A portable modular self-correcting rail 3D detection system

Abstract

The present invention belongs to the technical field of railway safe transportation, and relates to a system for automatic correction of rails, in particular to a portable modular self-correcting three-dimensional rail detection system, which can realize rapid assembly and detection. Its main structure includes two detection units for detection, a mounting bracket for support and connection, and a hand operator, wherein the main structure of the detection unit includes a running mechanism, a scanning mechanism, a control unit, a level detection sensor, a power supply unit, and a temperature and humidity sensor. The present invention can simultaneously complete the detection of the geometric characteristics of the rails, including: rail height, track gauge, rail profile, etc., and can also complete the detection of the geometric position of the rail relative to the earth. When the above detection is completed, the present invention simultaneously records the geographical coordinates of the detection point for tracing back the detection records. Its main concept is ingenious, the structural design is scientific and reasonable, the use and measurement are convenient, fast, and accurate, the application environment is friendly, and the market prospects are broad.

Description

Technical field:

[0001] The present invention belongs to the technical field of railway safe transportation, and relates to a system and method for automatic rail correction, in particular to a portable modular self-correcting rail three-dimensional detection system that can achieve rapid assembly and detection. Background technology:

[0002] Railway transportation is one of my country's primary modes of transportation. The performance of rails directly impacts railway safety. Measuring rail geometry is a crucial basis for verifying rail performance parameters. Currently, the most commonly used inspection method on railways is manual measurement, which requires manual operation of various measuring instruments. Test results are directly affected by operator skill and are susceptible to human interference. Automated inspection equipment is relatively large, making it difficult to transport manually. The assembly and calibration process is complex, making maintenance challenging. Equipment that can be easily assembled on-site can be affected by assembly errors, making accuracy difficult to guarantee. Without solving the above-mentioned problem, Chinese patent application number CN209605771U discloses a track profile detector that is easy to assemble and disassemble. The positioning assembly is connected to the bottom of the inspection box. The operating lever and the display screen are detachably connected and fixed by a locking member. The hand-held lever is connected to the operating lever. The operating lever can rotate along a vertical plane relative to the conversion head and the two are detachably connected. The locking member fixes the conversion head and the joint seat. The joint seat and the folding seat are respectively fixed to the connecting rod by locking members. The connecting plate is detachable from the connecting rod and the inspection box. Although the profile detector can be disassembled by its own structure and is easy to assemble, it can be quickly installed on the rails during railway inspection, which can facilitate the disassembly of the device and improve the detection effect and efficiency of rail damage. However, it is unable to achieve simultaneous detection of multi-dimensional data such as rail height, track gauge, and rail profile. Chinese Patent Application No. CN201610813938.4 discloses a high-precision method for measuring rail geometry. This method utilizes a combined measurement system constructed with multiple lasers and multiple cameras. Through joint calibration of the measurement systems and comprehensive processing of the collected data, the characteristic points of the rail geometry in two non-overlapping dynamic three-dimensional coordinate systems are accurately acquired. This detection device offers simple and flexible installation, high detection accuracy, and the ability to simultaneously calibrate the world coordinates of both rails, effectively completing three-dimensional detection of the rail geometry. The invention boasts simple and flexible installation, high detection accuracy, and the ability to simultaneously calibrate the world coordinates of both rails. However, this invention carries a heavy workload and is prone to errors introduced by manual operation. Chinese Patent Application No. CN106114553A discloses a method for photoelectric dynamic measurement of platform sway in a railway inspection vehicle. By utilizing multiple precise photoelectric rail displacement measurement systems installed on the inspection vehicle's equipment platform, the method can determine the relative displacement at multiple points on the rail, enabling dynamic measurement of the relative position and posture information between the rail surface and the measurement platform. The rail displacement precision photoelectric measurement system consists of a line laser, a point laser, and a camera. It uses a precision photoelectric displacement measurement method that combines point laser displacement measurement technology with line laser displacement profile triangulation technology to measure the precise displacement of the laser point. Line laser triangulation is used to determine the rail cross-sectional profile, and the position of the laser point on the rail surface profile is determined based on the image relationship between the point laser and line laser. However, during installation, this invention is prone to gaps between the mounting components, which can introduce errors in assembly and affect measurement results. Furthermore, the test equipment is large and difficult to move, making installation and commissioning complex.Therefore, the present invention seeks to design and provide a portable modular self-correcting rail three-dimensional detection system that can effectively solve the above problems. Summary of the invention:

[0003] The purpose of the present invention is to overcome the defects mentioned above and seek to design a portable modular self-correcting three-dimensional rail detection system. The system can simultaneously complete the detection of rail geometric characteristics, including rail height, track gauge, rail profile, etc., and can also complete the detection of the geometric position of the rail relative to the earth. After completing the above detection, the present invention simultaneously records the geographical coordinates of the detection points for tracing back the detection records.

