A calibration method of a dual-robot collaborative machining detection system

Abstract

The application provides a kind of double robot cooperative processing detection system's calibration method, comprising: control robot movement, and with standard ball as calibration object, calibration the homogeneous transformation relationship between the flange plate coordinate system of each robot and vision sensor coordinate system;Control the rotation of turntable, and with standard ball as calibration object, calibration the homogeneous transformation relationship between the base coordinate system of two robots, and calibration the homogeneous transformation relationship between the base coordinate system of each robot and turntable coordinate system;Based on the homogeneous transformation relationship between the vision sensor coordinate system of each robot and base coordinate system, the homogeneous transformation relationship between the base coordinate system of two robots, obtain the relationship between the center coordinates of tool head and the base coordinate system of each robot.Such, closed loop is formed between each module calibration, so as to effectively reduce error accumulation, realize the full parameter closed calibration of double robot cooperative processing detection system, so that the accuracy of calibration result is improved, and compatibility is improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing, and specifically to a calibration method for a dual-robot collaborative processing and inspection system. Background Technology

[0002] Intelligent equipment with robots as actuators is widely used in fields such as artificial intelligence, mechanical manufacturing, and automatic control due to its advantages of flexible configuration, automation, and high efficiency. It has become a new trend in the development of technologies such as national defense, energy, and transportation. Taking the processing of aero-engine blades as an example, engines are thin-walled structures with low stiffness and complex curved surfaces. Therefore, related technologies use dual-robot collaborative processing systems to overcome the difficulties in the processing of thin-walled parts. Parameter calibration is an important prerequisite for ensuring the processing quality of the robot system.

[0003] However, the calibration of robot systems in related technologies has the following problems:

[0004] First, in related technologies, each module is usually calibrated separately, which can easily lead to cumulative errors in the calibration results, causing the calibration results to be distorted and making the overall calibration process complicated.

[0005] Second, the calibration methods for each module in the relevant technologies are limited. For example, using a monocular camera to perform hand-eye calibration of the robot system is difficult to apply to other modules of the robot system. Other devices are needed to achieve full parameter calibration of the system. Summary of the Invention

[0006] This invention provides a calibration method for a dual-robot collaborative processing and inspection system, which can improve the accuracy of calibration.

[0007] This invention discloses a calibration method for a dual-robot collaborative machining and inspection system. The machining and inspection system includes a first robot, a second robot, and a turntable located between the first robot and the second robot. Each robot is equipped with a tool head, and a standard ball is mounted on the turntable. The method includes:

[0008] The first calibration process involves controlling the robot's movement and using a standard sphere as the calibration object to calibrate the homogeneous transformation relationship between the flange coordinate system and the vision sensor coordinate system of each robot.

[0009] The second calibration process involves controlling the rotation of the turntable and using a standard ball as the calibration object to calibrate the homogeneous transformation relationship between the base coordinate systems of the two robots, as well as the homogeneous transformation relationship between the base coordinate systems of each robot and the turntable coordinate system.

[0010] The third calibration process involves obtaining the relationship between the center coordinates of the tool head and the base coordinates of each robot based on the homogeneous transformation relationship between the coordinate systems of each robot's vision sensor and the base coordinate system, as well as the homogeneous transformation relationship between the base coordinate systems of the two robots.

[0011] Optionally, the homogeneous transformation relation includes the rotation transformation matrix and the translation transformation matrix;

[0012] The first calibration process includes:

[0013] Control the robot to move n times, and obtain the rotation transformation matrix between the base coordinate system and the flange coordinate system after the first robot moves to the i-th position, following the order i from 1 to n. Translation transformation matrix In addition to the 3D data of the standard sphere acquired by the first vision sensor; and the rotation transformation matrix between the base coordinate system and the flange coordinate system after the second robot moves to the i-th position. Rotation transformation matrix And the three-dimensional data of the standard sphere acquired by the second vision sensor;

[0014] Based on the three-dimensional data of the standard sphere collected by the first vision sensor at the i-th position of the first robot, the value S1.i of the center coordinate of the standard sphere in the first vision sensor is obtained; and based on the three-dimensional data of the standard sphere collected by the second vision sensor at the i-th position of the second robot, the value S2.i of the center coordinate of the standard sphere in the second vision sensor is obtained.

[0015] Based on S1.i, Obtain the value B1.i of the sphere's center coordinates in the first robot base coordinate system:

[0016]

[0017] Among them, B1.i=B1.i′, i≠i′, i′=1, 2...n, Let be the rotation transformation matrix between the flange coordinate system of the first robot and the first vision sensor. Let be the translation transformation matrix between the flange coordinate system of the first robot and the first vision sensor;

[0018] as well as,

[0019] Based on S2.i, and Obtain the value B2.i of the sphere's center coordinates in the second robot base coordinate system.

[0020]

[0021] Where B2.i = B2.i′, The rotation transformation matrix between the flange coordinate system of the second robot and the second vision sensor. Let be the translation transformation matrix between the flange coordinate system of the second robot and the second vision sensor;

[0022] Based on B1.i calibration and And, based on B2.i calibration and

[0023] Optionally, n ≥ 6.

[0024] Optionally, the second calibration process includes:

[0025] Control the turntable to rotate m times, and obtain the rotation transformation matrix between the first robot base coordinate system and the flange coordinate system after the j-th rotation of the turntable, following the order j from 1 to m. Translation transformation matrix In addition to the 3D data of the standard sphere acquired by the first vision sensor; and the rotation transformation matrix between the second robot base coordinate system and the flange coordinate system. Rotation transformation matrix And the three-dimensional data of a standard sphere acquired by a second vision sensor; where j = 1, 2, ..., m;

[0026] Based on the three-dimensional data of the standard sphere collected by the first vision sensor after the j-th rotation of the turntable, the value of the center coordinate of the standard sphere in the first vision sensor is obtained as Q1.j; and based on the three-dimensional data of the standard sphere collected by the second vision sensor after the j-th rotation, the value of the center coordinate of the standard sphere in the second vision sensor is obtained as Q2.j.

