Robot kinematics parameter error calibration method based on subspace measurement

Abstract

The invention discloses a robot kinematics parameter error calibration method based on subspace measurement, and the method comprises the steps: setting a measurement system, dividing the measurement system into a plurality of subspaces according to the working space of a robot, and placing standard part units in different subspaces; a monocular cross-structured light measurement unit fixed at the tail end of the robot is used for respectively measuring standard component units in each delimited subspace, an objective function is established, and an algorithm is used for iterating the objective function, so that the purpose of robot calibration is achieved. According to the invention, the problem of light shielding during measurement of the laser tracker is avoided. The problems that the calibration of the whole large space of the robot needs large workload and long time, the calibration precision of the large space is usually not high, and the specific small space cannot meet the calibration of the working space of the robot are solved. Meanwhile, according to the working requirements of different subspaces of the robot, independent calibration of the subareas is carried out, so that the robot meets the positioning precision requirements of different subspaces in different working modes.

Description

Technical Field

[0001] The present invention relates to the field of precision measurement, and particularly to a method for calibrating the kinematic parameter errors of a robot based on subspace measurement. Background Art

[0002] In the wave of accelerating transformation in the field of intelligent manufacturing, robots, with their excellent high repeatability and high efficiency, have become a key force in promoting the development of Industry 4.0, injecting powerful impetus into the transformation and upgrading of the manufacturing industry.

[0003] With the development of fields such as aerospace and automotive manufacturing, higher requirements are put forward for the positioning accuracy of robots. However, during the production, assembly, etc. of robots, they are inevitably affected by errors in link lengths, gear clearances, joint angle errors, etc., which directly restrict the positioning accuracy of robots.

[0004] Currently, the calibration of robot kinematic parameter errors has become a widely popular and highly cost-effective technical means, and is widely used in scenarios for reducing robot positioning errors.

[0005] However, for the calibration of robot kinematic parameter errors, it is carried out in the entire large space that the robot can reach, or just a simple specific small space. For example, for the calibration of the entire large space of the robot, this calibration process usually requires a large amount of calibration work and a long time, and at the same time, the calibration accuracy of the large space is usually less than that in a certain small space; for the calibration of a simple specific small space, the robot may not be limited to this small space during operation. After calibration in the small space, the positioning accuracy of the robot can be significantly improved only in this specific small space, while the accuracy improvement for the space outside the specific small space is limited. And generally, robots need to work in certain specific spaces, or the accuracy requirements are different for different spaces due to different working methods.

[0006] Therefore, there is an urgent need to propose a method for calibrating kinematic parameter errors based on subspace measurement to solve the problems that the calibration of the entire large space of the robot requires a large amount of work and a long time, and the calibration accuracy of the large space is usually not high, and the specific small space cannot meet the calibration of the robot working space.

[0007] Based on the above deficiencies existing in the prior art, the present invention is proposed. Summary of the Invention

[0008] Aiming at the above deficiencies existing in the prior art, the purpose of the present invention is to provide a method for calibrating the kinematic parameter errors of a robot based on subspace measurement to solve the problems that the calibration of the entire large space of the robot requires a large amount of work and a long time, and the calibration accuracy of the large space is usually not high, and the specific small space cannot meet the calibration of the robot working space.

[0009] To achieve the object of the present invention, the technical solution provided by the present invention is as follows:

[0010] A method for calibrating the kinematic parameter errors of a robot based on subspace measurement, characterized by including the setting of a measurement system, dividing the working space of the robot into multiple subspaces according to the working space of the robot, and placing the standard part units in different subspaces; measuring the standard part units in each of the delimited subspaces respectively by a monocular cross-structured light measurement unit fixed at the end of the robot, establishing an objective function and using an algorithm to iterate it to achieve the purpose of robot calibration;

[0011] The specific steps include:

[0012] Step S1, dividing the working space of the robot;

[0013] Divide the working space of the robot into multiple subspaces according to the working space of the robot;

[0014] Step S2, placing the standard part units in the subspaces divided in Step S1 respectively, and measuring the standard part units respectively by the monocular cross-structured light measurement unit carried at the end of the robot;