[0004] In order to achieve the above objectives, the portable modular self-correcting rail 3D detection system according to the present invention is implemented by the following technical solutions:

[0005] The main structure of the portable modular self-correcting rail three-dimensional detection system disclosed in the present invention includes two detection units for detection, a mounting bracket for support and connection, and a handheld operator. The main structure of the detection unit includes a running mechanism, a scanning mechanism, a control unit, a level detection sensor, a power supply unit, and a temperature and humidity sensor. The running mechanism is connected to the scanning mechanism to achieve simultaneous scanning while traveling; the level detection sensor is installed on the rear side of the scanning mechanism; the control unit communicates with the running mechanism, the scanning mechanism, and the level detection sensor to collect and process information from the above components; and the power supply unit supplies power to the running mechanism, the scanning mechanism, the control unit, and the level detection sensor.

[0006] The main structure of the running mechanism of the present invention includes a power wheel, a bracket, an auxiliary wheel, and an encoder; the power wheel is a servo wheel that can accurately control the running distance, and the encoder is arranged inside the power wheel, wherein the auxiliary wheel is arranged horizontally in front of the power wheel, and the bracket is a double-frame side-mounted plate structure, wherein the power wheel and the auxiliary wheel are located in the middle and lower part of the bracket and connected to its double support plates; the power supply and servo controller are both arranged in the middle of the frame side-mounted plate, wherein the servo controller is connected to the power wheel for control;

[0007] The main structure of the scanning mechanism includes a slide and laser 2D / 3D sensors. There are three groups of laser 2D / 3D sensors, which are installed in the upper front part between the two double-frame side panels and located above the slide. The slide can drive the laser 2D / 3D sensors to move longitudinally to adjust their angles.

[0008] The main structure of the control unit includes an industrial computer, a GPS positioning module, a 4G module, and a wireless communication module to form an integrated package, and can communicate with the computer through the GPS positioning module, the 4G module, and the wireless communication module;

[0009] The main structure of the mounting bracket includes a load-bearing frame, a locking mechanism, and a reference rod. The plug-in load-bearing frame has a triangular layout, with both ends plugged into the frame side mounting plates. The reference rod is located on the lower side of the load-bearing frame and is connected to the unit bracket through a locking mechanism. The main structure of the locking mechanism includes a magnetic concave plate, a top spring, and a base. The magnetic concave plate is connected to the base through the top spring to form an integral locking mechanism. The base of the locking mechanism is connected to the unit bracket, and the base is connected to the component side plate. The magnetic concave plate and top spring structure enable rapid installation while ensuring accurate and stable installation.

[0010] The handheld controller is a commercially available tablet computer with a related program installed in the computer;

[0011] Furthermore, the main structure of the bracket in the present invention includes two vertically arranged frame side mounting plates, a frame stabilizing link for connecting the two frame side mounting plates, and a bracket assembly component. The frame stabilizing link and the bracket assembly component are both located between the frame side mounting plates on both sides and connected thereto. There are three mounting bracket assembly components, which are arranged in a triangular manner to ensure the overall installation accuracy and stability of the system and achieve rapid assembly and disassembly. The bracket assembly components cooperate with the mounting bracket to reduce the error introduced by the system equipment.

[0012] Furthermore, in the present invention, the control unit and the power supply unit are arranged inside the detection unit and close to the upper part, wherein the control unit is located in front of the power supply unit, wherein the temperature and humidity sensor is located in front of the control unit, so as not to be blocked or interfered with by other components during measurement;

[0013] When using the portable modular self-correcting rail three-dimensional detection system according to the present invention, the specific operation method is as follows:

[0014] Step 1: Perform self-calibration

[0015] S1, the slide, the mobile laser 2D / 3D sensor, and the level detection sensor are rigidly connected. The mobile laser 2D / 3D sensor is calibrated. The resulting profile can be compared with the reference profile. If they match, the mobile laser 2D / 3D sensor is fault-free.

[0016] S2. The calibration components are rigidly connected to the reference rods, so the relative spatial position relationship between the calibration components is fixed. The detection unit can obtain the spatial coordinates of the reference calibration component contour through detection, calculate its own relative spatial posture position relative to the calibration component, and use this as a reference for detecting the contour spatial coordinates of the rail, thereby eliminating and correcting the installation error introduced by the quick installation clearance.

[0017] S3. The detection results of the level detection sensor are mutually verified according to the above correction calibration, and the system as a whole is corrected according to the level detection sensor reading;

[0018] Step 2. Specific calculation method:

[0019] There are reference calibration components on the left and right sides of the reference rod. The calibration components are triangular prisms. The straight lines where the convex edges of the prisms are located are coplanar. The straight lines where the convex edges of the prisms on the left and right sides are parallel, and are parallel to the straight lines where the convex edges of the prisms on the right side are located on the left and right sides; the distance between line 2 and line 3 is D, and the distances between the two ends of the calibrated parallel lines 1 and line 4 and between the two ends of the calibrated parallel lines 5 and line 8 are equal, both being D1;

[0020] S1. The left 3D image scanning coordinates form a matrix array

[0021]

