Six-axis manipulator multi-segment path corner correction method
The camera obtains image information and hand-eye calibration, and uses Bezier curves to fit smooth turning paths and fine-tune them, solving the problem of jitter of the robot at the corners of multiple lines, improving the motion accuracy and path accuracy.
Patent Information
- Application Number
- CN202411395403.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-08
- Publication Date
- 2025-01-03
AI Technical Summary
The rigid turning of the robot at the corners of the multi-line segment causes the equipment to shake, affecting the motion accuracy.
The image information is obtained through the camera, the hand-eye calibration is performed, the control points in the robot coordinate system are determined, the smooth turning path is fitted using the Bezier curve, and the curve transition corners are fine-tuned by the gradient descent method to replace the original rigid straight turning path.
The smooth transition of the robot at the corner is achieved, avoiding jitter, improving motion accuracy and reducing path errors.
Smart Images
Figure CN119238504A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of manipulators, and in particular relates to a method for correcting corners of multi-segment paths of a six-axis manipulator. Background Art
[0002] The movement path planning of the robot is a key technical issue. It involves how to make the movement of the robot from the starting point to the target point both efficient and safe.
[0003] In the related technology, the robot makes rigid turns at the corners of multiple line segments without smoothing or deceleration at the turning position. The speed or direction of movement changes suddenly at a point within a very small distance. The sudden changes in speed and direction of movement will cause the robot arm to shake and affect the accuracy of the robot. Summary of the invention
[0004] The purpose of the present invention is to provide a method for correcting corners of multi-segment paths of a six-axis manipulator, so as to solve the problem in the prior art that rigid corners may cause the manipulator of the equipment to shake.
[0005] To this end, the present invention provides a method for correcting the corners of a multi-segment path of a six-axis manipulator, comprising: a manipulator and a camera installed on the manipulator, and comprising the following steps: performing hand-eye calibration on the camera and the manipulator; acquiring image information through the camera; determining the coordinates of the control points of the moving path of the manipulator in the manipulator coordinate system according to the image information; fitting a smooth turning path according to the coordinates of the control points, the turning point of the smooth turning path being a curved transition fillet.
[0006] Preferably, hand-eye calibration is performed by using the camera and the manipulator to obtain a conversion relationship between the camera coordinate system and the manipulator coordinate system.
[0007] Preferably, the image information is RGB image information or depth image information. RGB image has pixel information of three channels, and depth image has depth information, and point cloud can be extracted. Coordinates in the camera coordinate system can be obtained according to pixel information or point cloud.
[0008] Preferably, the control points include a starting point, an end point and several waypoints.
[0009] Preferably, the coordinates of the control points in the image information in the camera coordinate system are converted into coordinates in the manipulator coordinate system according to the conversion relationship obtained by hand-eye calibration.
[0010] Preferably, fitting a smooth turning path according to the coordinates of the control points includes smoothing the corners of the moving path by fitting a Bezier curve according to the coordinates of the waypoints.
[0011] Preferably, the Bezier curve fitting formula includes:
[0012] B(t)=(1-t) 3 P0+3(1-t) 3 tP1+3(1-t) 3 t 3 tP2+t 3 P3
[0013] Among them, B(t) is the fitted smooth turning path, P0 is the starting point, P1 and P2 are the passing points, P3 is the end point, t is the control coefficient, t∈[0,1].
[0014] Preferably, the method further includes fine-tuning the curve transition radius to reduce the error with the original path.
[0015] Preferably, the fine-tuning of the curve transition radius is to iteratively optimize the moving path by a gradient descent method.
[0016] Preferably, the gradient descent method includes uniformly selecting a number of calculation points on a smooth turning path, calculating the Euclidean distance between the control point and the calculation point, and constructing a correction function; calculating the gradient of the correction function for each calculation point; updating the position of the calculation point in the reverse direction of the gradient; and repeating iterative calculations until the movement of the control point is less than a threshold.
[0017] Beneficial effects:
[0018] 1. The present invention provides a method for correcting the corners of a multi-segment path of a six-axis manipulator, which automatically generates a smooth turning path based on the image information of a camera, and replaces the original rigid straight line turning path with the smooth turning path, so that the manipulator can make a smooth transition and avoid jitter caused by changes in the turning position and direction.
[0019] 2. The present invention fine-tunes the curve transition radius of the smooth turning path to make it closer to the original route, thereby ensuring the accuracy of the smooth turning path. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0021] Figure 1 The present invention is a flowchart of a method for correcting corners of a multi-segment path of a six-axis manipulator.
[0022] Figure 2 Schematic diagram of the moving path of the robot in the related technology.
[0023] Figure 3 It is a schematic diagram of a calibration coordinate system for a method for correcting corners of a multi-segment path of a six-axis manipulator in the present invention.
[0024] Figure 4 The figure is a schematic diagram of the moving path of a method for correcting corners of a multi-segment path of a six-axis manipulator in the present invention. DETAILED DESCRIPTION
[0025] The content of the present invention can be more easily understood by selecting the following detailed description of the preferred implementation method of the present invention and the embodiments included. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those of ordinary skill in the art to which the present invention belongs. When there is a conflict, the definition in this specification shall prevail.
[0026] Embodiment 1:
[0027] A robot and a camera mounted on the robot, the movement path of the robot is as follows Figure 2 As shown, the moving path is a rigid corner, which is prone to large jitter when turning, thus affecting the movement accuracy of the manipulator.
[0028] like Figure 1 As shown, this embodiment provides a method for correcting corners of a multi-segment path of a six-axis manipulator, comprising the following steps:
[0029] S1. Perform hand-eye calibration on the camera and the manipulator: Through the hand-eye calibration of the camera and the manipulator, the conversion relationship between the camera coordinate system and the manipulator coordinate system is obtained. The hand-eye calibration is performed through a calibration board.
