Intelligent vehicle-mounted seamless track steel rail displacement detection device and method

Abstract

The invention discloses an intelligent vehicle-mounted seamless track steel rail displacement detection device and method, and relates to the technical field of track displacement detection devices.The intelligent vehicle-mounted seamless track steel rail displacement detection device comprises an image acquisition device, a vehicle-mounted host and a vehicle-mounted support, the image acquisition device is used for acquiring image information of the target points on the rail web and the target points on the foundation pile on one side of the rail; the on-board host is arranged in a carriage of the locomotive; the image acquisition device is in communication connection with the vehicle-mounted host and transmits image information to the vehicle-mounted host, and the vehicle-mounted host calculates the displacement of a track according to the image information; and the vehicle-mounted bracket is connected outside the locomotive and is used for providing a mounting position for the image acquisition device. The displacement detection device is installed on the locomotive, the displacement of the rail along the line is detected while the locomotive runs, manual one-by-one measurement on the ground is replaced, and the detection efficiency is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of track displacement detection devices, and in particular to an intelligent vehicle-mounted seamless line rail displacement detection device and method. Background Art

[0002] The railway engineering department mainly uses manual handheld equipment to measure the displacement of seamless line rails. First, displacement observation piles are buried next to the line, and a scale is attached to the corresponding rail waist. The inspection personnel fix the point laser or line laser detection equipment on the interface of the displacement observation pile. During measurement, the laser is irradiated on the rail waist scale, and the rail displacement is obtained by reading the reading on the scale. At least 3 people or more are required to measure, and the measuring equipment must be horizontal and perpendicular to the rail waist. The measurement of the displacement of existing seamless line rails is mainly based on manual measurement. During measurement, the detection equipment is placed on the displacement observation pile, and the equipment status is adjusted by manual experience. This detection method is not only inefficient and has high labor costs, but the detection accuracy is easily affected by many uncontrollable factors such as manual detection experience and equipment placement posture, which makes the detection efficiency low and repeatability difficult to guarantee. The problem is more prominent. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent vehicle-mounted seamless line rail displacement detection device and method to solve the problems existing in the above-mentioned prior art and improve efficiency and detection accuracy.

[0004] To achieve the above object, the present invention provides the following solutions: The present invention provides an intelligent vehicle-mounted seamless line rail displacement detection device, comprising: an image acquisition device, a vehicle-mounted host and a vehicle-mounted bracket, wherein the image acquisition device is used to acquire image information of target points on the rail waist and target points on a foundation pile on one side of the track; the vehicle-mounted host is arranged in a locomotive compartment; the image acquisition device is communicatively connected with the vehicle-mounted host and transmits image information to the vehicle-mounted host, and the vehicle-mounted host calculates the displacement of the track according to the image information; the vehicle-mounted bracket is connected to the outside of the locomotive and is used to provide an installation position for the image acquisition device.

[0005] Preferably, the image acquisition device includes three groups of binocular cameras, and the three groups of binocular cameras respectively acquire images of the two rail waists of the double-strand parallel rails and the target points on one side of the pile foundation, the on-board host establishes a pile foundation coordinate system according to the target points reconstructed from the pile foundation images, the on-board host establishes a local rail waist coordinate system according to the target points reconstructed from the rail waist images, and the on-board host performs coordinate system conversion on the reconstructed rail waist target points and the pile foundation target points to obtain the displacement of the rail.

[0006] Preferably, the two cameras in each group of binocular cameras are connected by a carbon fiber cross bar, wherein one group of binocular cameras is arranged on one side of the vehicle-mounted bracket, and the two carbon fiber cross bars in the other two groups of binocular cameras are connected through a solid column and then connected to the other side of the vehicle-mounted bracket.

[0007] Preferably, each camera is equipped with a coaxial light source, and a filter is configured inside the camera.

[0008] Preferably, each of the cameras is equipped with a constant temperature regulating mechanism.

[0009] Preferably, the vehicle-mounted host calibrates the three groups of binocular cameras in the same coordinate system in advance, and calculates the displacement of the track according to a point cloud registration method and a coordinate conversion method.

[0010] Preferably, it also includes a trigger controller, which is placed inside the carriage and controls the synchronous triggering of multiple groups of binocular cameras and multiple light sources.