[0022] S2. The horizontal coordinate sensor collects the inclination angle α1 around the X-axis and the inclination angle β1 around the Y-axis, and rotates the coordinate system according to α1 and β1 to obtain

[0023]

[0024] S3. Divide A2[n] into the following three regional coordinate matrix arrays according to the x coordinate

[0025] The value of A21[i] is A2[n] (P2 ≤ x1[n])

[0026] The value of A22[j] is A2[n] (x1[n] ≤ P1)

[0027] The value of A23[k] is A2[n] (P1 < x1[n] < P2)

[0028] S4. In A21[i], step by Δy, where Δy is twice the resolution in the Y-axis direction, search for the point with the maximum value of z2[i] within nΔy < y1[i] ≤ Δy(n + 1), and obtain the high points in the plane to form a new matrix array The points passed by the straight line 1 where the convex edge of the prism of this matrix array is located;

[0029] S5. Find the projection line of the matrix array B1[r] on the z = 0 plane

[0030]

[0031]

[0032] The obtained projection line equation on the z = 0 plane is y = b1x + a1, which is the projection equation of the straight line 1;

[0033] S6. Obtain the projection line y = b11z + a11 of B1[r] on the x = 0 plane according to the method in step S5;

[0034] S7, according to the method of step S4 and step S6, obtain the high point matrix array in the A22[j] plane The matrix array points are the points where the line 4 on which the prism convex side is located passes; the projection line y = b2x + a2 of the matrix array in the z = 0 plane and the projection line y = b22z + a22 in the x = 0 plane are obtained, that is, the projection equation of line 4;

[0035] S8. According to the method of step S4 and step S6, obtain the high point matrix array in the A23[k] plane The matrix array points are the points where the line 3 on which the prism convex side is located passes; the projection line y=b3x+a3 of the matrix array on the z=0 plane and the projection line y=b33z+a33 on the x=0 plane are obtained, that is, the projection equation of line 3;

[0036] S9. If y=b1x+a1 and y=b2x+a2 are not parallel and do not match the reference profile feature attributes, the system will fail the self-test. If they are parallel, calculate the rotation angle γ1 around the Z axis.

[0037] γ1=arccot(b1);

[0038] S10. Take the straight lines with the same distances as y=b1x+a1 and y=b2x+a2 That is, the projection equation of line 2 on the z=0 plane;

[0039] S11. Take the straight lines with the same distances as y=b11x+a11 and y=b22x+a22. That is, the projection equation of line 2 on the x=0 plane;

[0040] S12. Calculation The solution for the intersection with y=b3x+a3 is

[0041]

[0042] S13, substitute the value of X1 in step S10 into y=b33z+a33 to deduce

[0043]

[0044] S14, based on the above, the left coordinate system correction parameters are rotated around XYZ by α1, β1, γ1, and the translation is X1, Y1, Z1.

[0045] S15. According to the method of steps S1 to S14, the matrix array of the scanned coordinates of the right 3D image can be processed to derive the equations of the three projected lines on the z=0 plane: the projected line y=c1x+d1 on the z=0 plane and the projected line y=c11z+d11 on the x=0 plane, which are the projection equations of line 5.

[0046] Project the straight line y = c2x + d2 on the z = 0 plane, and project the straight line y = c22z + d22 on the x = 0 plane, which are the projection equations of line 8; project the straight line y = c3x + d3 on the z = 0 plane, and project the straight line y = c33z + d33 on the x = 0 plane, which are the projection equations of line 7; and pass the parallelism self-check of y = b11x + a11 and y = b22x + a22. If they are not parallel, the self-check of this unit cannot be passed;

[0047] Through self-checking, we can deduce that the projection line on the z=0 plane is Projecting a straight line onto the x=0 plane The projection equation of line 6 can be deduced as the coordinate system correction parameters on the right: the rotation angles around XYZ are α11, β11, γ11, and the translation amounts are X11, Y11, Z11;

[0048] S16. Find the distance between y=b1x+a1 and y=b2x+a2

[0049] Find the distance between y=c1x+d1 and y=c2x+d2

[0050] If d1 and e1 are not equal, it means that the standard rod is deformed or the scanning unit mechanical structure is deformed and deflected, and the system as a whole cannot pass the self-test. If they are equal, the self-test is passed. It is inferred that the center of the left 3D image coordinate system and the center of the right 3D image coordinate system are parallel to the X offset.

[0051]

[0052] S17. Find the difference between y=b11x+a11 and y=b22x+a22

[0053] Find the difference between y=c11x+d11 and y=c22x+d22

[0054] If d11 and e11 are not equal, it means that the standard rod is deformed or the scanning unit mechanical structure is deformed and deflected, and the system as a whole cannot pass the self-test. If they are equal, the self-test is passed. It is deduced that the center of the left 3D image coordinate system and the center of the right 3D image coordinate system are parallel to the Z offset.