[0030] like Figure 3 As shown in the figure, the base coordinate system is the base coordinate system of the manipulator, the manipulator coordinate system is the coordinate system of the end of the manipulator, the camera coordinate system is the coordinate system of the camera itself fixed on the manipulator, and the calibration object coordinate system is the coordinate system where the calibration plate is located. Move the manipulator arm twice arbitrarily. Since the calibration plate and the base of the manipulator arm are stationary, for a certain world point, its coordinate values in the base coordinate system and the calibration object coordinate system remain unchanged, while the coordinate values in the manipulator coordinate system and the camera coordinate system change with the movement of the manipulator arm. Based on this relationship, the transformation matrix between the manipulator coordinate system and the camera coordinate system can be solved, and then the transformation relationship between the machine coordinate system and the manipulator coordinate system can be obtained.
[0031] S2. Obtain image information through a camera; the image information is RGB image information or depth image information. RGB images have pixel information of three channels, and depth images have depth information, and point clouds can be extracted. Coordinates in the camera coordinate system can be obtained based on the pixel information or point clouds.
[0032] S3. Determine the coordinates of the control points of the moving path of the manipulator in the manipulator coordinate system according to the image information; the control points include a starting point, an end point, and a number of passing points. The coordinates of the control points in the image information in the camera coordinate system are converted into coordinates in the manipulator coordinate system according to the conversion relationship obtained by hand-eye calibration.
[0033] S4, fitting a smooth turning path according to the coordinates of the control points, where the turning point of the smooth turning path is a curved transition fillet. Figure 4 As shown, the corner of the moving path is smoothed by Bezier curve fitting. The formula of the Bezier curve fitting includes:
[0034] B(t)=(1-t) 3 P0+3(1-t) 3 tP1+3(1-t) 3 t 3 P2+t 3 P3
[0035] Among them, B(t) is the fitted smooth turning path, P0 is the starting point, P1 and P2 are the passing points, P3 is the end point, t is the control coefficient, t∈[0,1].
[0036] S5. Fine-tune the curve transition fillet to reduce the error with the original path; the fine-tune the curve transition fillet is to iteratively optimize the moving path through a gradient descent method.
[0037] The gradient descent method includes uniformly selecting a number of calculation points on a smooth turning path, calculating the Euclidean distance between the control point and the calculation point, and constructing a correction function; calculating the gradient of the correction function for each calculation point; updating the position of the calculation point in the reverse direction of the gradient; and repeating iterative calculations until the movement of the control point is less than a threshold.
[0038] Correction function J(R0,R1,…,R n ), where R0, R1, …, R n The correction function is used to measure the error between the curve and the target path, and the gradient is solved for each calculation point. gradient is a vector pointing to the direction where the function grows fastest. i Update the position of the calculated point along the reverse direction of the gradient, in The calculation point R i The updated position, α is the learning rate, which is used to control the amplitude of the control point update in each iteration. When , the iteration stops. ε is the tolerance.
[0039] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.
Claims
1. A method for correcting corners of a multi-segment path of a six-axis manipulator, comprising a manipulator and a camera mounted on the manipulator, characterized in that: The method comprises the following steps: performing hand-eye calibration on a camera and a manipulator; acquiring image information through a camera; determining the coordinates of control points of a moving path of the manipulator in a manipulator coordinate system according to the image information; fitting a smooth turning path according to the coordinates of the control points, wherein the turning point of the smooth turning path is a curve transition fillet.
2. A six-axis manipulator multi-segment path corner correction method according to claim 1, characterized in that: The camera and the manipulator are used to perform hand-eye calibration to obtain a conversion relationship between the camera coordinate system and the manipulator coordinate system.
3. A six-axis manipulator multi-segment path corner correction method according to claim 2, characterized in that: The image information is RGB image information or depth image information.
4. A six-axis manipulator multi-segment path corner correction method according to claim 3, characterized in that: The control points include a starting point, an end point and a number of waypoints.
5. A method for correcting corners of a multi-segment path of a six-axis manipulator according to claim 4, characterized in that: According to the conversion relationship obtained by hand-eye calibration, the coordinates of the control points in the image information in the camera coordinate system are converted into the coordinates in the manipulator coordinate system.
6. A method for correcting corners of a multi-segment path of a six-axis manipulator according to claim 4, characterized in that: The step of fitting a smooth turning path according to the coordinates of the control points includes smoothing the turning angle of the moving path by fitting a Bezier curve according to the coordinates of the waypoints.
7. A six-axis manipulator multi-segment path corner correction method according to claim 6, characterized in that: The formula for the Bezier curve fitting includes: B(t)=(1-t) 3 P0+3(1-t) 3 tP1+3(1-t) 3 t 3 P2+t 3 P3 Among them, B(t) is the fitted smooth turning path, P0 is the starting point, tP1 and tP2 are the passing points, tP3 is the end point, t is the control coefficient, t∈[0,1].
8. A method for correcting corners of a multi-segment path of a six-axis manipulator according to claim 1 or 7, characterized in that: It also includes fine-tuning the curve transition radius to reduce the error with the original path.
9. A method for correcting corners of a multi-segment path of a six-axis manipulator according to claim 8, characterized in that: The fine-tuning of the curve transition fillet is to iteratively optimize the moving path through a gradient descent method.
10. A method for correcting corners of a multi-segment path of a six-axis manipulator according to claim 9, characterized in that: The gradient descent method includes uniformly selecting a number of calculation points on a smooth turning path, calculating the Euclidean distance between the control point and the calculation point, and constructing a correction function; calculating the gradient of the correction function for each calculation point; and updating the position of the calculation point in the reverse direction of the gradient; The iterative calculation is repeated until the movement of the control point is less than the threshold.