[0015] The standard part unit includes a carbon fiber rod, and a plurality of standard balls are arranged on the carbon fiber rod;

[0016] The monocular cross-structured light measurement unit is fixed at the end of the robot, and measures a plurality of standard balls in the standard part units of different subspaces in sequence to obtain the radius of the standard balls and the coordinates of the centers of the balls in the camera coordinate system;

[0017] Step S3, collecting the robot joint angle data, performing robot error modeling, and obtaining an objective function;

[0018] Collect the robot joint angle data when the monocular cross-structured light measurement unit in Step S2 measures the standard part units, and through modeling the robot, obtain the positions of a plurality of standard balls in the standard part units in the robot coordinate system;

[0019] Establish an objective function according to the root mean square error between the distance values of the standard balls measured by the robot and the distance values of the corresponding standard balls provided by the standard part unit;

[0020] Step S4, using an algorithm to iterate to obtain the kinematic parameter errors of the robot in the subspace;

[0021] Iterate the objective function in Step S3 through an iterative algorithm to obtain the kinematic parameter errors of the robot when the minimum value of the robot measurement error is satisfied, and compensate for the kinematic parameter errors of the robot to achieve the purpose of calibration of this subspace;

[0022] Step S5: Calibrate multiple subspaces respectively to obtain the kinematic parameter errors of the robot in different subspaces.

[0023] After calibrating one subspace, replace it with a different subspace and repeat steps S2 - S4 to correspondingly compensate for the kinematic parameter errors of the robot in different subspaces, and obtain the results of compensating different parameter errors of the robot in each subspace.

[0024] Step S6: Determine whether all the subspaces divided in step S1 have been calibrated. If so, when all calibrations are completed, proceed to step S7; if not, when the calibration is not completed, replace the subspace where the standard part unit is placed and return to step S2 for measurement.

[0025] Step S7: Integrate the kinematic parameter errors of the robot in all subspaces to generate a kinematic model of the robot in the working space.

[0026] A further preferred technical solution provided by the present invention is:

[0027] In the above - mentioned step S1, when dividing the robot space, it is divided according to the working mode of the robot when working in a specific space or according to the positioning accuracy required by the robot in different subspaces; calibrate different subspaces respectively to meet the working requirements of the robot in different subspaces.

[0028] An even further preferred technical solution provided by the present invention is:

[0029] In the above - mentioned step S2, the carbon fiber rod is a carbon fiber rod with a low temperature coefficient of variation. The carbon fiber rod is arranged on the support frame, and the support frame is rotatably arranged on the turntable; through the rotation of the support frame and the turntable, the standard balls of the standard part unit are distributed throughout the subspace where they are located.

[0030] Another further preferred technical solution provided by the present invention is:

[0031] In the above - mentioned step S2, the monocular cross - structured light measurement unit is composed of a cross - structured light device and a camera, and the light strip of the cross - structured light device is captured by the camera to achieve the purpose of measurement.

[0032] The present invention also provides a technical solution for applying the method for calibrating the kinematic parameter errors of a robot based on subspace measurement in machine tool calibration.

[0033] Compared with the prior art, the beneficial effects of the present invention include:

[0034] 1) A method for calibrating the kinematic parameter errors of a robot based on subspace measurement according to the present invention uses materials and manufacturing devices with costs much lower than those of traditional instruments such as laser trackers for calibrating robots, and at the same time avoids the problem of light occlusion during measurement by the laser tracker.

[0035] 2) A method for calibrating the kinematic parameter errors of a robot based on subspace measurement according to the present invention solves the problems that large workload and long time are required for calibrating the entire large space of the robot, and the calibration accuracy of the large space is usually not high, and the specific small space cannot meet the calibration of the robot's working space.

[0036] 3) At the same time, the present invention can perform independent calibration of sub-regions according to the working requirements of different sub-spaces of the robot, so that the robot meets the positioning accuracy requirements when working in different ways in different sub-spaces. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0038] Figure 1 It is a schematic diagram of the calibration process of a method for calibrating the kinematic parameter errors of a robot based on subspace measurement according to the present invention;

[0039] Figure 2 It is a schematic diagram of the measurement process of a method for calibrating the kinematic parameter errors of a robot based on subspace measurement according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings.