[0055]

[0056] In summary, the system self-check and mutual verification are completed, and the self-calibration parameters of the left 3D image XYZ rotation angle are α1, β1, γ1, and the translation amount is X1, Y1, Z1. The self-calibration parameters of the right 3D image XYZ rotation angle are α11, β11, γ11, and the translation amount is X11, Y11, Z11. The translation amount of the right 3D image relative to the left 3D image is X12 and Z12. The above parameters are used to correct the installation error introduced by the quick installation clearance.

[0057] Step 3: Rail profile detection:

[0058] After the self-calibration is completed, the scanning mechanism scans the 3D profile of the rail and registers the test results of each unit based on the calibration results to form the overall detection profile;

[0059] S1. Track gauge: The track gauge is measured vertically from the edge of one rail to the edge of the other rail. Multi-point detection avoids errors caused by manual measurement without the measuring tool being perpendicular to the rail. Multi-point detection increases accuracy and can simultaneously verify the parallelism of the two rail sections.

[0060] S2, rail height: The rail height is calculated by taking the multi-point track gauge in S1 and the edge of the rail profile as the reference line and extending vertically to the other side of the rail edge and the horizontal angle between the ground;

[0061] S3, Track extension angle: The angle between the track extension direction and the ground level;

[0062] S4. Rail profile: After coordinate correction, the rail profile can be obtained by taking the plane perpendicular to the rail as the section. At the same time, the horizontal position of the rail relative to the ground can be obtained, providing horizontal information for further analysis of rail wear;

[0063] S5. Others: 3D spatial information of the two rail inspection sections and their relationship with the ground level, which can be further extracted as needed;

[0064] Step 4: Take measurements:

[0065] S1. On-site rapid assembly and testing of the portable modular self-correcting rail 3D detection system of the present invention. The handheld operator 3 wirelessly connects the two detection units to complete the overall initialization of the device.

[0066] S2, the left and right detection units respectively control the slide and move the front section of the laser 2D / 3D sensor to scan the calibration rod marker to obtain a 3D image;

[0067] S3. Compare the obtained data with the factory-calibrated reference 3D images on the left and right sides to determine whether it is within the error range. By verifying the data of the three laser 2D / 3D sensors, it can be inferred which sensor may be faulty. If a fault occurs, the alarm is exited;

[0068] S4. If there is no fault detection, the readings of the level detection sensor are further read and the three-dimensional space coordinate system of the unit is corrected respectively;

[0069] S5. Compare the data of the left and right detection units with the calibration rod data during factory calibration to determine whether the calibration rod is deformed. If deformation occurs, a fault alarm is issued and the system is exited.

[0070] S6. Correct the overall coordinate system based on the calibration rod, establish a coordinate system based on the geoid using the factory calibration information of the level detection sensor, and start scanning the entire 3D image of the railway track;

[0071] S7, temperature and humidity sensor detects current temperature and humidity;

[0072] S8. Identify characteristic information including the rail, rail top surface, and rail edge, and further calculate characteristic information including the track gauge, rail height, track extension angle, and track profile based on temperature and humidity corrections;

[0073] S9, the detection unit obtains GPS positioning information and the running information of the running encoder, and the control unit uploads the above information through the 4G module;

[0074] S10: Automatically move to the next detection point for detection.

[0075] Compared with the prior art, the present invention has the following beneficial effects:

[0076] 1. The present invention can simultaneously complete the detection of rail geometric characteristics, including rail height, track gauge, track profile, etc., and can also complete the detection of the rail's geometric position relative to the earth. After completing the above detection, the present invention simultaneously records the geographical coordinates of the detection point for tracing back the detection record.

[0077] 2. The modular design makes on-site assembly simple and fast. After assembly, it has an automatic calibration function to eliminate assembly errors, overcoming the detection errors introduced by similar portable products due to manual assembly or loose assembly.

[0078] 3. The detection results can form a three-dimensional image, which, combined with automatic correction and calibration, greatly improves the overall measurement accuracy and precision of the equipment. Compared with existing detection methods, the present invention can more comprehensively display the geometric size relationship of the rails within the detection range, the relative position relationship between the rails, and the relative position relationship between the rails and the earth through three-dimensional images, eliminating errors introduced by manual operation of the equipment.

[0079] 4. It can automatically move and detect on the rails and record the moving distance.

[0080] 5. The modular structure of the present invention is easy to carry and assemble. It has a reference rod. Based on the 3D profile imaging of the reference rod and the inclination sensor, the system can be self-calibrated to eliminate the introduction of errors.

[0081] 6. It has a complete self-calibration method, which can calibrate the results obtained by testing the reference components of each unit with the standard results to achieve self-inspection, and perform overall calibration by matching the reference components.

[0082] 7. The present invention relies on 3D detection to detect the relationship between various characteristics of the rails, which can effectively eliminate the errors introduced by manual operation that are difficult to avoid with traditional methods.

