Mechanical arm connecting rod error compensation method based on LM iteration method

By using a robotic arm link error compensation method based on the LM iterative method, and adjusting the DH parameters of the robotic arm using a laser tracker and the LM iterative algorithm, the problem of insufficient positioning accuracy of the robotic arm is solved, and high-precision positioning adaptability is achieved.

CN118990443APending Publication Date: 2024-11-22NANJING LANHAO INTELLIGENT TECH CO LTD
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202411210914.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-30
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The positioning accuracy of existing robotic arms is insufficient to meet high-precision requirements, resulting in poor positioning performance in high-precision work scenarios.

Method used

A robotic arm linkage error compensation method based on the LM iterative algorithm is adopted. The actual coordinates of the robotic arm during its movement are collected in real time by a laser tracker and compared with the theoretical coordinates. The DH parameters of the robotic arm are adjusted by the LM iterative algorithm until the positioning error meets the accuracy requirements.

Benefits of technology

It improves the repeatability and positioning accuracy of the robotic arm, ensuring high-precision positioning under different operating conditions and adapting to different operating scenarios and task requirements.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118990443A_ABST
    Figure CN118990443A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of mechanical arm positioning precision compensation, and particularly relates to a mechanical arm connecting rod error compensation method based on an LM iteration method, which comprises the following steps: writing a teaching point location program to enable a mechanical arm to run to a plurality of preset point locations; the mechanical arm executes the program, and meanwhile, the laser tracker collects actual coordinate values corresponding to all point positions in the movement process of the mechanical arm; theoretical coordinate values of the tail end of the mechanical arm are calculated according to connecting rod parameters of the mechanical arm; the actual coordinates are compared with the theoretical coordinates, and the positioning error of the mechanical arm is calculated; and adjusting the connecting rod parameters of the mechanical arm by using an LM iterative algorithm, and compensating the error until the error is smaller than a preset value. The repeated positioning precision of the mechanical arm is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of mechanical arm positioning accuracy compensation, and in particular relates to a mechanical arm connecting rod error compensation method based on LM iteration method. Background Art

[0002] With the continuous development of industry, industrial robotic arms are replacing manual labor in various industries to carry out production operations. Precision compensation is an important branch of robotics research. Higher robotic arm positioning accuracy allows the robotic arm to participate in more high-precision work. Robots in different working scenarios have different requirements for accuracy. In high-precision machining, the control accuracy of the robotic arm is particularly high, which places high requirements on the robotic arm processing technology.

[0003] At present, due to the difficulty in controlling the casting accuracy of the robot arm and the different assembly technologies of the robot arm, the positioning accuracy of the robot arm is difficult to achieve the desired effect. By compensating the robot arm parameters, the positioning accuracy of the robot arm can be improved on the existing basis so that it can complete the given work requirements. Summary of the invention

[0004] Purpose of the invention: The purpose of the present invention is to address the deficiencies in the prior art and provide a method for compensating the error of a robot arm connecting rod based on the LM iteration method, thereby greatly improving the repeatability of the positioning of the robot arm.

[0005] Technical solution: The robot arm connecting rod error compensation method based on the LM iteration method of the present invention comprises the following steps:

[0006] S1: Write a teaching point program to make the robot arm move to the preset point;

[0007] S2: After the laser tracker is preheated and initialized, the robot arm executes the teaching point program, and the laser tracker is used to collect the actual coordinate value of the point to which the robot arm moves;

[0008] S3: Calculate the theoretical coordinate value of the end of the robot arm according to the existing connecting rod parameters of the robot arm;

[0009] S4: Compare and calculate the actual coordinate value with the theoretical coordinate value to obtain the positioning error of the robot arm; use the LM iterative algorithm to adjust the connecting rod parameters of the robot arm to compensate for the positioning error until the positioning error is less than a predetermined value.

[0010] To further improve the above technical solution, in S1, for each teaching point, a corresponding joint movement instruction is written, and this operation is repeated until the teaching points are spread throughout the working space of the robot arm.