[0027] based on and Obtain the center coordinates of the sphere in the first robot base coordinate system, J1.j, where J1.j satisfies:

[0028]

[0029] And, based on and Obtain the sphere center coordinates J2.j in the second robot base coordinate system, where J2.j satisfies:

[0030]

[0031] Where J2.j=J2j′, J1.j=J1j′, j≠j′, j′=1、2……m;

[0032] Using Relations Calibration in, Let be the rotation transformation matrix between the two robot base coordinate systems. Let be the translation transformation matrix between the two robot base coordinate systems.

[0033] Optionally, the second calibration process includes:

[0034] The first fitted circular curve is constructed based on J1.1 to J1.m, and the center of the first fitted circular curve is...

[0035] Constructing the Z-axis of the turntable coordinate system r Axis, Z r Axis passes through the center of the circle And it is perpendicular to the plane containing the first fitted circular curve;

[0036] Constructing the X coordinate system of the turntable r Axis, X r The axis is from the center of the circle The vector pointing to point T1.1, where T1.1 is the projection point of J1.1 onto the plane containing the fitted circular curve;

[0037] Constructing the Y-axis of the turntable coordinate system r axis, Y r Axis, X r Axis and Z r The axes are perpendicular to each other, and the Y axis is perpendicular to each other. r Axis, X r Axis and Z r The points where the axes intersect each other are the centers of the circles.

[0038] Obtain the value Z1.j of the center coordinates of the standard sphere in the turntable coordinate system, where Z1.j satisfies:

[0039]

[0040] Where θ is the angle of rotation of the turntable each time, mθ = 360°, ||O r1 J1.j|| is the radius of the first fitted circular curve;

[0041] According to the relation and Calibration in, Let be the rotation transformation matrix between the first robot base coordinate system and the turntable coordinate system. Let be the translation transformation matrix between the first robot base coordinate system and the turntable coordinate system.

[0042] Optionally, the second calibration process includes:

[0043] Based on the second fitted circular curves J2.1 to J2.m, the center of the second fitted circular curve is...

[0044] Constructing the Z-axis of the turntable coordinate system r Axis, Z r Axis passes through the center of the circle And it is perpendicular to the plane containing the fitted circular curve;

[0045] Constructing the X coordinate system of the turntable r Axis, X r Axis along the center of the circle The vector pointing to T2.1, where T2.1 is the projection point of J2.1 onto the plane containing the second fitted circular curve;

[0046] Constructing the Y-axis of the turntable coordinate system r axis, Y r Axis, X r Axis and Z r The axes are perpendicular to each other, and the Y axis is perpendicular to each other. r Axis, X r Axis and Z r The points where the axes intersect each other are the centers of the circles.

[0047] Obtain the value Z2.j of the center coordinates of the standard sphere in the turntable coordinate system, where Z2.j satisfies:

[0048]

[0049] Where θ is the angle of rotation of the turntable each time, mθ = 360°, ||O r2 J2.j|| is the radius of the second fitted circular curve;

[0050] According to the relation and Calibration and in, Let be the rotation transformation matrix between the second robot base coordinate system and the turntable coordinate system. This is the translation transformation matrix between the second robot base coordinate system and the turntable coordinate system.

[0051] Optionally, the tool head is a wheel-shaped machining head, which includes a side plane and an outer arc surface, wherein the intersection of the side plane and the outer arc surface is the inner circle, and the outermost curve of the outer arc surface is the outer circle;

[0052] The third calibration process includes:

[0053] Adjust the robot positions so that the side plane of the tool head on one robot corresponds to the vision sensor on the other robot;

[0054] Several points on the side plane of the acquisition tool head are used as the first acquisition points;

[0055] Fit the first acquisition point to form a fitting plane;

[0056] Several points on the outer arc surface of the acquisition tool head are used as the second acquisition points;

[0057] Measure the distance between each second acquisition point and the fitting plane, and fit the second acquisition points with the largest distance from the fitting plane to form a third fitted circular curve;

[0058] The center coordinates of the tool head are obtained based on the third fitted circular curve;

[0059] Based on the homogeneous transformation relationship between the coordinate systems of each robot's vision sensor and the base coordinate system, and the homogeneous transformation relationship between the base coordinate systems of the two robots, the relationship between the center coordinates of the tool head and the base coordinate systems of each robot is obtained.

[0060] Optionally, the tool head is a spherical machining head.

[0061] The third calibration process includes:

[0062] Adjust the robot positions so that the spherical surface of the tool head on one robot corresponds to the vision sensor on the other robot;

[0063] Several points on the spherical surface of the acquisition tool head are used as the third acquisition points;

[0064] The third acquisition point is fitted to form a fitted sphere;

[0065] The center coordinates of the toolhead are obtained based on the fitted sphere.

[0066] Based on the homogeneous transformation relationship between the coordinate systems of each robot's vision sensor and the base coordinate system, and the homogeneous transformation relationship between the base coordinate systems of the two robots, the relationship between the center coordinates of the tool head and the base coordinate systems of each robot is obtained.

[0067] Optionally, between the second and third calibration steps, the calibration method may further include: removing the standard ball.

[0068] Optionally, after performing the third calibration step, the calibration method further includes: removing the first vision sensor or the second vision sensor; wherein the first robot is equipped with the first vision sensor and the second robot is equipped with the second vision sensor.

[0069] The beneficial effects of this invention are as follows:

[0070] This invention discloses a calibration method for a dual-robot collaborative machining and inspection system. The machining and inspection system includes a first robot, a second robot, and a turntable located between the first robot and the second robot. Each robot is equipped with a tool head, and a standard ball is mounted on the turntable. The method includes:

[0071] The first calibration process involves controlling the robot's movement and using a standard sphere as the calibration object to calibrate the homogeneous transformation relationship between the flange coordinate system and the vision sensor coordinate system of each robot.

[0072] The second calibration process involves controlling the rotation of the turntable and using a standard ball as the calibration object to calibrate the homogeneous transformation relationship between the base coordinate systems of the two robots, as well as the homogeneous transformation relationship between the base coordinate systems of each robot and the turntable coordinate system.

[0073] The third calibration process involves obtaining the relationship between the center coordinates of the tool head and the base coordinates of each robot based on the homogeneous transformation relationship between the coordinate systems of each robot's vision sensor and the base coordinate system, as well as the homogeneous transformation relationship between the base coordinate systems of the two robots.