[0083] 8. The present invention can provide more specific and complete spatial 3D geometric information of the rails, and can more comprehensively understand the spatial relationship of the rails, including: spatial horizontal and vertical relationships, transverse and longitudinal inclination relationships of the rails, and parallelism relationships of the rails, which correspond to rail wear. It can not only detect the degree of rail wear, but also provide basic data for wear factor analysis.

[0084] In summary, its main body is ingeniously conceived, the structural design is scientific and reasonable, the use and measurement are convenient, fast and accurate, the application environment is friendly, and the market prospects are broad. Description of the drawings:

[0085] Figure 1 This is a schematic diagram of the relative position relationship between the present invention and the rails.

[0086] Figure 2 It is a schematic diagram of the main structural principle of the present invention.

[0087] Figure 3 This is a schematic diagram of the main structural principle of the detection unit involved in the present invention.

[0088] Figure 4 This is a schematic diagram of the positional relationship of the laser 2D / 3D sensor involved in the present invention.

[0089] Figure 5 This is a schematic diagram of the main structural principle of the running mechanism involved in the present invention.

[0090] Figure 6 This is a schematic diagram of the main structural principle of the locking mechanism involved in the present invention.

[0091] Figure 7 This is a schematic diagram of the internal structure principle of the detection unit involved in the present invention.

[0092] Figure 8 The present invention provides a schematic diagram of the workflow of the portable modular self-correcting rail three-dimensional detection system.

[0093] Figure 9 This is a schematic diagram of the self-calibration principle involved in the present invention. Specific implementation method:

[0094] In order to clearly illustrate the technical features of this solution, the present invention is further described below in conjunction with embodiments.

[0095] Example 1

[0096] The main structure of the portable modular self-correcting railroad three-dimensional detection system involved in this embodiment includes two detection units 1 for detection, a mounting bracket 2 for supporting and connecting, and a handheld controller 3. The main structure of the detection unit 1 includes a running mechanism 1.1, a scanning mechanism 1.2, a control unit 1.3, a level detection sensor 1.4, a power supply unit 1.5, and a temperature and humidity sensor 1.6. The running mechanism 1.1 is connected to the scanning mechanism 1.2 to achieve simultaneous scanning while traveling; the level detection sensor 1.4 is installed on the rear side of the scanning mechanism 1.2; the control unit 1.3 communicates with the running mechanism 1.1, the scanning mechanism 1.2, and the level detection sensor 1.4 to collect and process information from the above components; and the power supply unit 1.5 supplies power to the running mechanism 1.1, the scanning mechanism 1.2, the control unit 1.3, and the level detection sensor 1.4.

[0097] The main structure of the running mechanism 1.1 in this embodiment includes a power wheel 1.1.1, a bracket 1.1.2, an auxiliary wheel 1.1.3, and an encoder 1.1.4; the power wheel 1.1.1 is a servo wheel that can accurately control the running distance, the encoder 1.1.4 is arranged inside the power wheel 1.1.1, the auxiliary wheel 1.1.3 is arranged horizontally in front of the power wheel 1.1.1, the bracket 1.1.2 is a double-frame side mounting plate 1.1.2.1 structure, the power wheel 1.1.1 and the auxiliary wheel 1.1.3 are located in the middle and lower part of the bracket 1.1.2 and are connected to its double support plates; the power supply 1.7 and the servo controller 1.8 are both arranged in the middle of the frame side mounting plate 1.1.2.1, and the servo controller 1.8 is connected to the power wheel 1.1.1 for control;

[0098] The main structure of the scanning mechanism 1.2 includes a slide 1.2.1 and laser 2D / 3D sensors 1.2.2. Three sets of laser 2D / 3D sensors 1.2.2 are located at the upper front between the two double-frame side panels 1.1.2.1 and above the slide 1.2.1. The slide 1.2.1 can drive the laser 2D / 3D sensors 1.2.2 to move longitudinally to adjust their angle.

[0099] The main structure of the control unit 1.3 includes an industrial computer, a GPS positioning module, a 4G module, and a wireless communication module to form an integrated package, and can communicate with the computer through the GPS positioning module, the 4G module, and the wireless communication module;

[0100] The main structure of the mounting bracket 2 includes a carrying frame 2.1, a locking mechanism 2.2, and a reference rod 2.3, wherein the plug-in carrying frame 2.1 is a triangular arrangement structure, and both ends are plug-in connected to the frame side mounting plates 1.1.2.1 respectively. The reference rod 2.3 is located on the lower side of the carrying frame 2.1 and is connected to the unit bracket 1.1.2 through the locking mechanism 2.2, wherein the main structure of the locking mechanism 2.2 includes a magnetic concave plate 2.2.1, a top spring 2.2.2, and a base 2.2.3. The magnetic concave plate 2.2.1 is connected to the base 2.2.3 through the top spring 2.2.2 to form an overall locking mechanism, the base 2.2.3 of the locking mechanism is connected to the unit bracket 1.1.2, and the base 2.2.3 is connected to the side plate of the 1.1.2.1 component; wherein the magnetic concave plate 2.2.1 and the top spring 2.2.2 structure can achieve rapid installation while ensuring accurate and stable installation;