[0041] Figure 1 It is a schematic diagram of the calibration process of the present invention, Figure 2 It is a schematic diagram of the measurement process of the present invention. Combining with the attached Figure 1 、 Figure 2 As shown in, a method for calibrating the kinematic parameter errors of a robot based on subspace measurement provided by an embodiment of the present invention includes setting a measurement system, specifically: dividing the working space of the robot 1 into multiple sub-spaces. In this embodiment, the first sub-space 4, the second sub-space 5, and the third sub-space 6 set in the attached are adopted, and the standard part unit 3 is placed in different sub-spaces;

[0042] The standard part unit 3 in the first subspace 4, the second subspace 5, and the third subspace 6 is measured by the monocular cross-structured light measurement unit 2 fixed at the end of the robot 1, and an objective function is established and iterated using an algorithm to achieve the purpose of calibrating the robot 1;

[0043] Iterate through the first subspace 4, the second subspace 5, and the third subspace 6 to obtain different kinematic parameter errors of the robot 1, so as to ensure the improvement of the positioning accuracy of the robot 1 in this subspace.

[0044] The specific steps are introduced as follows:

[0045] The calibration process schematic diagram of the present invention is as Figure 2 shown.

[0046] Step S1: Divide the working space of the robot 1 into multiple subspaces according to the working space of the robot 1;

[0047] During calibration, first fully consider the working space of the robot 1, and classify the working space into different subspaces. These subspaces can be divided according to the working mode of the robot when working in a specific space, or according to the positioning accuracy required by the robot in different subspaces; the purpose of the division is to calibrate different subspaces respectively to meet the working requirements of the robot in different subspaces;

[0048] In this embodiment, the working space of the robot 1 is divided into a first subspace 4, a second subspace 5, and a third subspace 6;

[0049] Step S2: Place the standard part unit 3 in the divided subspaces respectively, and the monocular cross-structured light measurement unit 2 carried at the end of the robot 1 measures the standard part unit 3 respectively;

[0050] According to the subspaces divided in step S1, as Figure 1 Figure 2 shown, place the standard part unit 3 in the first subspace 4;

[0051] Among them, the standard part unit 3 includes a carbon fiber rod 302, and a plurality of standard balls 301 are arranged on the carbon fiber rod 302;

[0052] The carbon fiber rod 302 is a carbon fiber rod 302 with a low temperature coefficient, the carbon fiber rod 302 is arranged on the support frame 303, and the support frame 303 is rotatably arranged on the turntable 304;

[0053] Through the rotation of the support frame 303 and the turntable 304, the standard balls 301 of the standard part unit 3 are scattered throughout the subspace where they are located;

[0054] The monocular cross-structured light measurement unit 2 is fixed at the end of the robot 1, and measures multiple standard balls 301 in the standard part unit 3 of different subspaces in sequence to obtain the radius of the standard ball and the coordinates of the ball center in the camera coordinate system;

[0055] Among them, the monocular cross-structured light measurement unit 2 is composed of a cross-structured light device 202 and a camera 201. By capturing the light strip of the cross-structured light device 202 with the camera 201, the purpose of measurement is achieved.

[0056] Step S3: Collect the joint rotation angle data of the robot 1, perform error modeling on the robot 1, and obtain the objective function;

[0057] Collect the joint angle data of the robot 1 when the monocular cross-structured light measurement unit 2 measures the standard part unit 3 in step S2. By modeling the robot 1, obtain the positions P of multiple standard balls 301 in the standard part unit 3 in the coordinate system of the robot 1;

[0058] The transformation relationship T between the (i - 1)-th joint and the i-th joint of the robot 1 i i-1 is expressed as:

[0059]

[0060] where, θ i represents the joint variable; d i represents the link offset; a i represents the link length; α i represents the link twist angle; β i represents the rotation variable about the Y-axis; cθ i represents cosθ i which is the cosine of θ i , cα i represents cosα i which is the cosine of α i ; sθ i represents sinθ i which is the sine of θ i , sα i represents sinα i which is the sine of α i ; Rot represents rotational motion, and Trans represents translational motion.

