An error compensation method for compensating for distance work accuracy of an industrial robot

Abstract

The application discloses an error compensation method for compensating distance work accuracy of an industrial robot, and comprises the following steps: obtaining actual grid point distance error and grid center point distance error according to nominal grid point distance, nominal grid center point distance, actual grid point distance and actual grid center point distance; optimizing grid point first weight and grid point second weight according to the introduced distance error weight adjustment factor under the corresponding grid in the unit of grid to obtain a comprehensive weight; obtaining sample point distance error by interpolation using the actual grid point distance error and the comprehensive weight; decomposing the sample point distance error in the nominal direction to obtain the position error of the sample point in the nominal direction; converting the position error into joint angle error through inverse solution; and compensating the joint angle error into a teach pendant to complete the compensation. The application overcomes the shortcoming that the calibration effect is limited by the openness of the controller, avoids the error caused by coordinate system conversion, and expands the calibration range to a certain extent.

Description

TECHNICAL FIELD

[0001] The present application relates to an error compensation method for compensating distance working accuracy of an industrial robot, and belongs to the field of industrial robot calibration. BACKGROUND

[0002] The movement distance of an industrial robot is to represent the moving distance in the working process of the robot, and the accuracy is crucial for the robot to carry out tasks such as carrying, assembling, and cutting and milling which have certain requirements for distance accuracy. At present, the method of calibrating kinematic parameters is usually used to improve the absolute position accuracy of the industrial robot, but for some working conditions that only require distance accuracy, improving the positioning accuracy will not only make the calibration work redundant and reduce the calibration efficiency, but also the modification of kinematic parameters is limited by the openness of the controller.

[0003] The spatial interpolation method is the most commonly used method in non-kinematic calibration, which usually uses the positioning error of the grid points around the sample points to interpolate the positioning error of the sample points. However, the process of using position information involves coordinate system conversion, which can introduce additional errors and reduce the accuracy of interpolation.

[0004] Therefore, the present application is proposed. SUMMARY

[0005] In view of the problems that the industrial robot is limited by the openness of the controller in traditional kinematic calibration and the coordinate system conversion in the process of using position information increases the error sources, the present application provides an error compensation method for compensating distance working accuracy of an industrial robot to solve the problem of insufficient accuracy in the weight calculation of the traditional spatial interpolation method.

[0006] The technical scheme of the present application is as follows:

[0007] According to the first aspect of the present application, an error compensation method for compensating distance working accuracy of an industrial robot is provided, which comprises the following steps:

[0008] Step 1: constructing a kinematic model of the industrial robot to obtain the kinematic parameters of the industrial robot;

[0009] Step 2: planning grid points, grid center points, sample points and reference points P0 in the working space of the industrial robot; obtaining the nominal distance of the grid points and the nominal distance of the grid center points according to the planned grid points, grid center points and reference points P0 in the working space of the industrial robot; obtaining the direction information from the reference points to the sample points as nominal direction information according to the sample points and the reference points P0;

[0010] Step 3, obtaining the actual distance from the grid point to the reference point as the grid point actual distance, and obtaining the actual distance from the grid center point to the reference point as the grid center point actual distance by measuring the device; obtaining the actual grid point distance error and the grid center point distance error according to the grid point nominal distance, the grid center point nominal distance, the grid point actual distance and the grid center point actual distance;

[0011] Step 4, calculating the grid point first weight according to the grid point distance error and the grid center point distance error under the corresponding grid in the unit of grid; calculating the grid point second weight by using the distance from the grid point to the sample point;

[0012] Step 5, optimizing the grid point first weight and the grid point second weight to obtain the comprehensive weight according to the introduced distance error weight adjustment factors μ1 and μ2 under the corresponding grid in the unit of grid;

[0013] Step 6, obtaining the sample point distance error by interpolation using the actual grid point distance error and the comprehensive weight;

[0014] Step 7, decomposing the sample point distance error in the nominal direction to obtain the position error of the sample point in the nominal direction;

[0015] Step 8, converting the position error into the joint angle error by inverse solution, and compensating the joint angle error into the teach pendant to complete the compensation.