[0011] Furthermore, in S2, the robot arm executes the teaching point program in a single step, waits for the laser tracker to collect data each time a joint movement instruction is executed, and executes the next instruction after the collection is completed, until all data collection is completed, and the actual coordinate set measPoints collected by the laser tracker is obtained.

[0012] Furthermore, S3 includes: using DH parameters as the robot arm link parameters to obtain initial DH parameters; for each collected joint angle set, applying the forward kinematics equation, calculating the transformation matrix through the DH parameters and the joint angles, extracting the world coordinates of the robot arm end from the transformation matrix, and recording them as the theoretical coordinate set theoPoints of the robot arm end.

[0013] Further, the S4 includes:

[0014] S4.1, calculating the spatial distance set theoL between adjacent points in the theoretical coordinate set theoPoints by using the distance calculation function calc_dis(p);

[0015] S4.2, calculating the spatial distance set measL between adjacent points in the actual coordinate set measPoints collected by the laser tracker through the distance calculation function calc_dis(p);

[0016] S4.3, performing error analysis on the set theoL and the set measL, calculating the error vector and the corresponding maximum error, average error and variance, and judging whether the scenario application requirements of the robot arm are met according to the set error criteria. If the analysis result meets the scenario application requirements of the robot arm, the compensation is completed; if the result does not meet the requirements, executing S4.4;

[0017] S4.4. If the error does not meet the application requirements, the new DH parameters are calculated using the LM iterative algorithm.

[0018] new_DH=calibration(old_DH,δ)

[0019] Among them, new_DH is the output DH parameter after compensation, old_DH is the DH parameter before compensation, and δ is the distance set;

[0020] S4.5. Superimpose the result calculated in step 4.4 on the current DH parameter and repeat the above steps until the error meets the requirement.

[0021] Further, the calculation of new DH parameters using the LM iterative algorithm includes:

[0022] Initialization: set the initial DH parameters and calculate the initial error;

[0023] Iterate: calculate the Jacobian matrix J and the error vector e;

[0024] Select the damping factor λ and solve the parameter update equation δ = -(J T J+λI) -1 J T e;

[0025] Use δ to update DH parameters and recalculate the error;

[0026] Adjust the damping factor λ according to the error change;

[0027] Repeat the iteration until the termination condition is met.

[0028] Furthermore, if the error decreases, accept the update and decrease λ; otherwise, increase λ and try again.

[0029] Beneficial effect: Compared with the prior art, the advantages of the present invention are: the present invention uses a laser tracker to collect the actual coordinates of each point in the movement process of the robot arm in real time, and compares them with the theoretical coordinates. This high-precision measurement method can accurately capture the slight deviation of the robot arm and ensure the accuracy of the error calculation. By introducing the LM iterative algorithm, the DH parameters of the robot arm are intelligently adjusted. The algorithm can automatically optimize the parameters according to the error data, so that the positioning error of the robot arm is gradually reduced until the set accuracy requirements are met. The method of dynamically adjusting the DH parameters is used, which means that the system can automatically adjust the geometric parameters of the robot arm according to the deviation data in the actual operation to adapt to different operation scenarios and task requirements. Through the real-time measurement of the laser tracker and the dynamic compensation of the LM algorithm, the scheme can adjust the parameters of the robot arm in real time during the actual operation to ensure high-precision positioning under different operating conditions. The present invention performs error compensation for specific connecting rod parameters, and the accuracy of compensation and the calculation speed of compensation parameters can be greatly improved by selecting constraint parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flowchart of the steps of the present invention;

[0031] Figure 2 is the positioning error of each point before and after calibration;

[0032] Figure 3 is the overall system error;

[0033] Figure 4 It is to calibrate the front and rear link parameters. DETAILED DESCRIPTION

[0034] The technical solution of the present invention is described in detail below with reference to the accompanying drawings, but the protection scope of the present invention is not limited to the embodiments.