[0074] In summary, the calibration method of the present invention has the following advantages:

[0075] First, in the dual-robot collaborative processing and inspection system of this application, the calibration of each module forms a closed loop. Specifically, taking the relevant calibration of the first robot 110 as an example, in the first calibration process, the rotation transformation matrix between the base coordinate system and the flange coordinate system of the first robot 110 is used. Translation transformation matrix To calibrate the rotation transformation matrix between the flange coordinate system of the first robot 110 and the first vision sensor 410. Translation transformation matrix In the second calibration process, the rotation transformation matrix between the flange coordinate system of the first robot 110 and the first vision sensor 410, which was calibrated in the first calibration process, is... Translation transformation matrix Solve for the value J1.j of the sphere center coordinates in the first robot's 110 base coordinate system, and then use J1.j to calibrate the rotation transformation matrix between the two robot base coordinate systems. Translation transformation matrix And to calibrate the rotation transformation matrix between the first robot's 110 base coordinate system and the turntable coordinate system. Translation transformation matrix In the third calibration process, the rotation transformation matrix between the first robot 110 base coordinate system and the flange coordinate system obtained in the first calibration process is... Translation transformation matrix and the rotation transformation matrix between the flange coordinate system of the first robot 110 and the first vision sensor 410. Translation transformation matrix Calculate the rotation transformation matrix between the vision sensor coordinate system and the base coordinate system of the first robot 110. Translation transformation matrix Combined with the rotation transformation matrix between the two robot base coordinate systems already calibrated in the second calibration process. Translation transformation matrix The homogeneous transformation relationship between the center coordinates of the first tool head 510 and the base coordinate system of the first robot 110 is obtained. This effectively reduces error accumulation, achieves full-parameter closed-loop calibration of the dual-robot collaborative processing and inspection system, improves the accuracy of calibration results, and simplifies the overall calibration process.

[0076] Second, the calibration method of this application can be effectively applied to various modules of the robot system, improving compatibility. Attached Figure Description

[0077] The above-described features and advantages of the present invention will be better understood after reading the following detailed description of embodiments of the present disclosure in conjunction with the accompanying drawings. In the drawings, components are not necessarily drawn to scale, and components having similar related characteristics or features may have the same or similar reference numerals.

[0078] Figure 1 This is a flowchart of the process of this invention;

[0079] Figure 2 This is a schematic diagram of the dual-robot collaborative processing and inspection system of the present invention;

[0080] Figure 3 This is a schematic diagram of the coordinate system calibration of the turntable in this invention;

[0081] Figure 4 This is a schematic diagram of the calibration of the wheel-shaped machining tool head of the present invention;

[0082] Figure 5 This is a schematic diagram of the calibration of the spherical machining tool head of the present invention.

[0083] Explanation of reference numerals in the attached figures:

[0084] 110 - First Robot, 120 - Second Robot

[0085] 200-turntable

[0086] 300-standard ball,

[0087] 410 - First visual sensor, 420 - Second visual sensor

[0088] 510 - First tool head, 520 - Second tool head, 501 - Side plane, 502 - Outer arc surface, 503 - Inner ring, 504 - Outer ring, 505 - Spherical surface

[0089] 610 - First mounting component, 620 - Second mounting component

[0090] 700 - Fastener. Detailed Implementation

[0091] The present invention will be further described below with reference to specific embodiments and accompanying drawings. More details are set forth in the following description in order to provide a full understanding of the present invention. However, the present invention can obviously be implemented in many other ways different from those described herein. Those skilled in the art can make similar extensions and derivations based on actual application situations without departing from the spirit of the present invention. Therefore, the scope of protection of the present invention should not be limited by the content of this specific embodiment.

[0092] It should be noted that these and other accompanying drawings are merely examples and are not drawn to scale, and should not be construed as limiting the scope of protection of the present invention.

[0093] The following is combined Figures 1-5 The technical solution of this application is described.

[0094] Figure 1 This is a flowchart of the process of this invention; Figure 2 This is a schematic diagram of the dual-robot collaborative processing and inspection system of the present invention. As shown in the figure, this application discloses a calibration method for a dual-robot collaborative processing and inspection system, wherein the dual-robot collaborative processing and inspection system includes a first robot 110, a second robot 120, and a turntable 200. The turntable 200 is located between the first robot 110 and the second robot 120. A first tool head 510 is provided on the first robot 110, and a second tool head 520 is provided on the second robot 120. The first tool head 510 and the second tool head 520 can be the same or different, for example, they can be set as follows: Figure 4 The wheel-shaped tool head shown Figure 5 The spherical tool head shown is an example. Calibration methods for machining and inspection systems include:

[0095] Step 1

[0096] Performing pre-installation procedures includes:

[0097] A standard ball 300 is mounted on the turntable 200 via a fixture 700, and vision sensors are mounted on each robot.

[0098] The fixing component 700 can be a clamp, clamp, or other device for fixing the standard sphere 300. The vision sensor is a surface structured light projection vision sensor, which projects light of a specific structure onto the object being photographed through an infrared laser. Then, an infrared camera collects the reflected structured light pattern and calculates depth information based on the principle of triangulation. The surface structured light projection vision sensor does not depend on the object's color and texture, and can achieve fast and accurate feature point matching, making it suitable for 3D imaging and reconstruction.

[0099] In the dual-robot collaborative processing and inspection system of this application, one vision sensor is included within the processing and inspection system itself, while the other is additionally installed to implement the calibration method of this invention. For example, the first vision sensor 410 installed on the first robot 110 is included in the system itself, while the second vision sensor 420 installed on the second robot 120 is additionally installed for calibration; or the second vision sensor 420 installed on the second robot 120 is included in the system itself, while the first vision sensor 410 installed on the first robot 110 is additionally installed for calibration. The first vision sensor 410 is installed on the first robot 110 via a first mounting member 610, and the second vision sensor 420 is installed on the second robot 120 via a second mounting member 620. The mounting members can be fixed to the robots before calibration. After calibration, one of the surface structured light projection vision sensors can be retained or removed as needed without disassembling the mounting members. Therefore, this has no impact on the dual-robot collaborative processing and inspection system itself, improving system flexibility and equipment utilization.