[0016] Further, the kinematics model includes but is not limited to D-H modeling and MD-H modeling.

[0017] Further, the step 2 specifically includes: dividing C cubic grids in the industrial robot workspace, each cubic grid being composed of K grid points, the kth grid point P ck in the cth grid, c=1, 2, …, C, k=1, 2, …, K; obtaining C grid center points Z c in the C grids, c=1, 2, …, C; randomly selecting one point in each of the C grids to obtain N sample points S n , n=1, 2, …, N, C=N; taking a point in the workspace different from the sample points and the grid points as a reference point P0; performing Euclidean distance operation on the coordinates of the grid points and the grid center points and the reference point to obtain the grid point nominal distance LP ck and the C grid center point nominal distance Lz c ; and obtaining the direction information from the reference point to the sample points as the nominal direction information.

[0018] Further, the step 3 specifically includes: collecting the actual distance from the grid point to the reference point as the grid point actual distance LPR ck, the actual distance from the grid center point to the reference point is collected as the grid center point actual distance LZR c ; the grid point actual distance LPR ck is subtracted from the grid point nominal distance LP ck to obtain the grid point distance error ALPR ck ; the grid center point actual distance LZR c is subtracted from the grid point nominal distance LZ c to obtain the grid center point distance error ALZR c .

[0019] Further, the grid point first weight obtained in step 4 is specifically:

[0020] Taking the first grid as an example, the first weight of each grid point in the first grid is calculated according to the grid point distance error and the grid center point distance error, and the calculation formula is:

[0021]

[0022] In the formula, w 1k , k = 1, 2, …, K is the first weight of the kth grid point in the first grid, w 1K is the first weight of the Kth grid point in the first grid; ALPR 1k , k = 1, 2, …, K is the distance error of the kth grid point in the first grid, ALPR 1K is the distance error of the Kth grid point in the first grid; ALZR1 is the first grid center point distance error; Cov() represents covariance calculation.

[0023] Further, the grid point second weight calculation formula is:

[0024]

[0025] In the formula, q ck represents the second weight of the kth grid point in the cth grid, d ck represents the Euclidean distance from the kth grid point in the cth grid to the sample point in the cth grid.

[0026] Further, the comprehensive weight expression is:

[0027] Q ck = μ1w ck + μ2q ck

[0028] In the formula, Q ck represents the comprehensive weight of the kth grid point in the cth grid, w ck represents the first weight of the kth grid point in the cth grid, qck represents the second weight of the kth grid point in the cth grid.

[0029] Further, the values of the distance error weight adjustment factors μ1 and μ2 depend on If the grid where the cth sample point is located satisfies the inequality, μ1 takes 1 and μ2 takes 0; otherwise, μ1 takes 0 and μ2 takes 1; wherein, ΔLZR c The distance error of the cth grid center point where the cth sample point is located, ΔLPR ck The distance error of the kth grid point in the cth grid where the cth sample point is located.

[0030] According to a second aspect of the present application, there is provided an error compensation system for compensating distance work accuracy of an industrial robot, comprising the modules of the method of any one of the above.

[0031] According to a third aspect of the present application, there is provided a terminal comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, the processor being configured to perform the steps of the error compensation method for compensating distance work accuracy of an industrial robot of any one of the above.

[0032] The present application has the following advantages:

[0033] 1) The method of the present application does not need to modify the kinematic parameters, overcoming the disadvantage that the calibration effect is limited by the openness of the controller;

[0034] 2) The method of the present application does not need to perform the conversion between the measurement coordinate system of the measurement device and the base coordinate system of the industrial robot, avoiding the error caused by the coordinate system conversion;

[0035] 3) The calibration process of steps 3-8 of the method of the present application is not limited by the position of the measurement device, to some extent, the calibration range is expanded.

