A high-load, high-precision mobile positioning robot and its positioning method

Abstract

This invention discloses a high-load, high-precision mobile positioning robot. The planar adjustment module is a 3-PRR parallel mechanism with three degrees of freedom: X-axis translation, Y-axis translation, and Z-axis rotation. Multiple bullseye bearings provide unconstrained vertical support and lateral guidance. The vision measurement module employs monocular vision measurement technology, combined with hand-eye calibration matrices to calculate pose errors, achieving precise coupling with the 3-PRR parallel mechanism. Pose deviation information is fed back to the 3-PRR parallel mechanism for accurate pose compensation. This invention introduces a homogeneous transformation matrix between the camera coordinate system and the static platform coordinate system of the 3-PRR parallel mechanism, establishing a quantified pose transformation relationship. This enables precise calculation and compensation of AGV pose deviations. Furthermore, it adds standardized detection steps for positioning accuracy, using a laser tracker to quantify the compensation accuracy, providing a theoretical basis and quantitative indicators for positioning accuracy verification.

Description

Technical Field

[0001] This invention belongs to the field of robotics technology, specifically relating to a high-load, high-precision mobile positioning robot and its positioning method. Background Technology

[0002] With the rapid advancement of major national projects such as aerospace and marine propulsion towards large-scale and integrated manufacturing, large components such as rocket modules, aircraft housings, and nuclear power plant cylinders require long-distance transport spanning tens or even hundreds of meters within the final assembly plant, and millimeter-level precision closure in narrow docking stations. Large-scale movement and high-precision docking have become common and essential requirements for high-end manufacturing. However, most Chinese enterprises still rely on the traditional method of overhead crane lifting combined with manual hand-cranked support frames: the support frames can only move in one direction, unable to achieve six-degree-of-freedom positional adjustment, the assembly process depends entirely on experience, resulting in low efficiency and high risk. Furthermore, the large size and weight of these components mean that conventional mobile robots, with their added attitude adjustment mechanisms, have a high center of gravity, making them difficult to maneuver in narrow, confined space areas such as module corridors or wing-fuselage docking zones, and unable to adapt to rapid docking operations in confined spaces. This severely restricts the flexible mass production and quality improvement of large equipment. To meet the demands for rapid, precise, and agile assembly of large components, designing a high-load, high-precision mobile positioning robot is an effective solution.

[0003] Currently, most heavy-duty mobile platforms on the market focus on transportation rather than assembly. Their positioning accuracy is only sufficient for workshop logistics requirements, lacking end-effector fine-tuning capabilities and unable to directly complete millimeter-level docking. When multi-joint robotic arms are added, their load-bearing capacity is insufficient due to the flexibility of the serial structure, and they are prone to significant deformation under heavy loads, making them unsuitable as rigid supports for the final docking process. If a general parallel attitude adjustment scheme is used, a six-degree-of-freedom mechanism is often stacked on the vehicle body, causing the vehicle's center of gravity to rise synchronously with the load. This results in poor stability when driving, turning, or braking in low or narrow spaces, and still poses a risk of tipping over and interference. Therefore, achieving a balance between high load, wide range, high precision, and low center of gravity is the core challenge that mobile assembly robots urgently need to overcome.

[0004] The purpose of this invention is to overcome the shortcomings of the prior art by organically combining the coarse positioning capability of AGV mobile robots with the fine compensation capability of large-load parallel mechanisms, and providing an autonomous mobile positioning robot with high load-bearing capacity, large adjustment range, high assembly accuracy, and low center of gravity. Summary of the Invention

[0005] This invention is proposed to solve the problems existing in the prior art, and its purpose is to provide a high-load, high-precision mobile positioning robot and its positioning method.

[0006] The technical solution of this invention is: A high-load, high-precision mobile positioning robot includes: A mobile trolley, on which a carrier body is mounted; The motion platform is used to output the pose adjustment results; The planar adjustment module installed on the carrier body is a 3-PRR parallel mechanism, including a planar support chain and a bullseye bearing; The planar branch group consists of three PRR-type branches, which are used to drive the moving platform to perform pose adjustment by translating in the X and Y directions and rotating around the Z axis; The bullseye bearing makes point contact with the moving platform and is laterally located outside the swing area of ​​the planar branch group, providing unrestrained vertical support and lateral guidance for the moving platform. A vision measurement module is installed on the carrier vehicle to acquire images of external targets and calculate the current pose information of the moving platform. The mobile positioning robot achieves high-precision positioning through the following operations: Collect images of the target at the workstation and calculate the pose compensation amount of the moving platform relative to the theoretical reference pose. The calculated pose compensation amount is compared with the preset theoretical reference pose to obtain the pose error of X-axis and Y-axis translation and rotation around the Z-axis. Determine whether the pose error exceeds the set threshold. If it does, drive the plane adjustment module to drive the moving platform to perform pose compensation. After compensation, re-acquire the target image and repeat the calculation and difference judgment operation until the error converges to the threshold range. Once the error reaches the target, determine whether the current workstation is the final target workstation. If so, the positioning is completed; otherwise, set the coordinates of the next workstation as the new theoretical reference pose and continue with subsequent positioning operations.