[0101] The handheld controller is a commercially available tablet computer with a related program installed in the computer;

[0102] Furthermore, the main structure of the bracket 1.1.2 in this embodiment includes two vertically arranged frame side mounting plates 1.1.2.1, a frame stabilizing link 1.1.2.2 for connecting the two frame side mounting plates 1.1.2.1, and a bracket assembly component 1.1.2.3. The frame stabilizing link 1.1.2.2 and the bracket assembly component 1.1.2.3 are both located between and connected to the frame side mounting plates 1.1.2.1 on both sides. There are three mounting bracket assembly components 1.1.2.3, which are arranged in a triangular manner to ensure the overall installation accuracy and stability of the system and achieve rapid assembly and disassembly; the bracket assembly components 1.1.2.3 cooperate with the mounting bracket 2 to reduce the error introduced by the system equipment;

[0103] Furthermore, in this embodiment, the control unit 1.3 and the power supply unit 1.5 are arranged inside the detection unit 1 and close to the upper part, wherein the control unit 1.3 is located in front of the power supply unit 1.5, and the temperature and humidity sensor 1.6 is located in front of the control unit 1.3, so as to avoid being blocked or interfered with by other components during measurement;

[0104] When using the portable modular self-correcting rail 3D detection system involved in this embodiment, the specific operation method is as follows:

[0105] Step 1: Perform self-calibration

[0106] S1, slide 1.2.1, mobile laser 2D / 3D sensor 1.2.2, and level detection sensor 1.4 are rigidly connected. The mobile laser 2D / 3D sensor 1.2.2 calibrates the components, and the resulting profile can be compared with the reference profile. If they match, the mobile laser 2D / 3D sensor 1.2.2 is not faulty.

[0107] S2. The calibration components are rigidly connected to the reference rod, so the relative spatial position relationship between the calibration components is fixed. The detection unit 1 can obtain the spatial coordinates of the contour of the reference calibration components through detection, calculate its relative spatial attitude position relative to the calibration components, and use this as a reference object for detecting the spatial coordinates of the rail contour, thereby eliminating and correcting the installation error introduced by the quick installation fit clearance;

[0108] S3. The detection results of the horizontal detection sensor 1.4 are mutually verified according to the above-mentioned corrected calibration, and the system as a whole is corrected according to the reading of the horizontal detection sensor 1.4;

[0109] Step 2. Specific calculation method:

[0110] As Figure 9 shown, there are reference calibration components on both the left and right sides of the reference rod. The calibration components are triangular prisms, and the straight lines where the convex edges of the prisms are located are coplanar. The straight lines where the convex edges of the prisms on the left and right sides are parallel, and are parallel to the straight lines where the convex edges of the prisms on the right side are located; the distance between line 2 and line 3 is D, and the distances between the two ends of the calibration parallel lines 1 and line 4 and between the two ends of the calibration parallel lines 5 and line 8 are equal, both being D1;

[0111] S1. The left 3D image scanning coordinates form a matrix array

[0112]

[0113] S2. The horizontal coordinate sensor collects the inclination angle α1 around the X-axis and the inclination angle β1 around the Y-axis, and obtains the coordinate system rotated according to α1 and β1

[0114]

[0115] S3. Divide A2[n] into the following three regional coordinate matrix arrays according to the x coordinate

[0116] A21[i] takes the value of A2[n] (P2 ≤ x1[n])

[0117] A22[j] takes the value of A2[n] (x1[n] ≤ P1)

[0118] A23[k] takes the value of A2[n] (P1 < x1[n] < P2)

[0119] S4. In A21[i], step according to Δy, where Δy is twice the resolution in the Y-axis direction, and search for the point with the maximum value of z21[i] within nΔy < y1[i] ≤ Δy(n + 1) to obtain the high points in the plane to form a new matrix array The points in this matrix array are the points passed by the straight line 1 where the convex edge of the prism is located;

[0120] S5. Find the projection straight line of the matrix array B1[r] on the z = 0 plane

[0121]

[0122]

[0123] The equation of the projected line on the z=0 plane is y=b1x+a1, which is the projection equation of line 1;

[0124] S6. Obtain the projection line y=b11z+a11 of B1[r] on the x=0 plane according to the method in step S5;

[0125] S7, according to the method of step S4 and step S6, obtain the high point matrix array in the A22[j] plane The matrix array points are the points where the line 4 on which the prism convex side is located passes; the projection line y = b2x + a2 of the matrix array in the z = 0 plane and the projection line y = b22z + a22 in the x = 0 plane are obtained, that is, the projection equation of line 4;

[0126] S8. According to the method of step S4 and step S6, obtain the high point matrix array in the A23[k] plane The matrix array points are the points where the line 3 on which the prism convex side is located passes; the projection line y=b3x+a3 of the matrix array on the z=0 plane and the projection line y=b33z+a33 on the x=0 plane are obtained, that is, the projection equation of line 3;

[0127] S9. If y=b1x+a1 and y=b2x+a2 are not parallel and do not match the reference profile feature attributes, the system will fail the self-test. If they are parallel, calculate the rotation angle γ1 around the Z axis.

