Intelligent rail weld grinding method

Abstract

The present application relates to a kind of steel rail weld intelligent polishing method, comprising the following steps: S1: conveying steel rail: the steel rail of completion welding is conveyed to polishing station and makes weld move to polishing range, using steel rail positioning device locks steel rail;S2: set steel rail type;S3: the robot of steel rail both sides uses 3D camera to the weld of polishing area is visually identified, intelligently identifies steel rail type, and compared with set steel rail type, if not consistent, then give prompt information, stop polishing;If consistent, then identify weld position, steel rail offset angle and weld thickness, calculate polishing speed and polishing path;S4: the robot of steel rail both sides drives respective connection polishing device according to polishing path and polishing speed and carries out polishing to steel rail weld;S5: after polishing, robot returns to initial position, prepares next polishing operation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of rail grinding, in particular to a rail weld intelligent grinding method. BACKGROUND

[0002] After the rail is welded, rough grinding is needed to polish the protruding weld beads and weld ribs so that the external size of the weld meets the requirements of subsequent process steps. The rough grinding process needs to polish the weld beads and weld ribs of the rail head, rail waist and rail bottom and meet the requirements of the operation guide book. Due to the special shape of the rail cross section, the grinding process requirements are very high. Simple profile automatic grinding equipment cannot identify and judge the grinding position, and cannot accurately control the grinding position and grinding thickness. Over-grinding or under-grinding occurs from time to time, and the grinding quality cannot meet the requirements of the rail welding base production standard. The traditional grinding wheel grinding method needs to change the tool multiple times to meet the grinding standard due to the limitation of the tool structure, and the grinding wheel grinding cannot cover the complex curved surface of the rail jaw, resulting in a blind area of operation. The control precision of the overlapping area between the mechanical arms is insufficient, and over-grinding or under-grinding is prone to occur. Therefore, the use of abrasive belt grinding has gradually attracted attention. However, how to plan a reasonable grinding path for abrasive belt grinding, especially the use of double abrasive belts to achieve high-precision grinding, is a problem that needs to be solved urgently. SUMMARY

[0003] The present application provides a rail weld intelligent grinding method, comprising the following steps:

[0004] S1: conveying the rail: conveying the welded rail to the grinding station and moving the weld into the grinding range, and locking the rail with a rail positioning device;

[0005] S2: setting the rail type;

[0006] S3: the robots on both sides of the rail use a 3D camera to visually identify the weld in the grinding area, intelligently identify the rail type, and compare it with the set rail type. If they are inconsistent, a prompt message is given and the grinding is stopped. If they are consistent, the weld position, rail offset angle and weld thickness are identified, and the grinding speed and grinding path are calculated;

[0007] S4: the robots on both sides of the rail drive the respective connected grinding devices to grind the rail weld according to the grinding path and grinding speed;

[0008] S5: after the grinding is completed, the robot returns to the initial position and prepares for the next grinding operation;

[0009] In the above step S3, the method of intelligently identifying the rail type includes: establishing a rail database model for various rail models, extracting point clouds from the pictures taken by the 3D camera, and matching them with the rail database model to intelligently identify the rail type.

[0010] Further, in the step S3, the planning method of the grinding path is as follows:

[0011] The outer surface of the rail is divided into several regions: the rail bottom region AB, the left rail foot bottom region AJ, the right rail foot bottom region BC, the left rail foot top surface and triangular region JI, the right rail foot top surface and triangular region CD, the left rail waist and rail jaw region IH, the right rail waist and rail jaw region DE, the left rail top side region GH, the right rail top side region EF, and the rail top region FG;

[0012] The left robot grinds the rail top region FG and the left rail top side region GH, and the right robot stands by;

[0013] The right robot grinds the rail bottom region AB, and the left robot stands by;

[0014] The left robot grinds the left rail foot bottom region AJ, and the right robot grinds the right rail foot bottom region BC;

[0015] The right robot grinds the right rail foot top surface and triangular region CD, and the left robot stands by;

[0016] The right robot grinds the right rail waist and rail jaw region DE and the right rail top side region EF, and the left robot grinds the left rail foot top surface and triangular region JI,

[0017] The left robot grinds the left rail waist and rail jaw region IH, and the right robot stands by.

[0018] Further, the grinding points of the left rail waist and the right rail waist are determined by the following method: taking the center position O of the rail as the base point, wherein the coordinates of O are (OX, OY, OZ), from the base point, a plurality of reference points Oi, i = 1, 2…n, are sequentially selected upwards and downwards along the surface of the rail, wherein n is determined according to the type of the rail, the coordinates of Oi are (OIX, OIY, OIZ), the center position of the rail obtained by visual recognition is Oc, the coordinates of Oc are (OCX, OCY, OCZ), the offset dx = OCX - OX, dy = OCY - OY, dz = OCZ - OZ between the center position of the rail obtained by visual recognition Oc and the center position O of the rail is calculated, and the coordinates of the plurality of reference points Oi are sequentially adjusted according to the above offset, the coordinates of the grinding point Oci are (OciX, OciY, OciZ), and the above grinding point Oci and Oc constitute the grinding path.