[0036] 4) The method of the present application proposes a comprehensive weight based on the first weight, the second weight and the introduction of the distance error weight adjustment factor, and the accuracy of the comprehensive weight is improved compared with the traditional weight calculation method. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 is a flow chart of the method of the present application;

[0038] Figure 2 is a kinematic model using the MD-H method in the embodiment;

[0039] Figure 3 is a grid point distribution diagram in the embodiment;

[0040] Figure 4 This is a simplified diagram of the experimental platform in the embodiment. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be arbitrarily combined with each other.

[0042] Example 1: As Figures 1-4 As shown, according to a first aspect of the present invention, an error compensation method for compensating for the distance working accuracy of an industrial robot is provided, comprising the following steps: Step 1, constructing a kinematic model of the industrial robot to obtain kinematic parameters of the industrial robot; Step 2, planning grid points, grid center points, sample points, and a reference point P0 in the workspace of the industrial robot; obtaining nominal distances between grid points and grid center points based on the planned grid points, grid center points, and reference point P0 in the workspace of the industrial robot; obtaining nominal direction information from the reference point to the sample point based on the sample points and the reference point P0 as nominal direction information; Step 3, obtaining the actual distance from the grid point to the reference point as the actual distance of the grid point and the actual distance from the grid center point to the reference point as the actual distance of the grid center point through a measuring device; and obtaining the actual distance from the grid center point to the reference point as the actual distance of the grid center point based on the nominal distances between the grid points, the nominal distances between the grid center points, and the actual distances between the grid points and the reference point P0; and obtaining the nominal direction information from the reference point to the sample point based on the nominal distances between the sample points and the sample point P0; Step 4, obtaining the nominal direction information based on the nominal distances between the grid points and the sample point P0; Step 5, obtaining the nominal direction information based on the nominal distances between the grid points and the sample point P0; Step 6, obtaining the nominal direction information based on the nominal distances between the grid points and the sample point P0; Step 7, obtaining the nominal direction information based on the nominal distances between the grid points and the sample point P0; Step 8, obtaining the nominal direction information based on the nominal distances between the grid points and the sample point P0; Step 9, obtaining the nominal direction information based on the nominal distances between the grid points and the sample point P0; Step 10, obtaining the nominal direction information based on the nominal distances between the grid points and the sample point P0; Step 11, obtaining the nominal direction information based on the nominal distances between the grid points and the sample point P0; Step Step 4: Using the grid as a unit, calculate the first weight of the grid point based on the grid point distance error and the grid center point distance error under the corresponding grid; calculate the second weight of the grid point using the distance from the grid point to the sample point; Step 5: Using the grid as a unit, optimize the first weight and the second weight of the grid point under the corresponding grid based on the introduced distance error weight adjustment factors μ1 and μ2 to obtain the comprehensive weight; Step 6: Use the actual grid point distance error and the comprehensive weight to interpolate and obtain the sample point distance error; Step 7: Decompose the sample point distance error in the nominal direction to obtain the position error of the sample point in the nominal direction; Step 8: Based on the industrial robot kinematic model established in Step 1, use the nominal kinematic parameters of the industrial robot to convert the position error into joint angle error through inverse kinematics, and compensate the joint angle error into the teach pendant to complete the compensation.

[0043] Furthermore, the kinematic model includes, but is not limited to, DH modeling and MD-H modeling.