[0100] Step 2

[0101] The first calibration procedure includes: controlling the robot's movement and, using a standard sphere 300 as the calibration object, calibrating the homogeneous transformation relationship between the flange coordinate system and the vision sensor coordinate system of each robot. The homogeneous transformation relationship includes rotation transformation matrices and translation transformation matrices. The specific details of the first calibration procedure are as follows:

[0102] Step 110: Control the robot to move n times, ensuring that the standard ball 300 is within the measurement range of the first vision sensor 410 and the second vision sensor 420 after each movement; during the movement, obtain the rotation transformation matrix between the base coordinate system and the flange coordinate system of the first robot 110 after it moves to the i-th position, following the order i from 1 to n. Translation transformation matrix The data includes the three-dimensional data of the standard sphere 300 acquired by the first vision sensor 410; and the rotation transformation matrix between the base coordinate system and the flange coordinate system of the second robot 120 after it moves to the i-th position. Rotation transformation matrix and the three-dimensional data of the standard sphere 300 acquired by the second vision sensor 420; among which, and and The three-dimensional data of the standard sphere 300 are recorded in the robot's control system for easy access at any time.

[0103] Step 120: Based on the three-dimensional data of the standard sphere 300 collected by the first vision sensor 410 at the i-th position of the first robot 110, fit the spherical equation of the standard sphere 300 corresponding to the three-dimensional data, and then obtain the value S1.i of the center coordinates of the standard sphere 300 in the first vision sensor 410 according to the spherical equation at the i-th position; and based on the three-dimensional data of the standard sphere 300 collected by the second vision sensor 420 at the i-th position of the second robot 120, fit the spherical equation of the standard sphere 300 corresponding to the three-dimensional data, and then obtain the value S2.i of the center coordinates of the standard sphere 300 in the second vision sensor 420 according to the spherical equation at the i-th position.

[0104] Step S130: Based on the value S1.i of the center coordinates of the standard sphere 300 in the first vision sensor 410, and the rotation transformation matrix between the base coordinate system and the flange coordinate system of the first robot 110. Translation transformation matrix Obtain the value B1.i of the sphere's center coordinates in the first robot's 110 base coordinate system, where B1.i satisfies:

[0105]

[0106] Among them, B1.i=B1.i′, i≠i′, i, i′=1, 2...n, Let be the rotation transformation matrix between the flange coordinate system of the first robot 110 and the first vision sensor 410. B1.i is the translation transformation matrix between the flange coordinate system of the first robot 110 and the first vision sensor 410; and The values ​​do not change with the position of the first robot 110.

[0107] Similarly, the value S2.i of the center coordinates of the standard sphere 300 in the second vision sensor 420, and the rotation transformation matrix between the base coordinate system and the flange coordinate system of the second robot 120, are also considered. Rotation transformation matrix Obtain the value B2.i of the sphere's center coordinates in the second robot's 120 base coordinate system, where B2.i satisfies:

[0108]

[0109] Where B2.i = B2.i′, The rotation transformation matrix between the flange coordinate system of the second robot 120 and the second vision sensor 420. B2.i is the translation transformation matrix between the flange coordinate system of the second robot 120 and the second vision sensor 420. and The values ​​do not change with the position change of the second robot 120.

[0110] Step S240: Calibration based on B1.i and And, based on B2.i calibration and The following is based on B1.i calibration. and For example, the specific details are as follows:

[0111] From B1.i = B1.i′, i ≠ i′, i, i′ = 1, 2, ..., n, we can obtain An equation, for example:

[0112] B1.1=B1.2, B1.1=B1.3, B1.1=B1.4...B1.1=B1.n, B1.2=B1.3, B1.2=B1.4...B1.2=B1.n...B1.n-1=B1.n, etc. and and There are a total of 7 unknowns, for example The following can be represented using quaternions:

[0113]

[0114] Therefore, the number of robot movements n mentioned above should satisfy: n≥6. Then, the Lagrange operator method is used for constraint optimization, and finally, the calibration is performed. and

[0115] Similarly,

[0116] From B2.i=B2.i′, i≠i′, i, i′=1, 2...n, we can get An equation, for example:

[0117] B2.1=B2.2, B2.1=B2.3, B2.1=B2.4...B2.1=B2.n, B2.2=B2.3, B2.2=B2.4...

[0118] B2.2 = B2.n...B2.n-1 = B2.n, etc. And... and There are 7 unknowns in total. Then, the Lagrange operator method is used for constraint optimization, and the final result can be calibrated. and

[0119] Step 3

[0120] After the first calibration step is completed, the second calibration step is performed: The turntable 200 is rotated, and using the standard sphere 300 as the calibration object, the homogeneous transformation relationship between the base coordinate systems of the two robots, as well as the homogeneous transformation relationship between the base coordinate systems of each robot and the turntable coordinate system, are calibrated. The specific details of the second calibration step are as follows:

[0121] Step S200: Calibrate the homogeneous transformation relationship between the two robot base coordinate systems, specifically including:

[0122] Step S210: Control the turntable 200 to rotate m times. During the rotation, control the turntable 200 to rotate at the same angle θ each time to ensure the calibration accuracy. At the same time, control θ ≤ 30° to ensure sufficient calibration points and further ensure calibration accuracy. In addition, control that after each rotation, the standard ball 300 is within the measurement range of the first vision sensor 410 and the second vision sensor 420.

[0123] During the rotation, following the sequence j from 1 to m, the rotation transformation matrix between the base coordinate system of the first robot 110 and the flange coordinate system after the j-th rotation of the turntable 200 is obtained. Translation transformation matrix The system also acquires the 3D data of the standard sphere 300 collected by the first vision sensor 410; and obtains the rotation transformation matrix between the base coordinate system of the second robot 120 and the flange coordinate system after the j-th rotation of the turntable 200. Rotation transformation matrix And the three-dimensional data of the standard sphere 300 collected by the second vision sensor 420; where j = 1, 2...m; m = 360° / θ; Both the three-dimensional data of the standard sphere 300 and the standard sphere 300 are recorded in the robot's control system.