[0061] Due to the influence of errors such as machining and assembly on the robot 1, there is a certain error between the actual value and the theoretical value of the kinematic parameters of the robot 1. Performing total differentiation on results in the transformation matrix error between joints being expressed by Equation (2).

[0062]

[0063] Among them, Δθ i , Δd i , Δa i , Δα i and Δβ i are the corresponding errors of the kinematic parameters θ i , d i , a i , α i and β i . Therefore, the position P of the standard sphere measured by the monocular cross-structured light measurement unit 2 in the coordinate system of the robot 1 can be further expressed as:

[0064]

[0065] Among them, is the matrix of the hand-eye calibration of the robot unit 1. [x1, y1, z1] is the position coordinate of the standard sphere measured by the monocular cross-structured light measurement unit 2. The hand-eye calibration process uses the method disclosed in the invention patent with the publication number CN 118559755 A and the invention name of a robot calibration method based on multi-line structured light applied by the applicant on June 13, 2024;

[0066] The theoretical position coordinates of the m-th and k-th standard sphere centers based on distance are respectively denoted as P Tm (x Tm , y Tm , z Tm ) and P Tk (x Tk , y Tk , z Tk ). Then, the spatial distance L Tmk between the theoretical positions of these two standard sphere centers is expressed as:

[0067]

[0068] The positions of these standard spheres 301 and the distances between any two standard spheres 301 have been measured by a coordinate measuring machine after the production and processing of the standard part unit 3 as the standard value L Tmk provided by the standard part unit 3;

[0069] Then, according to the root mean square error between the distance value L Amk between the standard spheres 301 measured by the robot 1 and the distance value L Tmk of the corresponding standard spheres 301 provided by the standard part unit 3, an objective function f1 is established;

[0070]

[0071] Where N is the total number of squared distance differences, q is the square of the qth distance difference, θ is the joint angle of the six rotary joints of robot 1 at two positions, and Δε1 is the kinematic parameter error Δθ of robot 1. i , Δd i , Δa i , Δα i and Δβ i Step S4, using the LM algorithm to iterate and obtain the kinematic parameter error of the robot 1 in the first subspace 4;

[0072] The LM algorithm is used to minimize the error function f1 to obtain the robot kinematic parameter error Δε1. The robot kinematic parameter error is compensated to achieve the purpose of calibration of the first subspace 4;

[0073] Step S5, calibrating multiple subspaces separately to obtain kinematic parameter errors of the robot 1 in different subspaces;

[0074] Repeat steps S2-S4, placing the standard component unit 3 in the second subspace 5 and the third subspace 6. Use the calibration method from the first subspace 4 to establish objective functions f2 and f3 in the second and third subspaces, respectively. Use the LM algorithm to minimize the error functions f2 and f3, respectively, to obtain the robot kinematic parameter errors Δε2 and Δε3. Compensate for the kinematic parameter errors of robot 1 in different subspaces, and obtain the compensation results for the different parameter errors of robot 1 in each subspace.

[0075] Step S6, determining whether all subspaces divided in step S1 have been calibrated. If so, all calibrations are completed and the process proceeds to step S7; if not, the process proceeds to the subspace where the standard unit 3 is placed and returns to step S2 for measurement.

[0076] The last step is to integrate the calibration of each subspace, aiming to perform different calibrations in different workspaces.

[0077] The robot kinematic errors obtained in step S5 are sorted out, and models are established for different subspaces respectively to generate measurement models of all subspaces to meet the kinematic model of the robot in the workspace.

[0078] It is not limited to these three subspaces, and multiple spaces can be divided according to the working needs of the robot.

[0079] Step S7 , sorting out the kinematic parameter errors of the robot in the first subspace 4 , the second subspace 5 , and the third subspace 6 , and generating a kinematic model of the robot in the workspace.

[0080] Organize the kinematic errors of the robot obtained in step S5, establish models for different subspaces respectively, generate measurement models for all subspaces, and satisfy the kinematic model of the robot in the workspace.