[0035] Example 1: Figure 1 The robot arm link error compensation method based on the LM iteration method shown in the figure comprises the following steps:

[0036] Step S1: First, let the robot arm run 51 points through teaching, and write the corresponding teaching point program PROG;

[0037] Step S1.1: Write the corresponding joint movement instructions for each point, and select the teaching points to cover the working space of the robot as much as possible;

[0038] Step S1.2: Repeat the above operation until 51 points are taught.

[0039] Step S2: After the laser tracker is preheated and initialized, the robot arm executes the program PROG, and the laser tracker collects the corresponding coordinate data;

[0040] Step S2.1: First, preheat the laser tracker and initialize the position;

[0041] Step S2.2: The robot executes the program PROG in a single step. Each time a joint movement instruction is executed, it waits for the laser tracker to collect data. After the collection is completed, it executes the next instruction until all the data are collected and the robot end coordinate set measPoints is obtained.

[0042] Step S3: Calculate the corresponding end coordinate set according to the existing link parameters of the robot arm;

[0043] Step S3.1: Calculate the world coordinates of the end of the robot arm corresponding to the acquisition joint angle through the initial robot arm link parameters, and record the coordinate values ​​of the end of the robot arm to obtain the theoretical coordinate set theoPoints of the end of the robot arm;

[0044] Step S3.2: By comparing the coordinates of the end of the robotic arm with the coordinates of the laser tracker, the positioning accuracy error of the robotic arm can be analyzed.

[0045] The connecting rod parameters adopt DH parameters. DH parameters (Denavit-Hartenberg parameters) are used to describe the geometric relationship between the joints of the robot arm. Assuming there are n joints, each connecting rod has four parameters: connecting rod length a, connecting rod torsion angle α, connecting rod offset d and joint angle θ. These parameters constitute the initial DH parameter table.

[0046] The initial DH parameters are based on the design of the robot or the results of previous calibration, and are used to determine the end position of the robot at any given joint angle configuration. When executing the teaching point program, the robot controller records the angle configuration of each joint. These joint angles are used to calculate the theoretical position of the end of the robot; the world coordinates of the end of the robot are calculated using the joint angle and DH parameters using the kinematic solution formula. This step is achieved through matrix transformation Ti. The state of each joint will form a 4x4 transformation matrix: Ti = Rot(z,θi)*Trans(z,di)*Trans(x,ai)*Rot(x,αi), and the total transformation matrix T is obtained by multiplication: T = T1*T2*...*Tn. The position information (the first three rows and the fourth column) is extracted from the total transformation matrix T to obtain the world coordinates of the end of the robot. The coordinates (x means,i ,y means,i 、z means,i ), where i represents different points; repeat the above process for each sampling point, and store the calculated terminal coordinates into the theoretical coordinate set theoPoints.

[0047] The laser tracker tracks the position of the end effector of the robot arm in real time and provides high-precision three-dimensional coordinate data (x theo,i ,y theo,i 、z theo,i ), this coordinate data represents the end position of the robot arm in actual operation, and the error vector e between the theoretical coordinate and the actual coordinate collected by the laser tracker is calculated, and the error vector e = actual coordinate - theoretical coordinate.

[0048] Step S4: Compensation of the connecting rod parameters is completed by performing error analysis on the theoretical coordinate set and the actual coordinate set of the laser tracker.

[0049] Step S4.1: Use the distance calculation function dis=calc_dis(p), where p is the input point set and dis is the output distance set. The distance calculation function calc_dis(p) is used to calculate the spatial distance set theoL between adjacent points in the theoretical coordinate set theoPoints, theoL=calc_dis(theoPoints).

[0050] S4.2. Calculate the spatial distance set measL between adjacent points in the actual coordinate set measPoints collected by the laser tracker through the distance calculation function calc_dis(p), measL=calc_dis(measPoints).