[0007] Furthermore, the planar adjustment module also includes a connecting plate connected to the carrier body, and the planar support chain and the bullseye bearing are disposed in the longitudinal space between the connecting plate and the moving platform.

[0008] Furthermore, the planar branch group consists of a first branch, a second branch, and a third branch with identical structures from the fixed end to the output end, used to realize translation in the X and Y directions and rotation around the Z axis.

[0009] Furthermore, each branch includes a sliding joint mounted on the connecting plate. The linear output part of the sliding joint is equipped with a first rotary joint, the rotary output part of the first rotary joint is connected to a connecting rod, and the other end of the connecting rod is equipped with a second rotary joint, which is connected to the moving platform.

[0010] Furthermore, the vision measurement module also includes a camera bracket connected to the vehicle body, on which an industrial camera is mounted.

[0011] Furthermore, the mobile trolley is equipped with wheels at its bottom, which provide support and propulsion for the overall movement of the trolley, enabling it to perform large-scale workstation movements and coarse positioning actions.

[0012] Another object of the present invention is to provide a positioning method for a high-load, high-precision mobile positioning robot, comprising the following steps: Step A: Based on the preset global navigation map of the workshop and workstation coordinate information, drive the mobile positioning robot to travel along the planned path to the set workstation area and brake to stop, thus completing the coarse positioning operation of the mobile positioning robot at the target workstation. Step B: Acquire the target image at the workstation, and combine the homogeneous transformation matrix from the camera coordinate system obtained by hand-eye calibration to the static platform coordinate system of the 3-PRR parallel mechanism to calculate the pose compensation amount of the moving platform of the 3-PRR parallel mechanism relative to the theoretical reference pose. Step C: Perform a difference calculation between the pose compensation amount calculated in Step B and the preset theoretical reference pose to obtain the pose errors of X-axis and Y-axis translation and rotation around the Z-axis. Step D: Determine whether the pose error calculated in step C exceeds the set threshold. If it does, drive the 3-PRR parallel mechanism to drive the moving platform to perform pose deviation compensation. After compensation, return to step B to re-acquire the target image at the workstation and repeat the subsequent steps until the error converges to the threshold range. Step E: If the pose error does not exceed the set threshold, read the task endpoint flag and determine whether the current workstation is the final target workstation; if so, the positioning is completed; otherwise, set the coordinates of the next workstation as the new theoretical reference pose and return to step A.

[0013] Furthermore, the specific process of calculating the pose compensation in step B includes: Define the ideal static platform coordinate system, the actual static platform coordinate system, the checkerboard coordinate system, and the camera coordinate system, and fix the industrial camera to the mobile positioning robot. The homogeneous transformation matrix from the camera coordinate system to the static platform coordinate system is obtained through hand-eye calibration. The homogeneous transformation matrix of the camera coordinate system relative to the checkerboard coordinate system is solved using the PnP algorithm. The transformation relationships between the ideal and actual static platform coordinate systems relative to the checkerboard coordinate system are established. The homogeneous transformation matrix is ​​expanded into a combination of a rotation matrix and a translation vector, and the Euler angles are obtained based on the rotation matrix. Ignoring the influence of the Z-axis dimension, the pose compensation amount of the 3-PRR parallel mechanism motion platform is finally calculated using the coordinate transformation formula in the two-dimensional plane. ; In the formula These are the coordinates of the moving platform in the actual static platform coordinate system. Translational compensation in the X and Y directions. This is the rotational compensation amount of the moving platform around the Z-axis. These are the Euler angles around the Z-axis under the ideal pose and the actual pose, respectively. in, Used for calculating translational pose error in step C. Used for calculating rotational pose error in step C.

[0014] Furthermore, after all station positioning actions are completed, the positioning accuracy is tested using a laser tracker. The specific process includes: When the moving platform is in its actual pose, establish a rectangular coordinate system at its plane center and set target points on the edge line of the moving platform. Measured by a laser tracker The target points are determined by solving for their coordinates in the ideal pose coordinate system of the moving platform, based on the geometric relationships between the target points. The coordinates in the ideal pose coordinate system are then calculated, and the rotation angle of the actual pose coordinate system of the moving platform relative to the ideal pose coordinate system, as well as the coordinates of the origin of the actual pose coordinate system of the moving platform in the ideal pose coordinate system, are then calculated. Finally, the actual positioning accuracy of the mobile positioning robot is obtained. If the detection result does not meet the preset accuracy requirements, the process returns to step D to re-perform pose deviation compensation.