[0011] The present invention also provides an intelligent vehicle-mounted seamless line rail displacement detection method, comprising: Preparation steps: Arrange target points on the piles and on the rail waist corresponding to the piles; install the detection device on the locomotive; obtain the spatial relative position relationship of each camera through global calibration in advance; Testing steps: S1, the detection device moves with the movement of the locomotive; S2: The onboard host controls the binocular camera to start taking pictures continuously when it is within 10 meters of the pile according to the GPS and locomotive mileage information, and stops taking pictures when it is 10 meters away from the pile; S3, the vehicle host pre-processes the acquired images to remove noise and improve image quality, and numbers each image according to the shooting order, and calculates the center position of the target point using the Hough transform algorithm; S4. Perform three-dimensional reconstruction of the pile target point and the rail waist target point, perform point cloud registration and alignment on the point set Q1 reconstructed from the pile target point image and the standard point set M1 used in the pile target point, calculate the transformation matrices R1 and T1. Since the system has completed global calibration, the three camera groups are rigidly connected and do not change their positions. Therefore, the reconstructed rail waist target point coordinates are converted to the pile coordinate system with the same transformation relationships R1 and T1, thereby realizing the unity of the coordinate systems of the rail waist target point and the pile target point. In the pile coordinate system, the spatial position relationship of the rail waist target point relative to the pile target point can be calculated, that is, the pile coordinate system is further converted to the rail waist target point through point-line-plane coordinate conversion, so as to calculate the total offset Y of the track.

[0012] Preferably, the total offset Y of the rail waist target point relative to the base pile target point in three-dimensional space can be calculated in S4. However, when the locomotive passes through the track, it exerts force on the track, causing the track attitude to change, thereby affecting the measurement accuracy. The track attitude compensation algorithm is used, that is, the total offset Y and the angle between the total offset vector and the longitudinal direction of the track are calculated. Calculate the displacement y in the longitudinal direction of the track; .

[0013] Preferably, when there is a need to calculate the locked rail temperature, after the longitudinal displacement calculation is completed, according to the formula: △t = y÷ al T = △t + t0 Calculate the current track temperature, where al is the conversion coefficient, y is the longitudinal displacement of the track, △t is the temperature change, T is the current track temperature, and t0 is the locked track temperature.

[0014] Compared with the prior art, the present invention has achieved the following technical effects: 1. The present invention installs a displacement detection device on a locomotive to detect the displacement of rails along the line while the locomotive is running, replacing manual measurement one by one on the ground, greatly improving detection efficiency.

[0015] 2. Use binocular vision algorithm to measure the total displacement and longitudinal displacement of the track, greatly improving the detection accuracy.

[0016] 3. Calculate the current track temperature through the longitudinal displacement of the track.

[0017] 4. Photogrammetry technology is used to calibrate multiple sets of binocular cameras. The relative position relationship of the cameras remains unchanged, and the locomotive vibration has no effect on the measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 A schematic diagram of an application scenario of an intelligent vehicle-mounted seamless railway rail displacement detection device provided by an embodiment of the present invention; Figure 2 This is a front view of the binocular camera and the carbon fiber rod; Figure 3 It is a schematic diagram of establishing a coordinate system on the rail waist plane; In the figure: 1-vehicle bracket; 2-track; 3-pile; 4-binocular camera; 41-camera; 5-carbon fiber crossbar; 6-solid column. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] Combine the following Figures 1 to 3 , describing an embodiment of the present invention.

[0023] The present invention provides an intelligent vehicle-mounted seamless line rail displacement detection device, comprising: an image acquisition device, a vehicle-mounted host and a vehicle-mounted bracket 1, wherein the image acquisition device is used to acquire image information of a target point on a rail waist and a target point on a foundation pile 3 on one side of a track 2; the vehicle-mounted host is arranged in a locomotive compartment; the image acquisition device is communicatively connected with the vehicle-mounted host and transmits the image information to the vehicle-mounted host, and the vehicle-mounted host calculates the displacement of the track 2 according to the image information; the vehicle-mounted bracket 1 is connected to the outside of the locomotive to provide an installation position for the image acquisition device, and the vehicle-mounted bracket 1 is fixed to a bottom cross beam of the locomotive by welding.

[0024] The present invention installs the displacement detection device on the locomotive, and detects the displacement of the rails along the line while the locomotive is running, replacing manual measurement one by one on the ground, thereby greatly improving the detection efficiency.

[0025] The locomotive may be a locomotive that originally runs on the track 2 , such as various trains. In addition, the locomotive may also be a track inspection vehicle, which usually runs on the track 2 and inspects the condition of the track 2 .