[0124] Step S220: Based on the three-dimensional data of the standard sphere 300 collected by the first vision sensor 410 after the j-th rotation of the turntable 200, fit the spherical equation of the standard sphere 300 corresponding to the three-dimensional data, and then obtain the value Q1.j of the center coordinate of the standard sphere 300 in the first vision sensor 410 based on the spherical equation of the standard sphere 300 after the j-th rotation.

[0125] Similarly, based on the three-dimensional data of the standard sphere 300 collected by the second vision sensor 420 after the j-th rotation, the spherical equation of the standard sphere 300 corresponding to the three-dimensional data is fitted, and then based on the spherical equation of the standard sphere 300 after the j-th rotation, the value Q2.j of the center coordinate of the standard sphere 300 in the second vision sensor 420 is obtained.

[0126] Step S230: Based on the rotation transformation matrix between the flange coordinate system of the first robot 110 and the first vision sensor 410, which has been calibrated in the first calibration process. Translation transformation matrix And the rotation transformation matrix between the first robot's 110 base coordinate system and the flange coordinate system after the j-th rotation of the turntable 200. Translation transformation matrix Obtain the center coordinates of the sphere in the 110 base coordinate system of the first robot, J1.j, where J1.j satisfies:

[0127]

[0128] Where J1.j=J1.j′, j≠j′, j=1、2......m, j′=1、2......m.

[0129] Similarly, the rotation transformation matrix between the flange coordinate system of the second robot 120 and the first vision sensor 410, which has already been calibrated in the first calibration process, is also used. Translation transformation matrix And the rotation transformation matrix based on the second robot's 120 base coordinate system and flange coordinate system. Translation transformation matrix Obtain the center coordinates of the sphere in the 120 base coordinate system of the second robot, J2.j, where J2.j satisfies:

[0130]

[0131] Where J2.j=J2.j′, j≠j′, j=1、2......m, j′=1、2......m.

[0132] Step S240: Using the relation We can obtain m equations, which are as follows:

[0133]

[0134] Based on these equations, it is possible to define in, Let be the rotation transformation matrix between the two robot base coordinate systems. Let be the translation transformation matrix between the two robot base coordinate systems.

[0135] After obtaining the ball center coordinates J1.j in the base coordinate system of the first robot 110 and J2.j in the base coordinate system of the second robot 120 in step S230, the second calibration process further includes steps B200 and M200. Step B200 involves calibrating the homogeneous transformation relationship between the base coordinate system of the first robot 110 and the turntable coordinate system; step M200 involves calibrating the homogeneous transformation relationship between the base coordinate system of the second robot 120 and the turntable coordinate system.

[0136] Specifically, step B200 includes:

[0137] Step B210: As Figure 3 As shown, a first fitted circular curve is constructed based on J1.1 to J1.m, with the center of the first fitted circular curve being...

[0138] Step B220: Construct the Z-axis of the turntable coordinate system r Axis, Z r Axis passes through the center of the circle And it is perpendicular to the plane containing the first fitted circular curve.

[0139] Step B230: Construct the X-axis of the turntable coordinate system r Axis, X r The axis is from the center of the circle The vector pointing to point T1.1, where T1.1 is the projection point of J1.1 onto the plane containing the fitted circular curve.

[0140] Step B240: Construct the Y-axis of the turntable coordinate system r axis, Y r Axis, X r Axis and Z r The axes are perpendicular to each other, and the Y axis is perpendicular to each other. r Axis, X r Axis and Z r The points where the axes intersect each other are the centers of the circles.

[0141] Step B250: Obtain the value Z1.j of the center coordinates of the standard sphere 300 in the turntable coordinate system, specifically as follows: Figure 3 As shown, Z1.j satisfies:

[0142]

[0143] Where θ is the angle of rotation of the turntable 200 in each rotation, mθ = 360°, ||O r1 J10.j|| is the radius of the first fitted circular curve.

[0144] Step B260: Based on the relation and We can construct m equations, setting the first recorded position to a 0° angle. The m equations are then:

[0145]

[0146] These equations can be used to determine... in, Let be the rotation transformation matrix between the 110 base coordinate system of the first robot and the turntable coordinate system. This is the translation transformation matrix between the 110 base coordinate system and the turntable coordinate system of the first robot.

[0147] Step M200 specifically includes:

[0148] Step M210: Based on the second fitted circular curves J2.1 to J2.m, the center of the second fitted circular curve is...

[0149] Step M220: Construct the Z-axis of the turntable coordinate system r Axis, Z r Axis passes through the center of the circle And it is perpendicular to the plane containing the fitted circular curve.

[0150] Step M230: Construct the X coordinate system of the turntable r Axis, X r Axis along the center of the circle The vector pointing to T2.1, where T2.1 is the projection point of J2.1 onto the plane containing the second fitted circular curve.

[0151] Step M240: Construct the Y-axis of the turntable coordinate system r axis, Y r Axis, X r Axis and Z r The axes are perpendicular to each other, and the Y axis is perpendicular to each other. r Axis, X r Axis and Z r The points where the axes intersect each other are the centers of the circles.

[0152] Step M250: Obtain the value Z2.j of the center coordinates of the standard sphere 300 in the turntable coordinate system. Z2.j satisfies:

[0153]

[0154] Where θ is the angle of rotation of the turntable 200 in each rotation, mθ = 360°, ||O r2 J2.j|| is the radius of the second fitted circular curve.

[0155] Step M260: Based on the relation and We can construct m equations, setting the first recorded position to a 0° angle, and the m equations are as follows:

[0156]

[0157] These equations can be used to determine... and in, Let be the rotation transformation matrix between the 120-base coordinate system of the second robot and the turntable coordinate system. This is the translation transformation matrix between the 120 base coordinate system and the turntable coordinate system of the second robot.

[0158] Step 4:

[0159] Remove the standard ball 300; After the second calibration procedure is completed, the standard ball 300 is no longer needed in subsequent work, so it can be removed to avoid interfering with subsequent calibration work and use.