[0015] The beneficial effects of this invention are as follows: This invention discloses a high-load, high-precision mobile positioning robot. Its planar adjustment module is a 3-PRR parallel mechanism with three degrees of freedom: X-axis translation, Y-axis translation, and Z-axis rotation. It provides unconstrained vertical support and lateral guidance through multiple sets of bullseye bearings. The vision measurement module adopts monocular vision measurement technology and combines hand-eye calibration matrix to calculate pose error, achieving precise coupling with the 3-PRR parallel mechanism. The pose deviation information is fed back to the 3-PRR parallel mechanism to complete accurate pose compensation.

[0016] This invention introduces a homogeneous transformation matrix between the camera coordinate system and the static platform coordinate system of the 3-PRR parallel mechanism, establishes a quantified pose transformation relationship, realizes accurate calculation and compensation of AGV pose deviation, and adds a standardized detection step for positioning accuracy. The compensation accuracy is quantified by using a laser tracker, providing a theoretical basis and quantitative indicators for positioning accuracy verification.

[0017] This invention combines the advantages of high overall rigidity, large working space, high positioning accuracy, and low center of gravity, enabling high-load, high-precision assembly applications. It solves the technical problems of the lack of pose compensation mechanism and the lack of quantitative detection method for positioning accuracy in existing mobile positioning robots. It is suitable for long-distance transportation and millimeter-level precise docking of large components in aerospace, marine power and other fields. Attached Figure Description

[0018] Figure 1This is a schematic diagram of the overall structure of the present invention; Figure 2 This is a schematic diagram of the mobile vehicle in this invention; Figure 3 This is a schematic diagram of the plane adjustment module in this invention; Figure 4 This is a schematic diagram of the branch of the planar adjustment module in this invention; Figure 5 This is a schematic diagram of the visual measurement module in this invention; Figure 6 This is a flowchart of the visual measurement module in this invention; Figure 7 This is a schematic diagram of the parallel mechanism for compensating for AGV posture deviation. Figure 8 This is a schematic diagram of the positioning accuracy detection principle for a parallel composite robot. in: 1. Moving platform 2. Mobile cart 3 Bullseye bearings 4 Vision measurement module 5. Planar adjustment module; 6. Carrier body 7 Control system 8 Wheels 9 Emergency Stop Button 10 Start / Stop Button 11 First branch 12 Second branch 13 Third branch 14 Connecting plate 15 Sliding joint 16 First rotary joint 17 Connecting rod 18 Second revolute joint 19 Servo motors 20 Camera brackets 21. Industrial cameras. Detailed Implementation

[0019] The present invention will now be described in detail with reference to the accompanying drawings and embodiments: Example

[0020] A high-load, high-precision mobile positioning robot includes: Mobile trolley 2, on which a carrier body 6 is mounted; Motion platform 1 is used to output the pose adjustment results; The planar adjustment module 5, which is a 3-PRR parallel mechanism, is installed on the carrier body 6 and is used to drive the moving platform 1 to perform posture adjustment; The visual measurement module 4 is installed on the carrier vehicle body 6 and is used to acquire external target images and calculate the current pose information of the moving platform 1. After the visual measurement module 4 converts the pose deviation information it calculates to the static platform coordinate system of the plane adjustment module 5, it transmits it to the plane adjustment module 5 to drive the moving platform 1 to perform pose compensation.

[0021] Mobile positioning robots achieve high-precision positioning through the following operations: Collect images of the target at the workstation and calculate the pose compensation amount of the moving platform 1 relative to the theoretical reference pose. The calculated pose compensation amount is compared with the preset theoretical reference pose to obtain the pose error of X-axis and Y-axis translation and rotation around the Z-axis. Determine whether the pose error exceeds the set threshold. If it does, drive the plane adjustment module 5 to drive the moving platform 1 to perform pose compensation. After compensation, re-acquire the target image and repeat the calculation and difference judgment operation until the error converges to the threshold range. Once the error reaches the target, determine whether the current workstation is the final target workstation. If so, the positioning is completed; otherwise, set the coordinates of the next workstation as the new theoretical reference pose and continue with subsequent positioning operations.

[0022] In this embodiment, as Figure 1 As shown, the structure of the whole machine of the present invention includes a moving platform 1, a moving trolley 2 that realizes a large stroke, a vision measurement module 4, and a plane adjustment module 5 that realizes a large load.