[0019] Further, after the step S4, the grinding quality is inspected, and if the inspection is unqualified, re-grinding or manual supplementary grinding is performed; after the inspection is qualified, the rail positioning device releases the rail, and the rail moves out of the grinding station;

[0020] The inspection method for the polishing quality comprises: the two sides of the robot are provided with 3D cameras for scanning the welding area again to detect the welding bump residual thickness, and if the welding bump residual thickness exceeds the set value, it is determined as unqualified.

[0021] The application further provides a steel rail intelligent polishing device, comprising a steel rail positioning device, a left robot and a right robot, wherein the left robot and the right robot are provided with a belt polishing device and a 3D camera at the front end, and the above polishing method is adopted.

[0022] Further, the steel rail intelligent polishing device further comprises a formula system, wherein the formula system stores different polishing forces corresponding to different steel rail types and the rotating speeds of the belt polishing device, and in step S3, after the steel rail type is determined to be consistent, the corresponding polishing force and the rotating speed of the belt polishing device are called to control the work of the belt polishing device.

[0023] Further, the steel rail positioning device is arranged along the extension direction of the steel rail, and is used for locking and positioning the steel rail to prevent the steel rail from being displaced during polishing.

[0024] Further, the 3D camera is a laser 3D industrial camera, and is further used for taking pictures of the welding position before the steel rail is locked and after the steel rail is released after polishing.

[0025] The steel rail polishing technology of the application adopts 3D visual recognition technology, intelligently recognizes the steel rail type, the welding position, the steel rail offset angle and the welding thickness, automatically plans a polishing path according to the set polishing process, realizes automatic intelligent polishing of the steel rail welding, and improves the polishing operation efficiency and the polishing operation quality. By setting a reasonable polishing path: (1) in the case of avoiding interference between the two mechanical arms, the non-interference area is worked synchronously to shorten the operation time as much as possible. (2) the polishing tool and the steel rail surface are kept in good contact to ensure the polishing effect. (3) the polishing path is continuously worked as much as possible to ensure the polishing effect and avoid the occurrence of wrong edges between different paths. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is a schematic diagram of a steel rail polishing path;

[0027] Figure 2 is a schematic diagram of the distribution of the coordinate points of the rail waist polishing path;

[0028] Figure 3 is a schematic diagram of the structure of a steel rail intelligent polishing device;

[0029] Figure 4 is a schematic diagram of the work of the belt polishing device. DETAILED DESCRIPTION

[0030] The steel rail welding intelligent polishing method of the application mainly comprises the following steps:

[0031] S1: conveying the steel rail: conveying the welded steel rail to the grinding station and moving the weld to the grinding range, locking the steel rail with a rail positioning device;

[0032] S2: setting the rail type and grinding intensity;

[0033] S3: the robots on both sides of the rail use a 3D camera to visually identify the weld in the grinding area, intelligently identify the rail type, and compare it with the set rail type. If they are not consistent, a prompt message is given and the grinding is stopped. If they are consistent, the weld position, rail offset angle, and weld thickness are identified, and the grinding speed and grinding path are calculated;

[0034] S4: the grinding devices connected to the robots on both sides of the rail grind the rail weld according to the grinding path and grinding speed;

[0035] S5: after grinding is completed, the robot returns to the initial position and prepares for the next grinding operation.

[0036] In the above step S2, the setting of the rail model is input by manual. In the above step S3, the method of intelligently identifying the rail type includes: for various rail models, a rail database model is established, the pictures taken by the 3D camera are point cloud extracted, and matched with the rail database model, so as to intelligently identify the rail type.

[0037] The grinding intensity and grinding material determine the specific grinding amount, and then the grinding speed can be determined according to the grinding amount. The grinding material and grinding intensity are determined according to the experience on site. At this time, visual identification is mainly used to identify the weld position, rail offset angle, and weld height, which are more complex manual measurement processes, and can also identify the type of rail. In this way, through double matching, on the one hand, it avoids input errors by personnel, and on the other hand, it combines manual experience and visual identification to quickly determine all key operating process parameters needed subsequently.

[0038] In the above step S3, the planning method of the grinding path is:

[0039] Referring to Figure 1 , first, the outer surface of the rail is divided into several regions: the rail bottom region AB, the left rail foot bottom region AJ, the right rail foot bottom region BC, the left rail foot top surface and triangular region JI, the right rail foot top surface and triangular region CD, the left rail waist and rail jaw region IH, the right rail waist and rail jaw region DE, the left rail top side region GH, the right rail top side region EF, and the rail top region FG.