[0044] Further, step 2 specifically involves: dividing the industrial robot's workspace into C cubic grids, each cubic grid consisting of K grid points, where the k-th grid point P in the c-th grid... ck Given c = 1, 2, ..., C, k = 1, 2, ..., K; obtain the center points Z of C grids in the C grids. c c = 1, 2, ..., C; randomly select one point from each of the C grids to obtain N sample points S. n n = 1, 2, ..., N, C = N (in this embodiment of the invention, a point is randomly selected as a sample point in each grid to construct N sample points); a point in the workspace that is different from the sample points and grid points is used as the reference point P0; the grid points and the grid center point are respectively subjected to Euclidean distance calculation with the reference point coordinates to obtain the nominal distance LP of the grid points. ck The nominal distance from the center point of each of the C grids to LZ c Simultaneously, obtain the direction information from the reference point to the sample point. As nominal directional information. Figure 3 Taking the workspace shown as an example, there are 64 cubic grids with a total of 125 grid points. Each cubic grid has K=8 grid points. The grid point coordinates, grid center point coordinates, sample point coordinates, and reference point coordinates are obtained through the industrial robot teach pendant.

[0045] Furthermore, the measuring equipment includes, but is not limited to, a laser tracker and a wire-type kinematic calibration system.

[0046] Furthermore, the actual distance from the grid point to the reference point is collected using measuring equipment as the actual distance between the grid points (LPR). ck Collect the actual distance from the grid center point to the reference point as the actual distance of the grid center point (LZR). c ; The actual distance of the grid points to LPR ck Nominal distance LP from grid point ck The difference is used to obtain the grid point distance error △LPR ck (expression is ΔLPR) ck =LPR ck -LP ck ); The actual distance from the grid center point to LZR c Nominal distance from grid point LZ c The difference is used to obtain the distance error △LZR between the grid center points. c (expression is ΔLZR) c =LZR c -LZ c ).

[0047] Further, taking the first grid as an example, the first weight of each grid point in the first grid is calculated according to the grid point distance error and the grid center point distance error under the first grid, and the calculation formula is as follows:

[0048]

[0049] In the formula, w 1k is the first weight of the kth grid point in the first grid, w 1K is the first weight of the Kth grid point in the first grid; ΔLPR 1k is the distance error of the kth grid point in the first grid, ΔLPR 1K is the distance error of the Kth grid point in the first grid; ΔLZR1 is the center point distance error of the first grid; Cov() represents covariance calculation. It should be noted that the present application constructs a cubic grid, and each cubic has 8 grid points. The calculation method of the first weight of each grid point under other grids is the same as above.

[0050] Further, the second weight calculation formula of the grid point is as follows:

[0051]

[0052] In the formula, q ck represents the second weight of the kth grid point in the cth grid, d ck represents the Euclidean distance from the kth grid point in the cth grid to the sample point in the cth grid.

[0053] The following example illustrates that the Euclidean distance expression from the first grid point in the first grid to the sample point in the first grid is as follows:

[0054]

[0055] In the formula, (x1, y1, z1) represents the coordinates of the sample point in the first grid, i.e. the nominal position of the sample point, which is obtained through the industrial robot teach pendant; (x 11 , y 11 , z 11 ) represents the coordinates of the first grid point in the first grid, i.e. the nominal position of the grid point, which is obtained through the industrial robot teach pendant.

[0056] Further, the comprehensive weight expression is as follows:

[0057] Q ck = μ1w ck + μ2q ck

[0058] In the formula, Q ckw represents the combined weight of the k-th grid point in the c-th grid. ck q represents the first weight of the k-th grid point in the c-th grid. ck This represents the second weight of the k-th grid point in the c-th grid.

[0059] The values ​​of the distance error weight adjustment factors μ1 and μ2 depend on If the grid containing the c-th sample point satisfies the inequality, then μ1 is 1 and μ2 is 0; otherwise, μ1 is 0 and μ2 is 1.

[0060] Furthermore, the interpolation formula is: Where, ΔLS n ΔLPR represents the distance error of the nth sample point. nk Q represents the distance error between the k-th grid point in the n-th grid. nk This represents the combined weight of the k-th grid point in the n-th grid.