[0160] Step 5:

[0161] The third calibration process includes: obtaining the relationship between the center coordinates of the tool head and the base coordinate systems of each robot, based on the homogeneous transformation relationship between the coordinate systems of each robot's vision sensor and the base coordinate system, and the homogeneous transformation relationship between the base coordinate systems of the two robots; wherein, the tool head set on the first robot 110 is the first tool head 510, and the tool head set on the second robot 120 is the second tool head 520. The specific details of the third calibration process are as follows:

[0162] If the tool head is Figure 4 The wheel-shaped machining head shown, i.e., the tool head, includes a side plane 501 and an outer arc surface 502. The intersection of the side plane 501 and the outer arc surface 502 is the inner ring 503, and the outermost curve of the outer arc surface 502 is the outer ring 504. Taking the first tool head 510 as an example, the third calibration process includes:

[0163] Step S310: Combining Figure 1 and Figure 4 As shown, the robot position is adjusted so that the side plane 501 of the first tool head 510 on the first robot 110 corresponds to the second vision sensor 420 on the second robot 120.

[0164] Step S320: The second vision sensor 420 collects several points on the side plane 501 of the first tool head 510 as the first collection points. In order to avoid collection errors, points located outside the inner circle 503 need to be excluded to ensure the subsequent fitting accuracy.

[0165] Step S330: Fit the first acquisition point to form a fitting plane, and the fitting plane corresponds to the side plane 501.

[0166] Step S340: Collect several points on the outer arc surface 502 of the first tool head 510 as the second collection points.

[0167] Step S350: Measure the distance between each second acquisition point and the fitting plane, and fit the second acquisition points with the largest distance from the fitting plane to form a third fitting circular curve, which corresponds to the outer circle 504.

[0168] Step S360: Obtain the center coordinates of the first tool head 510 based on the third fitted circular curve.

[0169] Step S370: Obtain the homogeneous transformation relationship between the vision sensor coordinate system and the base coordinate system of the first robot 110, including:

[0170] The rotation transformation matrix between the first robot's 110 base coordinate system and the flange coordinate system Translation transformation matrix and the rotation transformation matrix between the flange coordinate system of the first robot 110 and the first vision sensor 410. Translation transformation matrix The rotation transformation matrix between the vision sensor coordinate system and the base coordinate system of the first robot 110 was calculated. Translation transformation matrix The homogeneous transformation relationship between the vision sensor coordinate system and the base coordinate system of the first robot 110 includes: and

[0171] Step S380: Based on the homogeneous transformation relationship between the vision sensor coordinate system and the base coordinate system of the first robot 110, including the rotation transformation matrix between the vision sensor coordinate system and the base coordinate system of the first robot 110. Translation transformation matrix Furthermore, based on the homogeneous transformation relationship between the two robot base coordinate systems, this includes: the rotation transformation matrix between the two robot base coordinate systems. Translation transformation matrix between two robot base coordinate systems Obtain the homogeneous transformation relationship between the center coordinates of the first tool head 510 and the base coordinate system of the first robot 110.

[0172] Similarly, a third calibration process can be performed on the second tool head 520, specifically as follows:

[0173] Step S310′: Adjust the robot position so that the second tool head 520 on the second robot 120 corresponds to the first vision sensor 410 on the first robot 110.

[0174] Step S320': The first vision sensor 410 collects several points on the side plane 501 of the second tool head 520 as the first collection points.

[0175] Step S330′: Fit the first acquisition point to form a fitting plane.

[0176] Step S340′: Collect several points on the outer arc surface 502 of the second tool head 520 as the second collection points.

[0177] Step S350′: Measure the distance between each second acquisition point and the fitting plane, and fit the second acquisition points with the largest distance from the fitting plane to form a third fitting circular curve, which corresponds to the outer circle 504.

[0178] Step S360′: Obtain the center coordinates of the second tool head 520 based on the third fitted circular curve.

[0179] Step S370': Obtain the homogeneous transformation relationship between the second robot 120 vision sensor coordinate system and the base coordinate system, including:

[0180] The rotation transformation matrix between the second robot's 120 base coordinate system and the flange coordinate system Translation transformation matrix and the rotation transformation matrix between the flange coordinate system of the second robot 120 and the second vision sensor 420. Translation transformation matrix The rotation transformation matrix between the second robot's 120 vision sensor coordinate system and the base coordinate system was calculated. The homogeneous transformation relationship between the vision sensor coordinate system and the base coordinate system of the second robot 120 includes: and

[0181] Step S380': Based on the homogeneous transformation relationship between the vision sensor coordinate system and the base coordinate system of the second robot 120, including the rotation transformation matrix between the vision sensor coordinate system and the base coordinate system of the second robot 120. Translation transformation matrix Furthermore, based on the homogeneous transformation relationship between the two robot base coordinate systems, this includes: the rotation transformation matrix between the two robot base coordinate systems. Translation transformation matrix between two robot base coordinate systems Obtain the homogeneous transformation relationship between the center coordinates of the second tool head 520 and the base coordinate system of the second robot 120.

[0182] This completes all the calibrations for this application. Based on this information, the position and orientation of the tool head can be controlled during processing.

[0183] If the tool head is Figure 5 The spherical machining head shown, taking the first tool head 510 as an example, includes the following third calibration process:

[0184] Step B310: Combining Figure 1 and Figure 5 As shown, the robot position is adjusted so that the spherical surface 505 of the first tool head 510 on the first robot 110 corresponds to the second vision sensor 420 on the second robot 120.

[0185] Step B320: Collect several points on the spherical surface 505 of the first tool head 510 as the third collection points.

[0186] Step B330: Fit the third acquisition point to form a fitted sphere, which corresponds to the spherical surface 505.

[0187] Step B340: Obtain the center coordinates of the first tool head 510 based on the fitted sphere.

[0188] Step B350: Based on the homogeneous transformation relationship between the vision sensor coordinate system and the base coordinate system of the first robot 110; and based on the homogeneous transformation relationship between the two robot base coordinate systems, including: the rotation transformation matrix between the two robot base coordinate systems. Translation transformation matrix between two robot base coordinate systems Obtain the homogeneous transformation relationship between the center coordinates of the first tool head 510 and the base coordinate system of the first robot 110.

[0189] Similarly, a third calibration process can be performed on the second tool head 520, including:

[0190] Step B310: Adjust the robot position so that the second tool head 520 on the second robot 120 corresponds to the first vision sensor 410 on the first robot 110.

[0191] Step B320: Collect several points on the spherical surface 505 of the second tool head 520 as the third collection points.