[0051] S4.3. Compare the distance set of adjacent points between the theoretical trajectory (theoL) and the actual trajectory (measL). For each pair of corresponding distances (measLi, theoLi), calculate their difference: ei = measLi – theoLi. The calculated difference represents the deviation between the actual movement of the robot between two adjacent teaching points and the theoretical movement. Combine all ei into an error vector e, analyze the error vector, calculate the corresponding maximum error, average error, and variance to evaluate the accuracy and error of the robot during the movement. Determine whether the scenario application requirements of the robot are met according to the set error criteria. If the analyzed result meets the scenario application requirements of the robot, the compensation is completed; if the result does not meet the requirements, execute S4.4.

[0052] S4.4. If the error does not meet the application requirements, the new DH parameters are calculated using the LM iterative algorithm.

[0053] new DH =calibration(old DH ,δ)

[0054] Among them, new DH is the output DH parameter after compensation, old DH , is the DH parameter before compensation, and δ is the distance set.

[0055] S4.5: By repeating the above calculation until the calculated robot arm positioning error is less than the agreed value, the compensation of the DH parameters is completed.

[0056] The goal of the LM algorithm is to minimize ||e||2 (the sum of squares of the error vector). The new DH parameters calculated using the LM iterative algorithm include:

[0057] Initialization: set the initial DH parameters and calculate the initial error;

[0058] Iterate: Calculate the Jacobian matrix J and the error vector e, where J represents the partial derivative of e with respect to the DH parameters;

[0059] The damping factor λ is selected and, in each iteration, the equation δ = -(J T J+λI) -1 J T e to update DH parameters;

[0060] Use δ to update DH parameters and recalculate the error;

[0061] Adjust the damping factor λ according to the error change;

[0062] Repeat the iteration until the termination condition is met.

[0063] If the error decreases, accept the update and decrease λ; otherwise, increase λ and try again.

[0064] Embodiment 2: Using the method of embodiment 1 to perform error compensation,

[0065] Step 1: Complete the robot arm point acquisition through teaching and write the corresponding program PROG. The robot arm joint angles are shown in Table 1.

[0066] 96.794 12.198 3 -63.707 -92.006 58.694 46.598 -32.3 -76.107 -58.406 71.494 -43.66 -6.2 -98.305 -96.206 25.895 63.298 -18.9 -18.506 -61.006 0.095 59.198 -12.8 -8.206 -53.806 -98.605 61.698 -17.3 71.193 -95.006 -75.005 17.298 -6.9 88.893 -85.406 -63.905 62.598 -46.6 80.593 -64.006 -48.405 34.089 -24.4 60.493 -47.906 -33.905 61.79 -11.6 44.093 -69.306 -16.605 49.389 -3.5 23.393 -64.806 -48.204 -76.41 179.999 61.593 -99.506 -59.705 -35.01 -42.5 112.392 -101.005 -86.795 -61.32 11.62 75.693 -101.706 -86.795 -22.72 25.32 89.593 -96.29 -37.995 -65.22 25.32 118.193 -79.39 -23.895 -35.72 -15.68 160.093 -79.09 12.004 -52.92 4.42 204.793 -77.49 124.204 57.48 181.72 -78.406 -95.989 127.104 46.38 -15.58 -88.007 -105.19 71.49 64.38 -40.08 -77.207 -66.49 68.19 -79.319 160.52 -71.405 -84.89 119.19 -4.92 -37.68 -70.707 -109.99

[0067] Step 2: Power on the laser tracker and preheat it. After preheating, perform self-calibration of the laser tracker. After completing the above operations, start data collection.

[0068] The robot arm executes the program PROG and the laser tracker performs data acquisition. The coordinate set is shown in Table 2.

[0069]

[0070]

[0071] Table 2

[0072] Step 3: Calculate the end coordinate set of the robot arm through the joint angle, and compare it with the coordinate set of the laser tracker to obtain the positioning error of the current robot arm. If the error is less than the agreed value, the compensation of the connecting rod error is completed, otherwise execute step 4.