[0026] The foundation pile 3 is a reference for measuring displacement in the measurement system. Any fixed object with a certain height that remains stationary for a long time beside the track 2 can be used as the foundation pile 3, which can be a contact network pole beside the track or a displacement observation pile. In the later implementation of the scheme, the contact network pole can be used as the foundation pile 3. The present invention uses the marking point on the single-side foundation pile 3 as a reference to complete the measurement of two tracks 2 at the same time.

[0027] In some embodiments, the image acquisition device includes three groups of binocular cameras 4, which respectively acquire images of the two rail waists of the double parallel tracks and the target points on the piles 3 on one side. The vehicle-mounted host establishes a pile coordinate system according to the target points reconstructed from the pile images. The vehicle-mounted host establishes a local rail waist coordinate system according to the target points reconstructed from the rail waist images. The vehicle-mounted host performs coordinate system conversion on the reconstructed rail waist target points and the pile target points to obtain the displacement of the track.

[0028] The present invention utilizes binocular vision algorithm to measure the total displacement and longitudinal displacement of the track, greatly improving the detection accuracy.

[0029] In some embodiments, the two cameras in each group of binocular cameras 4 are connected by a carbon fiber cross bar 5, wherein one group of binocular cameras 4 is arranged on one side of the vehicle bracket 1, and the two carbon fiber cross bars 5 in the other two groups of binocular cameras 4 are connected by a solid column 6 and then connected to the other side of the vehicle bracket 1, and the solid column 6 is made of aluminum or alloy material.

[0030] The embodiment of the present invention realizes a rigid connection between all cameras, so that the relative position between each camera can be fixed and unchanged, and multiple groups of binocular cameras 4 are calibrated in advance, and the vibration of the locomotive has no effect on the measurement accuracy.

[0031] In some embodiments, in order to increase brightness and filter out stray light, each camera is equipped with a coaxial light source. The coaxial light source is not limited to blue light, infrared fill light, etc. The camera uses a corresponding filter inside to filter out stray light to the maximum extent.

[0032] In some embodiments, since changes in the internal temperature of the camera will cause changes in the optical axis, thereby affecting the measurement accuracy, the embodiments of the present invention add a constant temperature adjustment mechanism to ensure that the temperature inside the camera can quickly reach a constant temperature without the need for preheating, and to reduce the impact of the ambient temperature on the normal operation of the camera.

[0033] In some embodiments, the vehicle-mounted host pre-calibrates the three sets of binocular cameras 4 in the same coordinate system, and calculates the displacement of the track 2 according to the point cloud registration method and the coordinate conversion method.

[0034] Specifically, for each group of two binocular cameras, the Zhang Zhengyou calibration method is used to calibrate the internal and external parameters of the camera using a high-strength aluminum calibration plate; after the calibration of the two cameras of a single group of cameras is completed, a unified global calibration field is established. The global calibration field is mainly used for the global calibration of the three groups of binocular cameras 4 in the system. The calibration field contains the foundation piles 3 with target points, the laid coded points and non-coded points, the rail waist target points, and the standard measuring ruler with target points. Photogrammetry technology is used to obtain the image of the calibration field, and then the coded points and non-coded points on the image are reconstructed. The scale can be obtained according to the actual length and the reconstructed length of the standard measuring ruler, and then the reconstruction result is scaled to finally obtain the point set P. Then the point set Q reconstructed by each group of binocular cameras 4 is obtained, and the two point clouds are registered to obtain the transformation relationship (rotation and translation matrix) between the two point clouds. This transformation relationship represents the relative transformation relationship from the local coordinate system to the global coordinate system in each group of binocular cameras 4, so that all cameras are calibrated in one coordinate system. The global calibration obtains the spatial relative position relationship between each group.

[0035] Then the center position of the target point is calculated using the Hough transform algorithm, and the three-dimensional reconstruction of the target point of pile 3 and the target point of the rail waist is performed. The point set Q1 of the reconstructed image of the target point of pile 3 collected by the binocular camera 4 of pile 3 is aligned with the standard point set M1 used in the target point of pile 3 for point cloud registration, and the transformation matrices R1 and T1 are calculated.

[0036] Since the system has completed global calibration, the camera used to shoot the rail waist target point and the camera used to shoot the target point of the pile 3 are rigidly connected and do not change position. Therefore, the target point coordinates reconstructed by the two sets of binocular cameras 4 are converted to the pile 3 coordinate system with the same transformation relationship R1 and T1, so as to realize the coordinate system of the rail waist target point and the pile 3 target. Through point-line-surface coordinate conversion, the pile 3 coordinate system is further converted to the rail waist target point, so as to calculate the displacement of the track 2.