[0061] Furthermore, step 7 specifically involves: for each sample point, that is, the distance error of the sample point in the nominal direction... The above is decomposed to obtain the nth sample point. Position error in direction (△x) n ,△y n ,△z n ).

[0062] Furthermore, the joint angle error obtained from the inverse solution is input into the teach pendant to verify the validity of the results.

[0063] According to a second aspect of the embodiments of the present application, there is provided an error compensation system for compensating distance working accuracy of an industrial robot, comprising the modules of the method in any of the above embodiments, and specifically comprising: a first module configured to perform step 1, constructing a kinematic model of the industrial robot to obtain kinematic parameters of the industrial robot; a second module configured to perform step 2, planning grid points, grid center points, sample points and a reference point P0 in a working space of the industrial robot; obtaining a nominal distance of the grid points and a nominal distance of the grid center points according to the grid points, the grid center points and the reference point P0 in the working space of the industrial robot; obtaining direction information from the reference point to the sample points as nominal direction information according to the sample points and the reference point P0; a third module configured to perform step 3, obtaining an actual distance from the grid points to the reference point as an actual distance of the grid points and obtaining an actual distance from the grid center points to the reference point as an actual distance of the grid center points by using a measuring device; obtaining actual grid point distance errors and grid center point distance errors according to the nominal distance of the grid points, the nominal distance of the grid center points and the actual distance of the grid points and the actual distance of the grid center points; a fourth module configured to perform step 4, calculating a first weight of the grid points according to the grid point distance errors and the grid center point distance errors in a corresponding grid in a grid unit; calculating a second weight of the grid points by using a distance from the grid points to the sample points; a fifth module configured to perform step 5, optimizing the first weight of the grid points and the second weight of the grid points to obtain a comprehensive weight according to the introduced distance error weight adjustment factors μ1 and μ2 in the corresponding grid in the grid unit; a sixth module configured to perform step 6, obtaining a sample point distance error by using the actual grid point distance errors and the comprehensive weight; a seventh module configured to perform step 7, decomposing the sample point distance error in a nominal direction to obtain a position error of the sample points in the nominal direction; and an eighth module configured to perform step 8, converting the position error into a joint angle error by inverse solution, compensating the joint angle error into a teach pendant, and completing the compensation. The term "module" used in the above description can be a combination of software and / or hardware for realizing a predetermined function. Although the system described in the above embodiments is preferably realized in software, realization in hardware or a combination of software and hardware is also possible and contemplated. For the parts not described in detail, reference can be made to the related description of the embodiments.

[0064] According to a third aspect of the embodiments of the present application, there is provided a terminal comprising a processor, a memory, and a computer program stored in the memory and executable on the processor, the processor being configured to perform the steps of the error compensation method for compensating distance working accuracy of an industrial robot according to any of the above embodiments.

[0065] It should be apparent to those skilled in the art that the modules or steps of the present application described above can be implemented by general computing devices, which can be centralized on a single computing device or distributed on a network composed of multiple computing devices, and which can be implemented by program codes executable by computing devices, so that they can be stored in storage devices and executed by computing devices, and in some cases, the steps shown or described can be executed in different order, or they can be made into individual integrated circuit modules or a single integrated circuit module. Thus, the present application is not limited to any particular combination of hardware and software.

[0066] An alternative embodiment of the present application is described below, which specifically includes:

[0067] Step 1, establishing the kinematic model of the industrial robot.

[0068] The industrial robot used in this embodiment is a serial six-degree-of-freedom industrial robot, and the kinematic model is constructed by MD-H method, as shown in Figure 2 The transformation matrix of the i-th joint of the industrial robot relative to the i-1-th joint coordinate system is shown in formula (1):

[0069]

[0070] wherein, R X (α i-1 ) represents the angle of rotation around the X i-1 axis from Z i-1 to Z i ; D X (a i-1 ) represents the distance along the X i-1 axis from Z i-1 to Z i ; R Z (θ i ) is the angle of rotation around the Z i axis from X i-1 to X i ; D Z (d i ) is the distance along the Z i axis from X i-1 to X i . The nominal kinematic parameters of the industrial robot are shown in Table 1.