[0128] γ1=arccot(b1);

[0129] S10. Take the straight lines with the same distances as y=b1x+a1 and y=b2x+a2 That is, the projection equation of line 2 on the z=0 plane;

[0130] S11. Take the straight lines with the same distances as y=b11x+a11 and y=b22x+a22. That is, the projection equation of line 2 on the x=0 plane;

[0131] S12. Calculation The solution for the intersection with y=b3x+a3 is

[0132]

[0133]

[0134] S13. Substitute the value of X1 in step S10 into y=b33z+a33 to derive the following formula:

[0135]

[0136] S14, based on the above, the left coordinate system correction parameters are rotated around XYZ by α1, β1, γ1, and the translation is X1, Y1, Z1.

[0137] S15. According to the method of steps S1 to S14, the matrix array of the scanned coordinates of the right 3D image can be processed to derive the equations of the three projected lines on the z=0 plane: the projected line y=c1x+d1 on the z=0 plane and the projected line y=c11z+d11 on the x=0 plane, which are the projection equations of line 5.

[0138] Project the straight line y = c2x + d2 on the z = 0 plane, and project the straight line y = c22z + d22 on the x = 0 plane, which are the projection equations of line 8; project the straight line y = c3x + d3 on the z = 0 plane, and project the straight line y = c33z + d33 on the x = 0 plane, which are the projection equations of line 7; and pass the parallelism self-check of y = b11x + a11 and y = b22x + a22. If they are not parallel, the self-check of this unit cannot be passed;

[0139] Through self-checking, we can deduce that the projection line on the z=0 plane is Projecting a straight line onto the x=0 plane The projection equation of line 6 can be deduced as the coordinate system correction parameters on the right: the rotation angles around XYZ are α11, β11, γ11, and the translation amounts are X11, Y11, Z11;

[0140] S16. Find the distance between y=b1x+a1 and y=b2x+a2

[0141] Find the distance between y=c1x+d1 and y=c2x+d2

[0142] If d1 and e1 are not equal, it means that the standard rod is deformed or the scanning unit mechanical structure is deformed and deflected, and the system as a whole cannot pass the self-test. If they are equal, the self-test is passed. It is inferred that the center of the left 3D image coordinate system and the center of the right 3D image coordinate system are parallel to the X offset.

[0143]

[0144] S17. Find the difference between y=b11x+a11 and y=b22x+a22

[0145] Find the difference between y=c11x+d11 and y=c22x+d22

[0146] If d11 and e11 are not equal, it means that the standard rod is deformed or the scanning unit mechanical structure is deformed and deflected, and the system as a whole cannot pass the self-test. If they are equal, the self-test is passed. It is deduced that the center of the left 3D image coordinate system and the center of the right 3D image coordinate system are parallel to the Z offset.

[0147]

[0148] In summary, the system self-check and mutual verification are completed, and the self-calibration parameters of the left 3D image XYZ rotation angle are α1, β1, γ1, and the translation amount is X1, Y1, Z1. The self-calibration parameters of the right 3D image XYZ rotation angle are α11, β11, γ11, and the translation amount is X11, Y11, Z11. The translation amount of the right 3D image relative to the left 3D image is X12 and Z12. The above parameters are used to correct the installation error introduced by the quick installation clearance.

[0149] Step 3: Rail profile detection:

[0150] After the self-calibration is completed, the scanning mechanism 1.2 scans the 3D profile of the rail and registers the test results of each unit based on the calibration results to form an overall detection profile;

[0151] S1. Track gauge: The track gauge is measured vertically from the edge of one rail to the edge of the other rail. Multi-point detection avoids errors caused by manual measurement without the measuring tool being perpendicular to the rail. Multi-point detection increases accuracy and can simultaneously verify the parallelism of the two rail sections.

[0152] S2, rail height: The rail height is calculated by taking the multi-point track gauge in S1 and the edge of the rail profile as the reference line and extending vertically to the other side of the rail edge and the horizontal angle between the ground;

[0153] S3, Track extension angle: The angle between the track extension direction and the ground level;

[0154] S4. Rail profile: After coordinate correction, the rail profile can be obtained by taking the plane perpendicular to the rail as the section. At the same time, the horizontal position of the rail relative to the ground can be obtained, providing horizontal information for further analysis of rail wear;

[0155] S5. Others: 3D spatial information of the two rail inspection sections and their relationship with the ground level, which can be further extracted as needed;

[0156] Step 4: Take measurements:

[0157] S1. On-site rapid assembly and testing of the portable modular self-correcting rail 3D detection system involved in this embodiment. The handheld controller 3 is wirelessly connected to the two detection units 1 to complete the overall initialization of the device.