[0023] Specifically, the vision measurement module 4 is installed on the edge of the mobile car 2, using the mobile car 2 as the mounting base. After the mobile car 2 is positioned, the vision measurement module 4 performs pose error calculation. The plane adjustment module 5 is used to compensate for the pose error, thereby achieving high-precision position control of the mobile positioning robot.

[0024] Specifically, the planar adjustment module 5 is installed on the mobile trolley 2, using the mobile trolley 2 as the base. After the mobile trolley 2 has made a large-stroke position adjustment, the planar adjustment module 5 makes a small-range planar position adjustment.

[0025] In one implementation, the moving platform 1 in this embodiment is mounted above the plane adjustment module 5. However, depending on the actual application and adjustment needs, it can also be mounted on the side or below the plane adjustment module 5.

[0026] like Figure 2 As shown, Figure 2 This is a schematic diagram of the mobile trolley of the present invention. The mobile trolley 2 includes a supporting body 6, the frame structure of the supporting body 6 provides the mounting base, and wheels 8 are installed on both sides of the supporting body 6, that is, the bottom of the mobile trolley 2.

[0027] In one embodiment, the carrier body 6 is an electric drive vehicle. The carrier body 6 is equipped with an emergency stop button 9 and a start / stop button 10 located on the side. The emergency stop button 9 is used to cut off the power in an emergency, and the start / stop button 10 is used for normal power-on and power-off.

[0028] In one implementation, a control system 7 is provided in the carrier body 6. The control system 7 integrates data from the vision measurement module 4, the emergency stop button 9, and the start / stop button 10 to control the mobile trolley 2.

[0029] Furthermore, in the embodiments, the planar adjustment module 5 may include a connecting plate 14 connected to the carrier body 6. A planar support chain group and a bullseye bearing 3 are arranged in the longitudinal space between the connecting plate 14 and the moving platform 1. The planar support chain group consists of three PRR type supports, which are used to drive the moving platform 1 to perform position adjustment by translating in the X and Y directions and rotating around the Z axis.

[0030] In this embodiment, as Figure 3 As shown in the schematic diagram of the planar adjustment module 5, the planar adjustment module 5 includes a connecting plate 14, the lower end of which is connected to the carrier body 6. A planar support chain assembly and multiple bullseye bearings 3 are installed on the connecting plate 14. The planar support chain assembly adjusts the planar position of the moving platform 1, and the bullseye bearings 3 provide vertical support for the moving platform 1.

[0031] Furthermore, in the embodiments, the bullseye bearing 3 can be considered to be in point contact with the moving platform 1 and laterally located outside the swing area of ​​the planar branch group, for providing unconstrained vertical support and lateral guidance for the moving platform 1.

[0032] In this embodiment, the bullseye bearing 3 is arranged outside the swing area of ​​the planar branch assembly. The bullseye bearing 3 has freely rolling spherical crown balls inside, and the apex of the spherical crown forms point contact with the moving platform 1. There are multiple bullseye bearings 3, and the apexes of the spherical crowns of the multiple bullseye bearings 3 are located on the same horizontal plane, which together provide multi-point support for the moving platform 1.

[0033] Furthermore, in the embodiments, the planar branch group can be considered to consist of a first branch 11, a second branch 12, and a third branch 13, and the first branch 11, the second branch 12, and the third branch 13 have the same structure from the fixed end to the output end, which is used to realize translation in the X and Y directions and rotation around the Z axis.

[0034] In this embodiment, the planar branch group constitutes the motion branch of the 3-PRR parallel mechanism, which consists of three PRR-type branches to realize pose compensation for X-axis and Y-axis translation and rotation around the Z-axis.

[0035] The planar branch group includes a first branch 11, a second branch 12, and a third branch 13. The first branch 11, second branch 12, and third branch 13 have identical structures, differing only in their placement. Specifically, the direction of the bottom sliding joint of the second branch 12 is parallel to the longitudinal direction of the connecting plate 14. The directions of the bottom sliding joints in the first branch 11 and third branch 13 are at a certain angle to the direction of the bottom sliding joint in the second branch 12, and the directions of the bottom sliding joints in the first branch 11 and third branch 13 are symmetrical to the perpendicular direction of the bottom sliding joint in the second branch 12. The first branch 11, second branch 12, and third branch 13 have three degrees of freedom: translation in the X direction, translation in the Y direction, and rotation around the Z direction.

[0036] Furthermore, in the embodiments, each branch may include a sliding joint 15 mounted on the connecting plate 14. The linear output portion of the sliding joint 15 is equipped with a first rotating joint 16. The rotational output portion of the first rotating joint 16 is connected to a connecting rod 17. The other end of the connecting rod 17 is equipped with a second rotating joint 18, which is connected to the moving platform 1.