[0071] Table 1

[0072]

[0073] The homogeneous transformation matrix of each adjacent joint of the robot is multiplied to obtain the homogeneous transformation matrix of the robot end coordinate system relative to the base coordinate system, as shown in formula (2):

[0074]

[0075] Through formula (1) and formula (2), the transformation matrix in the base coordinate system of the industrial robot can be expressed by formula (3):

[0076]

[0077] Wherein, the first 3 rows of the matrix represent the robot end pose, and the last row represents the position of the robot end tool in the robot base coordinate system.

[0078] Step 2, as shown in Figure 3 , 64 cubic grid points are divided in the industrial robot workspace, each cubic grid point is composed of 8 grid points, and a total of 125 grid points; 64 grid center points Z c (c=1,2,…,64) are obtained in the 64 grids composed of 125 grid points; 64 points are selected as sample points S n (n=1,2,…,64), and a point in the space different from the grid points, grid center points and sample points is selected as the reference point P0, so the origin (0,0,0) is selected as the reference point.

[0079] The grid points, grid center points and sample points are subjected to Euclidean distance operation with the reference point to obtain the nominal distance of the grid points LP ck , the nominal distance of the 64 grid center points Lz c (c=1,2,…,64) and the nominal distance of the 64 sample points LS n (n=1,2,…,64); In addition, the direction from the reference point to the sample point can also be obtained:

[0080]

[0081] Step 3, the distance information of the grid points and the grid center points is collected by using the laser tracker, and the experimental platform diagram of the measurement equipment is shown in Figure 4 . The collected data is used as the actual distance data. The actual distance of the grid points LPR ck is subtracted from the nominal distance of the grid points LP ck to obtain the grid point distance error△LPR ck , and the formula is:

[0082] ΔLPR ck = LPR ck -LP ck (5)

[0083] The actual distance LZR of the grid center point c The nominal distance LZ of the grid point c The distance error ALZR of the grid center point is obtained by subtracting the actual distance LZR of the grid center point from the nominal distance LZ of the grid point c , and the formula is as follows:

[0084] The distance error ALZR of the grid center point c = LZR c - LZ c (6)

[0085] Step 4, in units of grids, the first weight of each grid point in the corresponding grid is calculated according to the distance error of the grid point and the distance error of the grid center point; the second weight of each grid point is calculated according to the distance from the sample point to the grid point; taking the first grid as an example, the first weight of each grid point in the first grid is calculated according to the distance error of the grid point and the distance error of the grid center point, and the formula is as follows:

[0086]

[0087] In the formula, w 1k ,k=1,2,...K is the first weight of the kth grid point in the first grid, w 1K is the first weight of the Kth grid point in the first grid; ALPR 1k ,k=1,2,...K is the distance error of the kth grid point in the first grid, ALPR 1K is the distance error of the Kth grid point in the first grid; ALZR1 is the distance error of the center point of the first grid. It should be noted that the present application constructs a cubic grid, and each cubic has 8 grid points. The calculation method of the first weight of each grid point in other grids is the same as above.

[0088] The formula for calculating the second weight of the grid point is as follows:

[0089]

[0090] In the formula, q ck represents the second weight of the kth grid point in the cth grid, d ck represents the Euclidean distance from the kth grid point in the cth grid to the sample point in the cth grid.

[0091] The following examples illustrate that the Euclidean distance expression from the first grid point in the first grid to the sample point in the first grid is as follows:

[0092]

[0093] In the formula, (x1, y1, z1) represents the coordinates of the sample point in the first grid, i.e. the nominal position of the sample point, obtained by an industrial robot teach pendant; (x 11 ,y 11 ,z 11 ) represents the coordinates of the first grid point in the first grid, i.e. the nominal position of the grid point, obtained by an industrial robot teach pendant.