[0040] Due to the interference area of the two robot arms, part of the area cannot be worked synchronously and needs to be avoided. Therefore, some paths can be worked synchronously, and some paths need to be worked individually. In the present application, the specific polishing sequence is as follows:

[0041] The left robot polishes the rail top area FG and the left rail top side area GH, and the right robot waits;

[0042] The right robot polishes the rail bottom area AB, and the left robot waits;

[0043] The left robot polishes the left rail foot bottom area AJ, and the right robot polishes the right rail foot bottom area BC;

[0044] The right robot polishes the right rail bottom foot upper surface and the triangular area CD, and the left robot waits;

[0045] The right robot polishes the right rail waist and jaw area DE and the right rail top side area EF, and the left robot polishes the left rail bottom foot upper surface and the triangular area JI,

[0046] The left robot polishes the left rail waist and jaw area IH, and the right robot waits.

[0047] The above path design can ensure the polishing effect, avoid misalignment, and has the highest efficiency. The rail top and rail bottom are the main contact working planes, so the best processing method is non-segmented processing, that is, the rail top area FG and the middle of the rail bottom area AB cannot be further divided, otherwise misalignment may occur at the segmented junction, affecting the polishing effect. Therefore, one mechanical arm is responsible for the rail top, and one mechanical arm is responsible for the rail bottom. The rail bottom corners, rail waist and rail jaw on both sides are respectively responsible for the corresponding side mechanical arm.

[0048] Affected by the robot arm posture and the gravity of the polishing tool itself, the convex curved surface and the horizontal plane (FG, AB area) move from far to near (pull back), so as to ensure good contact between the polishing tool and the rail surface. For the concave curved surface (C-D-E, H-I-J), move from near to far (push forward).

[0049] For the GH and EF areas, the two areas are perpendicular to the ground, so the principle of less segmentation is maintained, GH is continuously worked with FG, and EF is continuously worked with DE.

[0050] Further, for the polishing of the rail waist, refer to Figure 2For example, the right rail waist, the determination method of the coordinate points on the path is as follows: the center position O of the steel rail is selected as the base point, wherein the coordinates of O are (OX, OY, OZ), a plurality of reference points Oi, i=1, 2…n are sequentially selected upwards and downwards along the surface of the steel rail from the base point, wherein n is determined according to the type of the steel rail, the coordinates of Oi are (Oix, OiY, OiZ), the center position of the steel rail obtained by visual recognition is Oc, and the coordinates of Oc are (OcX, OcY, OcZ), the offset amount dx=OcX-OX, dy=OcY-OY, and dz=OcZ-OZ between the center position of the steel rail obtained by visual recognition and the center position O of the steel rail is calculated, and the coordinates of the plurality of reference points Oi are sequentially adjusted according to the above offset amount, and the coordinates of the polishing point Oci are (OiX+dx, OiY+dy, OiZ+dz), and the above polishing point Oci and Oc constitute a polishing path.

[0051] Wherein O and Oi are both preset positions, and Oc is a position obtained by visual recognition.

[0052] The polishing speed = the grinding amount / the weld thickness, wherein the grinding amount is determined in step S2.

[0053] Further, after the above step S4, the polishing quality is inspected, and if the inspection is unqualified, re-polishing or manual supplementary polishing is performed.

[0054] The method for inspecting the polishing quality includes that the two robots use 3D cameras to scan the weld area again to detect the residual thickness of the weld bumps, and if the residual thickness exceeds the set value (0.3mm), it is determined as unqualified.

[0055] Referring to Figures 3-4 The structure of the steel rail intelligent polishing device includes a steel rail positioning device 1, a left robot 2, and a right robot 3, the front ends of the left robot 2 and the right robot 3 are provided with a belt polishing device 4 and a 3D camera 5, the steel rail positioning device 1 is arranged along the extension direction of the steel rail, and a plurality of steel rail positioning devices 1 are arranged for locking and positioning the steel rail to prevent the steel rail from moving during polishing, and the functions thereof include hydraulic clamping and fixing of the steel rail and hydraulic pressing and fixing of the steel rail. Generally, the position of the weld reached during each transmission and locking of the steel rail should be ensured within a range of ±20cm, the steel rail positioning device should first perform a hydraulic clamping action, and then perform a hydraulic pressing action, so as to realize locking and fixing of the steel rail and avoid movement of the steel rail during polishing.

[0056] In the present application, according to the characteristics of the double robots, the two belt polishing devices 4 cooperate with each other, combine with the 3D visual recognition technology, and realize full-domain dead-angle-free high-precision polishing of the rail top, rail waist, and rail bottom of the steel rail through dynamic path planning.

[0057] Further, the 3D camera is a laser 3D industrial camera, which can also be used to take pictures of the weld position after the steel rail is locked and before the steel rail is released after grinding is completed.

