ t ·S 钢轨 =247.8·Δt·S 钢轨 Among them, σ t =247.8·Δt Where: Pt is the temperature force (N); S 钢轨 is the cross-sectional area of ​​the rail (cm2); σt is the temperature stress (N); Δt is the variation of the actual rail temperature relative to the construction locked rail temperature (℃).

9. A storage medium, characterized in that: The storage medium stores program codes that can implement the method as claimed in any one of claims 1 to 8.

10. A seamless rail risk assessment system based on vehicle-mounted inspection, characterized in that: The system executes the method according to any one of claims 1 to 8, and the system comprises: A dynamic acquisition module, comprising a camera device and a communication unit disposed at the bottom of the vehicle carrier, which is configured to dynamically acquire image data including rail displacement observation piles and scales and transmit the image data to the data processing module during the movement of the vehicle carrier relative to the measured rail; The data processing module is fixedly arranged at the bottom or inside of the vehicle carrier, and is connected to the dynamic acquisition module and the detection control center for communication. It is used to process the acquired image data and determine the centroid of the displacement observation pile and the scale in the image, and then calculate the longitudinal displacement of the corresponding seamless rail according to the distribution of the centroid combined with the installation distance between the displacement observation pile and the scale, and determine the rail temperature stress index based on the calculation, so as to realize the risk assessment of broken rails and perform visual output of relevant data; The data processing module comprises: An image division unit is configured to perform binarization processing on the collected image data using the Otsu algorithm to divide and determine the foreground area and the background area; A feature extraction unit configured to determine the position and contour features of the displacement observation pile and the scale in the image based on a Canny edge detection algorithm; A calculation center determination unit is configured to normalize the contours of the displacement observation pile and the scale in the image and the template object, and then match and determine the center positions of the displacement observation pile and the scale; A longitudinal displacement calculation unit, configured to determine a detection straight-line distance between the center of the displacement observation pile and the center of the scale, and calculate the longitudinal displacement of the corresponding seamless rail in combination with the installation distance between the center of the displacement observation pile and the center of the scale; A real-time rail temperature analysis unit is configured to calculate the change of the actual rail temperature at the current position of the seamless rail relative to the construction locked rail temperature based on the longitudinal displacement combined with the expansion and contraction properties of the rail, and then determine the temperature stress index at different positions of the rail based on the change, wherein the temperature stress index includes temperature stress and full-section temperature force; The risk assessment and processing unit is configured to determine the matching rail risk assessment result based on the distribution of the temperature stress index, provide a corresponding processing solution in combination with the set decision principle, and output it visually to the user.

Images