[0192] Step B330: Fit the third acquisition point to form a fitted sphere, with spherical surface 505 corresponding to spherical surface 505.

[0193] Step B340: Obtain the center coordinates of the tool head based on the fitted sphere.

[0194] Step B350: Based on the homogeneous transformation relationship between the second robot 120's vision sensor coordinate system and the base coordinate system; and, based on the homogeneous transformation relationship between the two robot base coordinate systems, including: the rotation transformation matrix between the two robot base coordinate systems. Translation transformation matrix between two robot base coordinate systems Obtain the homogeneous transformation relationship between the center coordinates of the second tool head 520 and the base coordinate system of the second robot 120.

[0195] Step 6:

[0196] After all calibrations are completed, remove the first vision sensor 410 or the second vision sensor 420; leaving one is sufficient to meet the product inspection needs of the dual-robot collaborative processing system, increasing system flexibility and equipment utilization.

[0197] In summary, the calibration method of the present invention has the following advantages:

[0198] First, in the dual-robot collaborative processing and inspection system of this application, the calibration of each module forms a closed loop. Specifically, taking the relevant calibration of the first robot 110 as an example, in the first calibration process, the rotation transformation matrix between the base coordinate system and the flange coordinate system of the first robot 110 is used. Translation transformation matrix To calibrate the rotation transformation matrix between the flange coordinate system of the first robot 110 and the first vision sensor 410. Translation transformation matrix In the second calibration process, the rotation transformation matrix between the flange coordinate system of the first robot 110 and the first vision sensor 410, which was calibrated in the first calibration process, is... Translation transformation matrix Solve for the value J1.j of the sphere center coordinates in the first robot's 110 base coordinate system, and then use J1.j to calibrate the rotation transformation matrix between the two robot base coordinate systems. Translation transformation matrix And to calibrate the rotation transformation matrix between the first robot's 110 base coordinate system and the turntable coordinate system. Translation transformation matrix In the third calibration process, the rotation transformation matrix between the first robot 110 base coordinate system and the flange coordinate system obtained in the first calibration process is... Translation transformation matrix and the rotation transformation matrix between the flange coordinate system of the first robot 110 and the first vision sensor 410. Translation transformation matrix Calculate the rotation transformation matrix between the vision sensor coordinate system and the base coordinate system of the first robot 110. Translation transformation matrix Combined with the rotation transformation matrix between the two robot base coordinate systems already calibrated in the second calibration process. Translation transformation matrix The homogeneous transformation relationship between the center coordinates of the first tool head 510 and the base coordinate system of the first robot 110 is obtained. This effectively reduces error accumulation, achieves full-parameter closed-loop calibration of the dual-robot collaborative processing and inspection system, improves the accuracy of calibration results, and simplifies the overall calibration process.

[0199] Second, the calibration method of this application can be effectively applied to various modules of the robot system, improving compatibility.

[0200] While the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the invention. Any variations and modifications can be made by those skilled in the art without departing from the spirit and scope of the invention. Therefore, any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention, without departing from the scope of the invention, fall within the protection scope defined by the claims of the present invention.

Claims

1. A calibration method for a dual-robot collaborative processing and inspection system, characterized in that, The processing and inspection system includes a first robot (110), a second robot (120), and a turntable (200) located between the first robot (110) and the second robot (120). Each robot is equipped with a tool head, and a standard ball (300) is mounted on the turntable (200). The method includes: First calibration process: Control the robot's movement and use the standard ball (300) as the calibration object to calibrate the homogeneous transformation relationship between the flange coordinate system and the vision sensor coordinate system of each robot; The second calibration process is to control the rotation of the turntable (200) and use the standard ball (300) as the calibration object to calibrate the homogeneous transformation relationship between the base coordinate systems of the two robots, as well as the homogeneous transformation relationship between the base coordinate systems of each robot and the turntable coordinate system. The third calibration process involves obtaining the relationship between the center coordinates of the tool head and the base coordinates of each robot based on the homogeneous transformation relationship between the coordinate systems of each robot's vision sensor and the base coordinate system, as well as the homogeneous transformation relationship between the base coordinate systems of the two robots.

2. The calibration method according to claim 1, characterized in that, The homogeneous transformation relationship includes rotation transformation matrix and translation transformation matrix; The first calibration process includes: Control the robot to move n times, and obtain the rotation transformation matrix between the base coordinate system and the flange coordinate system after the first robot (110) moves to the i-th position, following the order i from 1 to n. Translation transformation matrix and the three-dimensional data of the standard sphere (300) acquired by the first vision sensor (410); and the rotation transformation matrix between the base coordinate system and the flange coordinate system after the second robot (120) moves to the i-th position. Rotation transformation matrix and the three-dimensional data of the standard sphere (300) acquired by the second vision sensor (420); Based on the three-dimensional data of the standard sphere (300) collected by the first vision sensor (410) at the i-th position of the first robot (110), the value S1.i of the center coordinates of the standard sphere (300) in the first vision sensor (410) is obtained; and based on the three-dimensional data of the standard sphere (300) collected by the second vision sensor (420) at the i-th position of the second robot (120), the value S2.i of the center coordinates of the standard sphere (300) in the second vision sensor (420) is obtained. Based on S1.i, Obtain the value B1.i of the sphere center coordinates in the base coordinate system of the first robot (110): Among them, B1.i=B1.i′, i≠i′, i′=1, 2...n, Let be the rotation transformation matrix between the flange coordinate system of the first robot (110) and the first vision sensor (410). Let be the translation transformation matrix between the flange coordinate system of the first robot (110) and the first vision sensor (410); as well as, Based on S2.i, and Obtain the value B2.i of the sphere center coordinates in the base coordinate system of the second robot (120). Where B2.i = B2.i′, The rotation transformation matrix between the flange coordinate system of the second robot (120) and the second vision sensor (420) is given. Let be the translation transformation matrix between the flange coordinate system of the second robot (120) and the second vision sensor (420); Based on B1.i calibration and And, based on B2, i calibration and 3. The calibration method according to claim 2, characterized in that, n≥6。 4. The calibration method according to claim 2, characterized in that, The second calibration process includes: Control the turntable (200) to rotate m times, and obtain the rotation transformation matrix between the base coordinate system of the first robot (110) and the flange coordinate system after the j-th rotation of the turntable (200) in the order of j from 1 to m. Translation transformation matrix and the three-dimensional data of the standard sphere (300) acquired by the first vision sensor (410); and the rotation transformation matrix between the base coordinate system and the flange coordinate system of the second robot (120). Rotation transformation matrix and the three-dimensional data of the standard sphere (300) acquired by the second vision sensor (420); where j = 1, 2...m; Based on the three-dimensional data of the standard sphere (300) collected by the first vision sensor (410) after the j-th rotation of the turntable (200), the value Q1.j of the center coordinates of the standard sphere (300) in the first vision sensor (410) is obtained; and based on the three-dimensional data of the standard sphere (300) collected by the second vision sensor (420) after the j-th rotation, the value Q2.j of the center coordinates of the standard sphere (300) in the second vision sensor (420) is obtained. based on and Obtain the sphere center coordinates J1.j in the base coordinate system of the first robot (110), where J1.j satisfies: And, based on and Obtain the sphere center coordinates J2.j in the base coordinate system of the second robot (120), where J2.j satisfies: Where J2.j=J2.j′, J1.j=J1.j′, j≠j′, j′=1、2......m; Using Relations Calibration in, Let be the rotation transformation matrix between the two robot base coordinate systems. Let be the translation transformation matrix between the two robot base coordinate systems.