[0081] Meanwhile, the measurement method for robot error calibration based on subspaces provided by the present invention is applicable not only to robots but also to the calibration and measurement of machine tools.

[0082] The described embodiments are only a part of the embodiments of this application, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts shall fall within the scope of protection of this application.

Claims

1. A method for calibrating the kinematic parameter errors of a robot based on subspace measurement, characterized in that It includes measurement system setting. According to the working space of the robot, it is divided into multiple sub-spaces, and the standard part units are placed in different sub-spaces. The standard part units in each of the delimited sub-spaces are measured separately by the monocular cross-structured light measurement unit fixed at the end of the robot, and an objective function is established and iterated using an algorithm to achieve the purpose of robot calibration. The specific steps include: Step S1, division of the robot working space; According to the robot working space, the robot working space is divided into multiple sub-spaces; Step S2, the standard part units are respectively placed in the sub-spaces divided in Step S1, and the monocular cross-structured light measurement unit carried at the end of the robot measures the standard part units respectively; Among them, the standard part unit includes a carbon fiber rod, and multiple standard balls are arranged on the carbon fiber rod; The monocular cross-structured light measurement unit is fixed at the end of the robot, and measures multiple standard balls in the standard part units of different sub-spaces in sequence to obtain the radius of the standard balls and the coordinates of the ball centers in the camera coordinate system; Step S3, collect the robot joint angle data, perform robot error modeling, and obtain the objective function; Collect the robot joint angle data when the monocular cross-structured light measurement unit in Step S2 measures the standard part unit, and through modeling the robot, obtain the positions of multiple standard balls in the standard part unit in the robot coordinate system; Establish an objective function based on the root mean square error between the distance values of the standard balls measured by the robot and the distance values of the corresponding standard balls provided by the standard part unit; Step S4, use an algorithm to iterate and obtain the kinematic parameter error of the robot in the sub-space; Iterate the objective function in Step S3 through an iterative algorithm to obtain the kinematic parameter error of the robot when the minimum robot measurement error is satisfied, and compensate for this kinematic parameter error of the robot to achieve the purpose of calibration of this sub-space; Step S5, calibrate multiple sub-spaces respectively to obtain the kinematic parameter errors of the robot in different sub-spaces; After one sub-space calibration is completed, replace with different sub-spaces, repeat Steps S2 - S4, and perform corresponding compensation on the kinematic parameter errors of the robot in different sub-spaces to obtain the results of different parameter error compensations of the robot in each sub-space; Step S6, determine whether all the sub-spaces divided in Step S1 have been calibrated. If so, when all calibrations are completed, enter Step S7; if not, when the calibration is not completed, replace the sub-space where the standard part unit is placed, and return to Step S2 for measurement; Step S7, fuse the kinematic parameter errors of the robot in all sub-spaces to generate the kinematic model of the robot in the working space.

2. A method for calibrating the kinematic parameter errors of a robot based on subspace measurement according to claim 1, characterized in that In the said Step S1, the division of the robot space is carried out according to the working mode of the robot when working in a specific space, or according to the positioning accuracy required by the robot in different sub-spaces; calibrate different sub-spaces respectively to meet the working requirements of the robot in different sub-spaces.

3. A kinematic parameter error calibration method for a robot based on subspace measurement according to claim 1, characterized in that In the step S2, the carbon fiber rod is a carbon fiber rod with a low temperature coefficient of variation. The carbon fiber rod is arranged on a support frame, and the support frame is rotatably arranged on a turntable. By rotating the support frame and the turntable, the standard balls of the standard part unit are distributed throughout the subspace where they are located.

4. A method for calibrating the kinematic parameter error of a robot based on subspace measurement according to claim 1, characterized in that In the step S2, the monocular cross-structured light measurement unit is composed of a cross-structured light device and a camera. By capturing the light strip of the cross-structured light device with the camera, the purpose of measurement is achieved.

5. An application of a method for calibrating the kinematic parameter errors of a robot based on subspace measurement according to any one of claims 1-4 in machine tool calibration.

Images