[0037] In this embodiment, as Figure 4 As shown in the schematic diagram of the planar adjustment module branches, each of the first branch 11, the second branch 12, and the third branch 13 includes a sliding joint 15. The fixed and guiding parts of the sliding joint 15 are fixed to the connecting plate 14. The sliding joint 15 includes a linear output part capable of planar movement, and the linear output part has a flat upper end. A first rotary joint 16 is installed at the upper end of the linear output part. The axis of the first rotary joint 16 is perpendicular to the movement direction of the sliding joint 15. The connecting rod 17 is the rotational follower of the first rotary joint 16. At the same time, a second rotary joint 18 is installed at the far end of the connecting rod 17 to realize the rotational connection between the intermediate moving platform 1 and the second rotary joint 18.

[0038] In one implementation, the movable pair 15 is driven by a servo motor 19, which is connected to the movable pair 15 via a coupling. The servo motor 19 can also be connected by a direct drive or a transmission belt.

[0039] Furthermore, in the embodiments, the vision measurement module 4 may include a camera bracket 20 connected to the vehicle body 6, on which an industrial camera 21 is mounted.

[0040] In this embodiment, as Figure 5 As shown in the schematic diagram of the vision measurement module, the vision measurement module 4 is installed on the edge of the carrier body 6 and includes an industrial camera 21 and a camera bracket 20. The industrial camera 21 is connected and fixed to the carrier body 6 through both ends of the camera bracket 20.

[0041] As one implementation method, the industrial camera 21 can use, but is not limited to, QR codes, checkerboard patterns, coded markers, etc., to measure the target.

[0042] Furthermore, in the embodiments, the visual measurement module 4 is used to acquire the target image at the workstation, calculate the pose error between the current pose information and the theoretical reference pose, and transmit the pose error as the pose deviation information to the plane adjustment module 5.

[0043] Furthermore, in the embodiments, the bottom of the mobile trolley 2 is equipped with wheels 8, which provide support and propulsion for the overall movement of the mobile trolley, and work together with the mobile trolley to achieve a wide range of workstation movement and coarse positioning.

[0044] As one implementation method, wheel 8 can be, but is not limited to, Mecanum wheels, steering wheels, casters, and differential wheels. When wheel 8 is a Mecanum wheel, it is driven by a direct-drive motor. Example

[0045] like Figure 6 , Figure 7 , Figure 8 As shown, Figure 6 This is a flowchart of the visual measurement module of the present invention. Figure 7 This is a schematic diagram illustrating the principle of a parallel mechanism for compensating for AGV posture deviations. Figure 8 This is a schematic diagram of the positioning accuracy detection principle of a parallel composite robot, which enables the mobile positioning robot to move over a wide area and achieve high-precision positioning. The positioning method includes the following steps: A. Based on the preset global navigation map of the workshop and the coordinate information of the workstation, the mobile positioning robot is driven to move along the planned path to the set workstation area and stop. During the movement, the robot's pose is corrected in real time through the machine's LiDAR to ensure that the workstation target is completely in the field of view of the industrial camera of the vision measurement module, thus completing the coarse positioning operation of the robot at the target workstation.

[0046] B. Acquire images of the target at the workstation and extract features, combining this with the homogeneous transformation matrix from the camera coordinate system obtained through hand-eye calibration to the static platform coordinate system of the 3-PRR parallel mechanism. .

[0047] Define four types of coordinate systems: the coordinate system of the static platform of the 3-PRR parallel mechanism when the AGV is in its initial ideal pose. (Original point) Located in the center of the quiet platform, shaft and The axes are parallel to the two perpendicular sides of the static platform. The axes are determined by the right-hand rule; the coordinate system of the parallel mechanism's static platform when the AGV is in its actual position. (Original point) Located in the center of the quiet platform, shaft and The axes are parallel to the two perpendicular sides of the static platform. (The axes are determined by the right-hand rule) chessboard coordinate system and camera coordinate system Furthermore, the camera is fixedly connected to the AGV. The three coordinate axes are parallel to each other; the pose relationships of the four types of coordinate systems are established through the above homogeneous transformation matrix: ; In the formula Let be the homogeneous transformation matrix of the ideal static platform coordinate system relative to the chessboard coordinate system. This is the homogeneous transformation matrix of the camera coordinate system relative to the checkerboard coordinate system (obtained by the PnP algorithm). Let be the homogeneous transformation matrix of the ideal static platform coordinate system relative to the camera coordinate system. Let be the homogeneous transformation matrix of the actual static platform coordinate system relative to the chessboard coordinate system. Let be the homogeneous transformation matrix of the actual static platform coordinate system relative to the camera coordinate system. This formula is used to indirectly solve the ideal and actual pose relationship of the static platform relative to the chessboard grid.