5. The calibration method according to claim 4, characterized in that, The second calibration process includes: The first fitted circular curve is constructed based on J1.1 to J1.m, with its center at O. r1 ; Constructing the Z-axis of the turntable coordinate system r Axis, Z r The axis passes through the center O of the circle r1 And it is perpendicular to the plane containing the first fitted circular curve; Constructing the X-axis of the turntable coordinate system r Axis, X r The axis is from the center O r1 The vector pointing to point T1.1, where T1.1 is the projection point of J1.1 onto the plane containing the fitted circular curve; Constructing the Y-axis of the turntable coordinate system r axis, Y r Axis, X r Axis and Z r The axes are perpendicular to each other, and the Y axis is perpendicular to each other. r Axis, X r Axis and Z r The intersection point between any two axes is the center O of the circle. r1 ; Obtain the coordinates of the center of the standard sphere (300) in the turntable coordinate system, Z1.j, where Z1.j satisfies: Where θ is the angle of rotation of the turntable (200) each time, mθ=360°, ||O r1 J 1.j || is the radius of the first fitted circular curve; According to the relation and Calibration in, Let be the rotation transformation matrix between the base coordinate system of the first robot (110) and the turntable coordinate system. Let be the translation transformation matrix between the base coordinate system of the first robot (110) and the turntable coordinate system.

6. The calibration method according to claim 4, characterized in that, The second calibration process includes: Based on the second fitted circular curves J2.1 to J2.m, the center of the second fitted circular curve is O. r2 ; Constructing the Z-axis of the turntable coordinate system r Axis, Z r The axis passes through the center O of the circle r2 And it is perpendicular to the plane containing the fitted circular curve; Constructing the X-axis of the turntable coordinate system r Axis, X r Axis along the center O r2 The vector pointing to T2.1, where T2.1 is the projection point of J2.1 onto the plane containing the second fitted circular curve; Constructing the Y-axis of the turntable coordinate system r axis, Y r Axis, X r Axis and Z r The axes are perpendicular to each other, and the Y axis is perpendicular to each other. r Axis, X r Axis and Z r The intersection point between any two axes is the center O of the circle. r2 ; Obtain the value Z2.j of the center coordinates of the standard sphere (300) in the turntable coordinate system, where Z2.j satisfies: Where θ is the angle of rotation of the turntable (200) each time, mθ=360°, ||O r2 J2.j|| is the radius of the second fitted circular curve; According to the relation and Calibration and in, Let be the rotation transformation matrix between the base coordinate system of the second robot (120) and the turntable coordinate system. Let be the translation transformation matrix between the base coordinate system and the turntable coordinate system of the second robot (120).

7. The calibration method according to claim 1, characterized in that, The tool head is a wheel-shaped machining head, which includes a side plane (501) and an outer arc surface (502). The intersection of the side plane (501) and the outer arc surface (502) is the inner ring (503), and the outermost curve of the outer arc surface (502) is the outer ring (504). The third calibration process includes: Adjust the position of the robots so that the side plane (501) of the tool head on one of the robots corresponds to the vision sensor on the other robot; Several points on the side plane (501) of the acquisition tool head are used as the first acquisition points; Fit the first acquisition point to form a fitting plane; Several points on the outer arc surface (502) of the acquisition tool head are used as the second acquisition points; Measure the distance between each second acquisition point and the fitting plane, and fit the second acquisition points with the largest distance from the fitting plane to form a third fitted circular curve; The center coordinates of the tool head are obtained based on the third fitted circular curve; Based on the homogeneous transformation relationship between the coordinate systems of each robot's vision sensor and the base coordinate system, and the homogeneous transformation relationship between the base coordinate systems of the two robots, the relationship between the center coordinates of the tool head and the base coordinate systems of each robot is obtained.

8. The calibration method according to claim 1, characterized in that, The tool head is a spherical machining head. The third calibration process includes: Adjust the position of the robots so that the spherical surface (505) of the tool head on one of the robots corresponds to the vision sensor on the other robot; Several points on the spherical surface (505) of the acquisition tool head are used as the third acquisition points; The third acquisition point is fitted to form a fitted sphere; The center coordinates of the toolhead are obtained based on the fitted sphere. Based on the homogeneous transformation relationship between the coordinate systems of each robot's vision sensor and the base coordinate system, and the homogeneous transformation relationship between the base coordinate systems of the two robots, the relationship between the center coordinates of the tool head and the base coordinate systems of each robot is obtained.

9. The calibration method according to claim 1, characterized in that, Between the second and third calibration steps, the calibration method further includes: removing the standard ball (300).

10. The calibration method according to claim 1, characterized in that, After performing the third calibration step, the calibration method further includes: removing the first vision sensor (410) or the second vision sensor (420); wherein the first robot (110) is equipped with the first vision sensor (410), and the second robot (120) is equipped with the second vision sensor (420).

Images