[0058] Further, the intelligent grinding device further comprises a formula system, which stores different grinding intensities corresponding to different steel rail types and the rotating speed of the belt sander, and in step S3, after the steel rail type is determined to be consistent, the corresponding grinding intensity and the rotating speed of the belt sander are called to control the work of the belt grinding device.

[0059] The above embodiments are only used to illustrate the technical solutions of the present application, but not to limit it; although the foregoing embodiments of the present application have been described in detail, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for intelligent grinding of rail welds, characterized in that, Includes the following steps: S1: Rail conveying: The welded rail is conveyed to the grinding station and the weld is moved into the grinding area. The rail is locked using a rail positioning device. S2: Set the rail type; S3: The robots on both sides of the rail use 3D cameras to visually identify the weld seams in the grinding area, intelligently identify the rail type, and compare it with the set rail type. If they are inconsistent, a prompt message is given and grinding is stopped; if they are consistent, the weld seam position, rail offset angle, and weld thickness are identified, and the grinding speed and grinding path are calculated. S4: The robots on both sides of the rail drive their respective connected grinding devices to grind the rail welds according to the grinding path and grinding speed. S5: After polishing is completed, the robot returns to its initial position, ready for the next polishing operation; In step S3 above, the method for intelligently identifying rail types includes: establishing a rail database model for various rail models, extracting point clouds from images captured by a 3D camera, and matching them with the rail database model to intelligently identify rail types; in step S3 above, the method for planning the grinding path is as follows: The outer surface of the rail is divided into several regions: rail base region AB, left rail foot base region AJ, right rail foot base region BC, left rail base upper surface and triangular area JI, right rail base upper surface and triangular area CD, left rail web and rail jaw region IH, right rail web and rail jaw region DE, left rail top side region GH, right rail top side region EF, and rail top region FG. The robot on the left is grinding the top rail area FG and the side rail area GH, while the robot on the right is in standby mode. The robot on the right is grinding the bottom area AB of the track, while the robot on the left is in standby mode. The left robot has its left track foot area AJ, and the right robot has its right track foot area BC polished. The robot on the right is polishing the upper surface of the right rail base and the triangular area CD, while the robot on the left is in standby mode. The robot on the right grinds the right rail web and jaw area DE, as well as the right rail top side area EF, while the robot on the left grinds the upper surface of the left rail bottom foot and the triangular area JI. The left robot grinds the left rail waist and rail jaw area IH, while the right robot stands by. The grinding points for the left and right rail waists are determined by the following method: The center position O of the rail is selected as the base point, with coordinates (OX, OY, OZ). Starting from the base point, multiple reference points Oi are selected sequentially upwards and downwards along the surface of the rail, i=1,2…n, where n is determined according to the type of rail. The coordinates of Oi are (Oix, OiY, OiZ). The center position of the rail is obtained by visual recognition as Oc, with coordinates (OcX, OcY, OcZ). The offsets between the visually recognized center position Oc and the center position O are calculated as follows: dx=OcX-OX, dy=OcY-OY, dz=OcZ-OZ. Based on these offsets, the coordinates of the multiple reference points Oi are adjusted sequentially. The coordinates of the grinding point Oci are (OiX+dx, OiY+dy, OiZ+dz). The grinding points Oci and Oc constitute the grinding path.

2. The intelligent grinding method for rail welds according to claim 1, characterized in that, After step S4 above, the grinding quality is inspected. If the inspection fails, the grinding is repeated or manual grinding is performed. After the inspection is passed, the rail positioning device releases the rail and moves the rail out of the grinding station. The method for inspecting the grinding quality includes: the robots on both sides use 3D cameras to scan the weld area again to detect the thickness of the weld bead residue. If it exceeds the set value, it is judged as unqualified.

3. A rail intelligent grinding device, comprising a rail positioning device (1), a left robot (2), and a right robot (3), wherein the front ends of the left robot (2) and the right robot (3) are equipped with a belt grinding device (4) and a 3D camera (5), characterized in that, The invention includes an intelligent rail grinding system for performing the grinding method described in any one of claims 1-2.

4. The intelligent rail grinding equipment according to claim 3, characterized in that, The intelligent grinding equipment also includes a formula system, which stores the grinding intensity and belt sander speed corresponding to different rail types. In step S3, after the rail types are determined to be consistent, the corresponding grinding intensity and belt sander speed are retrieved to control the operation of the belt sander (4).

5. The intelligent rail grinding equipment according to claim 3, characterized in that, Multiple rail positioning devices are arranged along the extension direction of the rail to lock and position the rail, preventing it from shifting during the grinding process.

6. The intelligent rail grinding equipment according to claim 5, characterized in that, The 3D camera is a laser 3D industrial camera. The 3D camera is also used to take pictures of the weld position after the rail is locked and before the rail is released after grinding.

Images