[0048] Expand the homogeneous transformation matrix into a combination of a rotation matrix and a translation vector: ; In the formula These are the rotation matrices of the ideal and actual static platform coordinate systems relative to the checkerboard coordinate system, respectively, used to describe the rotational relationship between the coordinate systems; These are the translation vectors of the ideal and actual static platform coordinate systems relative to the checkerboard coordinate system, respectively, used to characterize the translation relationship between the coordinate systems, thereby achieving the separation of rotation and translation components in the pose relationship.

[0049] By using the basic rotation matrix around the X, Y, and Z axes: ; In the formula Let cosine and sine be trigonometric functions respectively. For the rotation angle about the corresponding coordinate axis, based on this basic rotation matrix, around a fixed coordinate system The axes rotate sequentially The composite rotation matrix can be represented as: ; Euler angles are obtained by solving the rotation matrix: ; The rotation matrix can be obtained using the above formula. The three Euler angles are Rotation matrix The three Euler angles are .

[0050] Since the parallel mechanism only performs in-plane motion and rotation about the Z-axis, we assume a Cartesian coordinate system. Coplanar, ignoring the influence of the Z-axis dimension, and the Euler angles of the two poses satisfy the following relationship. Simultaneously, define the translation vector of the ideal static platform coordinate system. Translation vector of the actual static platform coordinate system coordinate system origin of coordinates In a two-dimensional plane coordinate system The coordinates in are respectively and Through the coordinate transformation formula in a two-dimensional plane: ; In the formula ; coordinates Revolved coordinate system of Axis rotation Two-dimensional rotation matrix, coordinate system arrive A two-dimensional translation vector is used to achieve coordinate transformation of the origin in different coordinate systems.

[0051] Finally, the pose compensation amount of the 3-PRR parallel mechanism motion platform is obtained through coordinate transformation: ; In the formula These are the coordinates of the moving platform in the actual static platform coordinate system. Translational compensation in the X and Y directions. This is the rotational compensation amount of the moving platform around the Z-axis. These are the Euler angles about the Z-axis under ideal and actual poses, respectively. This formula provides a quantitative basis for pose compensation of the 3-PRR parallel mechanism, which moves the platform to [the desired position / position] via a motor. This allows for the compensation of AGV posture deviation.

[0052] Step C: Perform a difference calculation between the initial pose obtained in Step B and the preset theoretical reference pose to calculate the current pose error, and obtain the specific error values ​​for translation in the X and Y directions and rotation around the Z axis. The error calculation is based on the theoretical reference pose to accurately quantify the degree of deviation between the actual pose and the ideal state.

[0053] Step D: Determine whether the pose error calculated in Step C exceeds a set threshold, which is set according to the actual assembly accuracy requirements; if it exceeds the threshold, start the 3-PRR parallel mechanism, and generate motion commands for each branch through inverse kinematics based on the pose compensation amount obtained in Step B, drive the servo motor to drive the sliding joint and revolute joint to move in coordination, and then drive the moving platform to complete the pose deviation compensation; after compensation, return to Step B, collect the target image at the workstation again and repeat the subsequent pose calculation and error calculation steps until the error converges to the threshold range; if the pose error does not exceed the set threshold, proceed directly to Step E.

[0054] Step E: Read the task endpoint marker and determine whether the current workstation is the final target workstation; if so, lock the moving trolley brake and complete this positioning action; otherwise, set the coordinates of the next workstation as the new theoretical reference pose and return to step A to continue the next round of coarse and fine positioning operations.

[0055] After all positioning actions are completed, a laser tracker is used to test the positioning accuracy. The core idea of ​​the test is to solve for the actual pose coordinate system of the moving platform. The origin In the ideal pose coordinate system coordinates in and the rotation angle between the two coordinate systems The specific detection process is as follows: when the moving platform is in its actual pose, a rectangular coordinate system is established at its plane center. ( shaft and The axes are parallel to the two perpendicular sides of the moving platform, and target points are set on the edges of the moving platform. ,in for The bottom right corner point was measured using a laser tracker. Ideal pose coordinate system of the moving platform ( shaft and The coordinates of the axes (which are parallel to the two perpendicular sides of the moving platform) Based on the geometric relationship between target points (vectors) with vector Perpendicular, vector with vector Establish coordinate relationship equations (parallel): ; In the formula Target The coordinates in the ideal pose coordinate system provide a mathematical basis for solving the coordinates of the unknown target point; then, by solving the above system of equations using Cramer's rule, the unknown target point can be obtained. Coordinates: ; In the formula Target The coordinate difference; then, the rotation angle of the actual pose coordinate system of the moving platform relative to the ideal pose coordinate system is solved using the target point coordinates: ; In the formula This represents the rotational deviation angle of the actual pose coordinate system about the Z-axis. Target The distance along the Y-axis is used to quantify the rotational deviation; then, based on the right triangle... geometric relationship ( right-angled side and (axis parallel) See Figure 8 It is the actual pose coordinate system of the moving platform. The origin In the ideal pose coordinate system Zhongyu same edge The projection point.