[0094] The comprehensive weight expression is:

[0095] Q ck = μ1w ck + μ2q ck (10)

[0096] In the formula, Q ck represents the comprehensive weight of the kth grid point in the cth grid, w ck represents the first weight of the kth grid point in the cth grid, and q ck represents the second weight of the kth grid point in the cth grid.

[0097] The values of the distance error weight adjustment factors μ1 and μ2 depend on If the grid in which the cth sample point is located satisfies the inequality, μ1 takes 1 and μ2 takes 0; otherwise, μ1 takes 0 and μ2 takes 1.

[0098] Step 6, the sample point distance error is obtained by using the actual grid point distance error and the comprehensive weight interpolation; the interpolation formula is:

[0099]

[0100] In the formula, ΔLS n represents the distance error of the nth sample point, ΔLPR nk represents the distance error of the kth grid point in the nth grid, and Q nk represents the comprehensive weight of the kth grid point in the nth grid.

[0101] The results after interpolation are shown in Table 2. Compared with before calibration, the maximum value precision is improved by 59.02%, the average value precision is improved by 67.70%, and the standard deviation precision is improved by 58.64%; compared with the traditional interpolation method, the maximum value precision is improved by 41.89%, the average value precision is improved by 64.67%, and the standard deviation precision is improved by 41.67%.

[0102] Table 2 Comparison of distance errors before and after calibration

[0103]

[0104] Step 7, for each sample point, i.e. the sample point distance error in the nominal direction The position error (△x n ,△y n ,△z n ) of the nth sample point in the x direction is obtained by decomposing the position error (△x n ,△y n ,△z n ) of the nth sample point in the x direction.

[0105] Step 8: The position error is converted into joint angle error (△θ1,△θ2,…,△θ6) by inverse solution and input into the teach pendant to complete compensation.

[0106] By using the technical scheme, the method of the present application does not need to modify the kinematics parameters, overcomes the shortcoming that the calibration effect is limited by the openness of the controller, does not need to convert between the measurement coordinate system of the measuring equipment and the base coordinate system of the industrial robot, avoids the error caused by the coordinate system conversion, is not limited by the position of the measuring equipment during the calibration process, and to some extent, expands the calibration range. The calculation method of the comprehensive weight is proposed, and the accuracy is improved compared with the traditional weight calculation method.

[0107] The specific embodiments of the present application are described in detail above in combination with the drawings, but the present application is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the purpose of the present application.

Claims

1. An error compensation method for compensating for a distance work accuracy of an industrial robot, characterized by, The method comprises the following steps: Step 1, constructing a kinematic model of the industrial robot to obtain kinematic parameters of the industrial robot; Step 2, planning grid points, grid center points, sample points and a reference point P0 in a working space of the industrial robot; obtaining nominal distances of the grid points and the grid center points according to the grid points, the grid center points and the reference point P0 in the working space of the industrial robot; obtaining direction information from the reference point to the sample points as nominal direction information according to the sample points and the reference point P0; Step 3, obtaining actual distances from the grid points to the reference point as actual distances of the grid points and obtaining actual distances from the grid center points to the reference point as actual distances of the grid center points through a measuring device; Obtaining actual grid point distance errors and grid center point distance errors according to the nominal distances of the grid points and the grid center points and the actual distances of the grid points and the grid center points; Step 4, calculating first weights of the grid points according to the grid point distance errors and the grid center point distance errors in a corresponding grid in a grid unit; Calculating second weights of the grid points by using distances from the grid points to the sample points; Step 5, optimizing the first weights of the grid points and the second weights of the grid points according to introduced distance error weight adjustment factors μ1 and μ2 to obtain comprehensive weights in a corresponding grid in a grid unit; Step 6, obtaining sample point distance errors by interpolation using the actual grid point distance errors and the comprehensive weights; Step 7, decomposing the sample point distance errors in a nominal direction to obtain position errors of the sample points in the nominal direction; Step 8, converting the position errors into joint angle errors by inverse solution and compensating the joint angle errors into a teach pendant to complete compensation. The first weights of the grid points obtained in the step 4 are specifically: Taking the first grid as an example, the first weights of the grid points in the first grid are calculated according to the grid point distance errors and the grid center point distance errors in the first grid, and the calculation formula is: where w 1k is the first weight of the kth grid point in the first grid, w 1K is the first weight of the Kth grid point in the first grid; ΔLPR 1k is the distance error of the kth grid point in the first grid, ΔLPR 1K is the distance error of the Kth grid point in the first grid; ΔLZR1 is the distance error of the center point of the first grid; Cov() represents the covariance calculation; The second weight of the grid point is calculated according to the following formula: In the formula, q ck represents the second weight of the kth grid point in the cth grid, d ck represents the Euclidean distance from the kth grid point in the cth grid to the sample point in the cth grid.