[0037] In addition, when the locomotive passes over the track 2, a certain force will be exerted on the track 2, causing the track 2 to change its posture. Therefore, the above process can calculate the total offset Y of the rail waist target point relative to the target point of the pile 3 in three-dimensional space. If the locked rail temperature is to be calculated, the displacement y in the longitudinal direction of the track 2 needs to be calculated. The angle θ between the total offset vector and the longitudinal direction of the track 2.

[0038] Specifically, three target points are arranged in the longitudinal direction of track 2 (one of which is the positioning target point), and the 321 coordinate transformation method is used to establish a coordinate system on the rail waist. After the rail waist coordinate system is established, the component of the longitudinal displacement of track 2 can be calculated based on the total offset and the angle θ between the total offset vector and the longitudinal direction of track 2.

[0039] After the longitudinal displacement calculation is completed, according to the formula: △t = y÷ al T = △t + t0 Calculate the current track temperature, where al is the conversion coefficient, y is the longitudinal displacement of track 2, △t is the temperature change, T is the current track 2 temperature, and t0 is the locked track temperature.

[0040] In some embodiments, the target point is a reflective point. The purpose of using the reflective point is that the light source is coaxial with the camera, and after the light source irradiates the reflective point, more light is reflected into the camera, which can effectively increase the target brightness.

[0041] In some embodiments, the embodiments of the present invention further include a trigger controller, which is placed inside the vehicle compartment and controls the synchronous triggering of multiple sets of binocular cameras 4 and multiple light sources.

[0042] The present invention also provides an intelligent vehicle-mounted seamless line rail displacement detection method, comprising: Preparation steps: arranging target points on the foundation pile 3 and on the part of the rail waist corresponding to the foundation pile 3; installing the detection device on the locomotive; obtaining the spatial relative position relationship of each camera through global calibration in advance; Testing steps: S1, the detection device moves with the movement of the locomotive; S2, the vehicle host controls the binocular camera 4 to start taking pictures continuously when it reaches within 10 meters of the pile 3 according to the GPS and locomotive mileage information, and stops taking pictures when it is 10 meters away from the pile; S3, the vehicle-mounted host pre-processes the acquired images to remove noise and improve image quality, and numbers each image according to the shooting order, and calculates the center position of the target point using the Hough transform algorithm; S4. Perform three-dimensional reconstruction of the target points of pile 3 and rail waist target points, perform point cloud registration and alignment on the point set Q1 of the reconstructed image of the target points of pile 3 and the standard point set M1 used in the target points of pile 3, calculate the transformation matrices R1 and T1. Since the system has completed global calibration, the three groups of cameras are rigidly connected and do not change their positions. Therefore, the coordinates of the reconstructed rail waist target points are converted to the pile 3 coordinate system with the same transformation relationships R1 and T1, thereby realizing the unity of the coordinate systems of the rail waist target points and the pile 3 target points; in the pile coordinate system, the spatial position relationship of the rail waist target points relative to the pile target points can be calculated, that is, the pile 3 coordinate system is further converted to the rail waist target points through point-line-plane coordinate conversion, so as to calculate the displacement of track 2.

[0043] In some embodiments, the total offset Y of the rail waist target point relative to the base pile target point in three-dimensional space can be calculated in S4. However, when the locomotive passes through the track, it exerts force on the track, causing the track attitude to change, thereby affecting the measurement accuracy. A track attitude compensation algorithm is used, that is, the total offset Y and the angle between the total offset vector and the longitudinal direction of the track are used to compensate for the total offset Y. Calculate the displacement y in the longitudinal direction of the track; .

[0044] In some embodiments, when there is a need to calculate the locked rail temperature, after the longitudinal displacement calculation is completed, according to the formula: △t = y÷ al T = △t + t0 Calculate the current track temperature, where al is the conversion coefficient, y is the longitudinal displacement of the track, △t is the temperature change, T is the current track temperature, and t0 is the locked track temperature.

[0045] The present invention uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only used to help understand the method and core ideas of the present invention. At the same time, for those skilled in the art, according to the ideas of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. An intelligent vehicle-mounted seamless track rail displacement detection device, characterized in that: include: An image acquisition device is used to acquire image information of target points on the rail waist and target points on the foundation piles on one side of the track; The onboard host is installed in the locomotive compartment; The image acquisition device is in communication connection with the vehicle-mounted host and transmits image information to the vehicle-mounted host, and the vehicle-mounted host calculates the displacement of the track according to the image information; as well as The vehicle-mounted bracket is connected to the outside of the locomotive and is used to provide an installation position for the image acquisition device.