[0056] Solve for the origin of the actual pose coordinate system of the moving platform. Coordinates in the ideal pose coordinate system: ; In the formula Let be the coordinates of the origin of the actual pose coordinate system in the ideal pose coordinate system. The lengths of the long and short sides of the moving platform. It is half the length of the diagonal of the moving platform. For the ideal pose of the moving platform, the diagonal and The included angle of the axes enables precise positioning of the actual pose origin; finally, the actual positioning accuracy of the mobile positioning robot is obtained through the above calculations: ; In the formula The X and Y direction positioning deviations of the actual center of the moving platform in the ideal pose coordinate system. The formula provides a quantitative detection result for the positioning accuracy, which is the rotational positioning deviation around the Z-axis. If the detection result does not meet the preset accuracy requirements, the process returns to step D to re-perform pose deviation compensation.

[0057] This invention discloses a high-load, high-precision mobile positioning robot. Its planar adjustment module is a 3-PRR parallel mechanism with three degrees of freedom: X-axis translation, Y-axis translation, and Z-axis rotation. Multiple bullseye bearings provide unconstrained vertical support and lateral guidance. The vision measurement module employs monocular vision measurement technology, combined with hand-eye calibration matrices to calculate pose errors, achieving precise coupling with the 3-PRR parallel mechanism. The pose deviation information is fed back to the 3-PRR parallel mechanism for accurate pose compensation. This invention introduces a homogeneous transformation matrix between the camera coordinate system and the static platform coordinate system of the 3-PRR parallel mechanism, establishing a quantified pose transformation relationship. This enables precise calculation and compensation of AGV pose deviations. Furthermore, it adds standardized detection steps for positioning accuracy, using a laser tracker to quantify the compensation accuracy, providing a theoretical basis and quantitative indicators for positioning accuracy verification.

[0058] This invention combines the advantages of high overall rigidity, large working space, high positioning accuracy, and low center of gravity, enabling high-load, high-precision assembly applications. It solves the technical problems of the lack of pose compensation mechanism and the lack of quantitative detection method for positioning accuracy in existing mobile positioning robots. It is suitable for long-distance transportation and millimeter-level precise docking of large components in aerospace, marine power and other fields.

Claims

1. A positioning method for a high-load, high-precision mobile positioning robot, characterized in that: The mobile positioning robot includes a mobile vehicle (2) on which a carrier body (6) is mounted; a moving platform (1); a plane adjustment module (5) mounted on the carrier body (6) for driving the moving platform (1) to perform pose adjustment by translating in the X and Y directions and rotating around the Z axis; and a vision measurement module (4) mounted on the carrier body (6). The positioning method includes the following steps: Step A: Based on the preset global navigation map of the workshop and workstation coordinate information, drive the mobile positioning robot to travel along the planned path to the set workstation area and brake to stop, thus completing the coarse positioning operation of the mobile positioning robot at the target workstation. Step B: Acquire the target image at the workstation, and combine the homogeneous transformation matrix from the camera coordinate system obtained by hand-eye calibration to the static platform coordinate system of the 3-PRR parallel mechanism to calculate the pose compensation amount of the moving platform of the 3-PRR parallel mechanism relative to the theoretical reference pose. Step C: Perform a difference calculation between the pose compensation amount calculated in Step B and the preset theoretical reference pose to obtain the pose errors of X-axis and Y-axis translation and rotation around the Z-axis. Step D: Determine whether the pose error calculated in step C exceeds the set threshold. If it does, drive the 3-PRR parallel mechanism to drive the moving platform to perform pose deviation compensation. After compensation, return to step B to re-acquire the target image at the workstation and repeat the subsequent steps until the error converges to the threshold range. Step E: If the pose error does not exceed the set threshold, read the task endpoint flag and determine whether the current workstation is the final target workstation; if so, the positioning is completed; otherwise, set the coordinates of the next workstation as the new theoretical reference pose and return to step A. The specific process of calculating the pose compensation in step B includes: Define the ideal static platform coordinate system, the actual static platform coordinate system, the checkerboard coordinate system, and the camera coordinate system, and fix the industrial camera to the mobile positioning robot. The homogeneous transformation matrix from the camera coordinate system to the static platform coordinate system is obtained through hand-eye calibration. The homogeneous transformation matrix of the camera coordinate system relative to the checkerboard coordinate system is solved using the PnP algorithm. The transformation relationships between the ideal and actual static platform coordinate systems relative to the checkerboard coordinate system are established. The homogeneous transformation matrix is ​​expanded into a combination of a rotation matrix and a translation vector, and the Euler angles are obtained based on the rotation matrix. Ignoring the influence of the Z-axis dimension, the pose compensation amount of the 3-PRR parallel mechanism motion platform is finally calculated using the coordinate transformation formula in the two-dimensional plane. ; In the formula These are the coordinates of the moving platform in the actual static platform coordinate system. Translational compensation in the X and Y directions. This is the rotational compensation amount of the moving platform around the Z-axis. These are the Euler angles around the Z-axis under the ideal pose and the actual pose, respectively. in Used for calculating translational pose error in step C. Used for calculating rotational pose error in step C.