2. The error compensation method for compensating for a distance work accuracy of an industrial robot according to claim 1, characterized by, The kinematic model comprises D-H modeling and MD-H modeling.

3. The error compensation method for compensating for a distance work accuracy of an industrial robot according to claim 1, characterized by, The step 2 is specifically: dividing C cubic grids in the industrial robot workspace, each cubic grid being composed of K grid points, the kth grid point P ck in the cth grid, c=1, 2, …, C, k=1, 2, …, K; obtaining C grid center points Z c in the C grids, c=1, 2, …, C; randomly selecting one point in each of the C grids to obtain N sample points S n , n=1, 2, …, N, C=N; A reference point P0 is selected from the workspace points that are different from the sample points and grid points. The Euclidean distance between the grid points, the grid center point, and the reference point is calculated to obtain the nominal distance LP between the grid points. ck The nominal distance from the center point of each of the C grids to LZ c Simultaneously, obtain the direction information from the reference point to the sample point. As nominal directional information.

4. The error compensation method for compensating for distance work accuracy of an industrial robot according to claim 1, characterized by, Step 3 specifically involves: collecting the actual distance from the grid point to the reference point using a measuring device as the actual distance LPR between the grid points. ck Collect the actual distance from the grid center point to the reference point as the actual distance of the grid center point (LZR). c ; The actual distance of the grid points to LPR ck Nominal distance LP from grid point ck The difference is used to obtain the grid point distance error △LPR ck ; The actual distance from the grid center point to LZR c Nominal distance from grid point LZ c The difference is used to obtain the distance error △LZR between the grid center points. c .

5. The error compensation method for compensating for distance work accuracy of an industrial robot according to claim 1, characterized by, The comprehensive weight expression is: Q ck = μ1w ck + μ2q ck In the formula, Q ck represents the first weight of the kth grid point in the cth grid, q ck represents the first weight of the kth grid point in the cth grid, q ck represents the second weight of the kth grid point in the cth grid.

6. The error compensation method for compensating for a distance work accuracy of an industrial robot according to claim 5, characterized by, The values of the distance error weight adjustment factors μ1 and μ2 depend on If the grid in which the cth sample point is located satisfies the inequality, then μ1 takes the value 1 and μ2 takes the value 0; otherwise, μ1 takes the value 0 and μ2 takes the value 1; where ΔLZR c The distance error of the cth grid center point in which the cth sample point is located, ΔLPR ck The distance error of the kth grid point in the cth grid in which the cth sample point is located.

7. An error compensation system for compensating for distance to work accuracy of an industrial robot, characterized by A module comprising the method of any one of claims 1-6.

8. A terminal, characterized by comprising: A processor, a memory and a computer program stored in the memory and executable on the processor, the processor being configured to execute the steps of the error compensation method for compensating distance working accuracy of the industrial robot according to any one of claims 1-6.

Images