[0073] Step 4: Calculate the compensation parameters of the connecting rod parameters by analyzing the errors calculated in step 3, and jump to step 3.

[0074] As described above, although the present invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the present invention itself. Various changes in form and details may be made without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A robot arm link error compensation method based on LM iteration method, characterized in that: The steps include: S1: Write a teaching point program to make the robot arm move to the preset point; S2: After the laser tracker is preheated and initialized, the robot arm executes the teaching point program, and the laser tracker is used to collect the actual coordinate value of the point to which the robot arm moves; S3: Calculate the theoretical coordinate value of the end of the robot arm according to the existing connecting rod parameters of the robot arm; S4: Compare and calculate the actual coordinate value with the theoretical coordinate value to obtain the positioning error of the robot arm; use the LM iterative algorithm to adjust the connecting rod parameters of the robot arm to compensate for the positioning error until the positioning error is less than a predetermined value.

2. The robot arm connecting rod error compensation method based on LM iteration method according to claim 1, characterized in that: In S1, for each teaching point, a corresponding joint movement instruction is written, and this operation is repeated until the teaching points are spread throughout the working space of the robot arm.

3. The robot arm connecting rod error compensation method based on LM iteration method according to claim 2, characterized in that: In S2, the robot arm executes the teaching point program in a single step, and waits for the laser tracker to collect data each time a joint movement instruction is executed. After the collection is completed, the next instruction is executed until all data are collected, and the actual coordinate set measPoints collected by the laser tracker is obtained.

4. The robot arm connecting rod error compensation method based on LM iteration method according to claim 3 is characterized in that: The S3 includes: Using DH parameters as the robot arm connecting rod parameters to obtain initial DH parameters; For each collected joint angle set, the forward kinematics equation is applied to calculate the transformation matrix through the DH parameters and the joint angles, and the world coordinates of the end of the robot are extracted from the transformation matrix and recorded as the theoretical coordinate set theoPoints of the end of the robot.

5. The robot arm connecting rod error compensation method based on LM iteration method according to claim 4 is characterized in that: The S4 includes: S4.1, calculating the spatial distance set theoL between adjacent points in the theoretical coordinate set theoPoints by using the distance calculation function calc_dis(p); S4.2, calculating the spatial distance set measL between adjacent points in the actual coordinate set measPoints collected by the laser tracker through the distance calculation function calc_dis(p); S4.3, performing error analysis on the set theoL and the set measL, and judging whether the scenario application requirements of the robot arm are met according to the set error criteria. If the analysis result meets the scenario application requirements of the robot arm, the compensation is completed; if the result does not meet the requirements, executing S4.4; S4.

4. If the error does not meet the application requirements, the new DH parameters are calculated using the LM iterative algorithm. in, is the output DH parameter after compensation, is the DH parameter before compensation, is the distance set; S4.

5. Superimpose the result calculated in step 4.4 on the current DH parameter and repeat the above steps until the error meets the requirement.

6. The robot arm connecting rod error compensation method based on LM iteration method according to claim 5, characterized in that: The calculation of new DH parameters using the LM iterative algorithm includes: Initialization: set the initial DH parameters and calculate the initial error; Iterate: calculate the Jacobian matrix J and the error vector e; Select the damping factor , solve the parameter update equation ; use Update DH parameters and recalculate errors; Adjust the damping factor according to the error change ; Repeat the iteration until the termination condition is met.

7. The robot arm connecting rod error compensation method based on LM iteration method according to claim 6, characterized in that: If the error decreases, accept the update and reduce Otherwise, increase and try again.

Citation Information

Cited By

  • Industrial robot calibration method and device based on LM algorithm of PID controller

    CN119927917A

  • Manipulator tail end high-precision dynamic compensation positioning method based on industrial vision

    CN121468571A

  • A high-precision dynamic compensation positioning method for the end of a manipulator based on industrial vision

    CN121468571B