[0158] S2, the left and right unit detection units 1 respectively control the slide 1.2.1 and the front section of the mobile laser 2D / 3D sensor 1.2.2 to scan the calibration rod marker to obtain a 3D image;

[0159] S3. Compare the obtained data with the factory-calibrated reference 3D images on the left and right sides to determine whether it is within the error range. By verifying the data from the three laser 2D / 3D sensors in 1.2.2, it can be inferred which sensor may be faulty. If a fault occurs, the alarm is exited.

[0160] S4, if there is no fault detection, the reading of the level detection sensor 1.4 is further read, and the three-dimensional space coordinate system of the unit is corrected respectively;

[0161] S5. Compare the data of the left and right detection units 1 with the calibration rod data during factory calibration to determine whether the calibration rod is deformed. If so, a fault alarm is issued and the system is exited.

[0162] S6. Correct the overall coordinate system based on the calibration rod, establish a coordinate system based on the geoid using the factory calibration information of the level detection sensor 1.4, and start scanning the entire 3D image of the railway track;

[0163] S7, temperature and humidity sensor 1.6 detects the current temperature and humidity;

[0164] S8. Identify characteristic information including the rail, rail top surface, and rail edge, and further calculate characteristic information including the track gauge, rail height, track extension angle, and track profile based on temperature and humidity corrections;

[0165] S9, the detection unit 1 obtains GPS positioning information and the running encoder running information, and the control unit 1.3 uploads the above information through the 4G module;

[0166] S10: Automatically move to the next detection point for detection.

Claims

1. A portable modular self-correcting rail 3D detection system, characterized by The invention is realized by the following technical solution: the main structure includes two detection units for detection, a mounting bracket for support and connection, and a handheld operator, wherein the main structure of the detection unit includes a running mechanism, a scanning mechanism, a control unit, a level detection sensor, a power supply unit, and a temperature and humidity sensor. The running mechanism is connected to the scanning mechanism to realize scanning while traveling; the level detection sensor is installed in the scanning mechanism; the control unit communicates with the running mechanism, the scanning mechanism, and the level detection sensor to collect and process information from the above components; the power supply unit supplies power to the running mechanism, the scanning mechanism, the control unit, and the level detection sensor; The main structure of the running mechanism includes a power wheel, a bracket, an auxiliary wheel, and an encoder. The power wheel is a servo wheel that can accurately control the running distance. The encoder is set inside the power wheel, and the auxiliary wheel is set horizontally in front of the power wheel. The bracket has a double-frame side-mounted plate structure. The power wheel and auxiliary wheel are located in the middle and lower part of the bracket and connected to its double support plates. The power supply and servo controller are both set in the middle of the frame side mounting plate, and the servo controller is connected to the power wheel for control. The main structure of the scanning mechanism includes a slide and laser 2D / 3D sensors. There are three groups of laser 2D / 3D sensors, which are installed in the upper front part between the two double-frame side panels and located above the slide. The slide can drive the laser 2D / 3D sensors to move longitudinally to adjust their angles. The main structure of the control unit includes an industrial computer, a GPS positioning module, a 4G module, and a wireless communication module to form an integrated package, and can communicate with the computer through the GPS positioning module, the 4G module, and the wireless communication module; The main structure of the mounting bracket includes a load-bearing frame, a locking mechanism, and a reference rod. The plug-in load-bearing frame is a triangular arrangement structure, and the two ends are respectively plug-in connected to the side mounting plates of the frame. The reference rod is located on the lower side of the load-bearing frame and is connected to the unit bracket through a locking mechanism. The main structure of the locking mechanism includes a magnetic concave plate, a top spring, and a base. The magnetic concave plate is connected to the base through the top spring to form an overall locking mechanism. The base of the locking mechanism is connected to the unit bracket, and the base is connected to the side plate of the component. The magnetic concave plate and top spring structure can achieve rapid installation while ensuring accurate and stable installation.

2. A portable modular self-correcting rail 3D detection system according to claim 1, characterized in that The handheld controller is a commercially available tablet computer with a related program installed in the computer.

3. The portable modular self-correcting rail 3D detection system according to claim 1 is characterized in that The main structure of the bracket includes two vertically arranged frame side mounting plates, a frame stabilizing connecting rod for connecting the two frame side mounting plates, and a bracket assembly component. The frame stabilizing connecting rod and the bracket assembly component are both located between the frame side mounting plates on both sides and connected to them. There are three mounting bracket assembly components, which are arranged in a triangular manner to ensure the overall installation accuracy and stability of the system and achieve rapid disassembly and assembly; the bracket assembly components cooperate with the mounting bracket to reduce the error introduced by the system equipment.

4. The portable modular self-correcting rail 3D detection system according to claim 1 is characterized in that The control unit and the power supply unit are arranged inside the detection unit and close to the upper part, wherein the control unit is located in front of the power supply unit, and the temperature and humidity sensor is located in front of the control unit so as not to be blocked or interfered by other components during measurement.

Images