2. The intelligent vehicle-mounted seamless track rail displacement detection device according to claim 1 is characterized in that: The image acquisition device includes three groups of binocular cameras, which respectively acquire images of two rail waists of the double-strand parallel rails and target points on one side of the pile foundation. The vehicle-mounted host establishes a pile foundation coordinate system according to the target points reconstructed from the pile foundation image. The vehicle-mounted host establishes a local rail waist coordinate system according to the target points reconstructed from the rail waist image. The vehicle-mounted host performs coordinate transformation on the reconstructed rail waist target points and the pile foundation target points to obtain the displacement of the rail.

3. The intelligent vehicle-mounted seamless track rail displacement detection device according to claim 2 is characterized in that: The two cameras in each group of binocular cameras are connected by a carbon fiber cross bar, one group of binocular cameras is arranged on one side of the vehicle-mounted bracket, and the two carbon fiber cross bars in the other two groups of binocular cameras are connected by a solid column and then connected to the other side of the vehicle-mounted bracket.

4. The intelligent vehicle-mounted seamless track rail displacement detection device according to claim 3 is characterized in that: Each camera is equipped with a coaxial light source, and a filter is arranged inside the camera.

5. The intelligent vehicle-mounted seamless track rail displacement detection device according to claim 3 is characterized in that: Each of the cameras is equipped with a constant temperature regulating mechanism.

6. The intelligent vehicle-mounted seamless track rail displacement detection device according to claim 3 is characterized in that: The vehicle-mounted host calibrates the three groups of binocular cameras in the same coordinate system in advance, and calculates the displacement of the track according to the point cloud registration method and the coordinate conversion method.

7. The intelligent vehicle-mounted seamless track rail displacement detection device according to claim 4 is characterized in that: It also includes a trigger controller, which is placed inside the carriage and controls multiple groups of binocular cameras and multiple light sources to trigger their operation synchronously.

8. An intelligent vehicle-mounted seamless track rail displacement detection method, characterized in that: include: Preparation steps: Arrange target points on the piles and on the rail waist corresponding to the piles. The target points on the rail waist are connected to form a right angle, and one side of the right angle extends longitudinally along the rail waist. Install the detection device on the locomotive. Obtain the spatial relative position relationship of each camera through global calibration in advance. Testing steps: S1, the detection device moves with the movement of the locomotive; S2: The onboard host controls the binocular camera to start taking pictures continuously when it is within 10 meters of the pile according to the GPS and locomotive mileage information, and stops taking pictures when it is 10 meters away from the pile; S3, the vehicle host pre-processes the acquired images to remove noise and improve image quality, and numbers each image according to the shooting order, and calculates the center position of the target point using the Hough transform algorithm; S4. Perform three-dimensional reconstruction of the pile target point and the rail waist target point, perform point cloud registration and alignment on the point set Q1 reconstructed from the pile target point image and the standard point set M1 used in the pile target point, calculate the transformation matrices R1 and T1. Since the system has completed global calibration, the three camera groups are rigidly connected and do not change their positions. Therefore, the reconstructed rail waist target point coordinates are converted to the pile coordinate system with the same transformation relationships R1 and T1, thereby realizing the unity of the coordinate systems of the rail waist target point and the pile target point. In the pile coordinate system, the spatial position relationship of the rail waist target point relative to the pile target point can be calculated, that is, the pile coordinate system is further converted to the rail waist target point through point-line-plane coordinate conversion, so as to calculate the total offset Y of the track.

9. The intelligent vehicle-mounted seamless track rail displacement detection method according to claim 8, characterized in that: In S4, the total offset Y of the rail waist target point relative to the pile target point in three-dimensional space can be calculated. However, when the locomotive passes through the track, it exerts force on the track, causing the track attitude to change, which in turn affects the measurement accuracy. The track attitude compensation algorithm is used, that is, the total offset Y and the angle between the total offset vector and the longitudinal direction of the track are used to compensate for the total offset Y. Calculate the displacement y in the longitudinal direction of the track; 。 10. The intelligent vehicle-mounted seamless track rail displacement detection method according to claim 9, characterized in that: When there is a need to calculate the locked rail temperature, after the longitudinal displacement calculation is completed, according to the formula: △t = y÷ al T = △t + t0 Calculate the current track temperature, where al is the conversion coefficient, y is the longitudinal displacement of the track, △t is the temperature change, T is the current track temperature, and t0 is the locked track temperature.

Images