2. The positioning method for a high-load, high-precision mobile positioning robot according to claim 1, characterized in that, After all station positioning actions are completed, the positioning accuracy is tested using a laser tracker. The specific process includes: When the moving platform is in its actual pose, establish a rectangular coordinate system at its plane center and set target points on the edge line of the moving platform. Measured by a laser tracker The target points are determined by solving for their coordinates in the ideal pose coordinate system of the moving platform, based on the geometric relationships between the target points. The coordinates in the ideal pose coordinate system are then calculated, and the rotation angle of the actual pose coordinate system of the moving platform relative to the ideal pose coordinate system, as well as the coordinates of the origin of the actual pose coordinate system of the moving platform in the ideal pose coordinate system, are then calculated. Finally, the actual positioning accuracy of the mobile positioning robot is obtained. If the detection result does not meet the preset accuracy requirements, the process returns to step D to re-perform pose deviation compensation.

3. The positioning method for a high-load, high-precision mobile positioning robot according to claim 1, characterized in that: The mobile trolley (2) is equipped with a carrier body (6). The moving platform (1) is used to output the pose adjustment result; The planar adjustment module (5) is a 3-PRR parallel mechanism, including a planar branch group and a bullseye bearing (3). The planar branch group consists of three PRR-type branches, which are used to drive the moving platform (1) to perform X-axis and Y-axis translation and rotation around the Z-axis for posture adjustment; The bullseye bearing (3) is in point contact with the moving platform (1) and is laterally located outside the swing area of ​​the planar branch group, and is used to provide unrestrained vertical support and lateral guidance for the moving platform (1). The visual measurement module (4) is installed on the carrier vehicle body (6) and is used to acquire external target images and calculate the current pose information of the moving platform (1). The mobile positioning robot achieves high-precision positioning through the following operations: Collect images of the target at the workstation and calculate the pose compensation of the moving platform (1) relative to the theoretical reference pose. The calculated pose compensation amount is compared with the preset theoretical reference pose to obtain the pose error of X-axis and Y-axis translation and rotation around the Z-axis. Determine whether the pose error exceeds the set threshold. If it does, drive the plane adjustment module (5) to drive the moving platform (1) to perform pose compensation. After compensation, re-acquire the target image and repeat the calculation and difference judgment operation until the error converges to the threshold range. Once the error reaches the target, determine whether the current workstation is the final target workstation. If so, the positioning is completed; otherwise, set the coordinates of the next workstation as the new theoretical reference pose and continue with subsequent positioning operations.

4. The positioning method for a high-load, high-precision mobile positioning robot according to claim 3, characterized in that: The planar adjustment module (5) also includes a connecting plate (14) connected to the carrier body (6), and the planar branch group and the bullseye bearing (3) are located in the longitudinal space between the connecting plate (14) and the moving platform (1).

5. The positioning method for a high-load, high-precision mobile positioning robot according to claim 4, characterized in that: The planar branch group consists of a first branch (11), a second branch (12), and a third branch (13) with identical structures from the fixed end to the output end, used to realize translation in the X and Y directions and rotation around the Z axis.

6. The positioning method for a high-load, high-precision mobile positioning robot according to claim 5, characterized in that: Each branch includes a sliding joint (15) mounted on the connecting plate (14). The linear output part of the sliding joint (15) is equipped with a first rotating joint (16). The rotational output part of the first rotating joint (16) is connected to a connecting rod (17). The other end of the connecting rod (17) is equipped with a second rotating joint (18). The second rotating joint (18) is connected to the moving platform (1).

7. A positioning method for a high-load, high-precision mobile positioning robot according to claim 1 or 3, characterized in that: The vision measurement module (4) also includes a camera bracket (20) connected to the carrier body (6), on which an industrial camera (21) is mounted.

Images