Robust low-cost robot tool calibration method

Abstract

The invention relates to the technical field of robot tool calibration, in particular to a robust low-cost robot tool calibration method, which comprises the following steps: (1) according to the touch between the tail end of a robot tool and a key point, acquiring the relative position transformation of the tail end of a robot joint relative to a robot base coordinate system through a robot controller; (2) establishing coordinates of each key point on the calibration plate under a robot base coordinate system; (3) establishing an error function of robot tool calibration, wherein the error function comprises a distance error and a coplane error; and (4) optimizing and solving the TCP position by using an LM method according to the recorded robot joint end data, the real distance and the established error function. According to the scheme, the target acquisition mode is simple, the cost is low, the calculation result precision is high, and the method has certain robustness for noise existing in the calibration process; under the condition that noise exists, an approximate optimal solution can still be obtained through the scheme.

Description

Technical Field

[0001] The present invention relates to the technical field of robot tool calibration, and in particular to a robust and low-cost robot tool calibration method. Background Art

[0002] Robot tool calibration refers to the process of determining the transformation matrix of the tool coordinate system relative to the robot end link coordinate system. The purpose of this process is to facilitate the operator to flexibly adjust the posture of the robot end actuator and more accurately control the robot's motion trajectory. The most widely used method is to control the robot to make the tool end TCP point contact a fixed space point multiple times in different postures, record the rotation angles of each robot joint in contact, and use the robot kinematic position constraints to establish an equation relationship to solve the robot tool end position.

[0003] Existing solutions use specific needle-tip targets or plane devices, such as tool targets with fixed needle tips, or resistive screens with ranging functions, to obtain the true distance between all touch points; or metal plates with sensing capabilities that can send feedback signals to stop the robot from moving after the robot tool touches the plane. These solutions all require the customization of special targets, which are not only difficult to obtain but also costly. In addition, most of the existing solution methods are based on direct solution of equations. These methods can obtain accurate results under ideal conditions without noise and errors; however, due to the inevitable data errors in the actual process, the solution results may be biased.

[0004] Therefore, a new technical solution is urgently needed to solve the above technical problems. Summary of the invention

[0005] The purpose of the present invention is to overcome the above-mentioned problems of the prior art and to provide a robust, low-cost robot tool calibration method to solve the technical problem that the existing technology needs to customize special targets, which are not only difficult to obtain but also costly; and the existing solution methods are mostly based on direct solution of equations, which cannot be performed under ideal conditions without noise and error, and thus the solution results may be biased.

[0006] The above objectives are achieved through the following technical solutions:

[0007] A robust and low-cost robot tool calibration method includes a horizontally arranged calibration plate for the robot tool end to touch, wherein a plurality of key points are arranged on the calibration plate; and the method includes the following steps:

[0008] Step (1) According to the contact between the robot tool end and the key point, the relative position transformation of the robot joint end relative to the robot base coordinate system is obtained through the robot controller

[0009] Where i∈(1,n); n is the number of key points, n>5, and the positions of key points are not collinear;

[0010] Step (2) Establish the coordinates of each key point on the calibration plate in the robot base coordinate system As follows:

[0011]

[0012] Then we have:

[0013]

[0014] in, is the translation of the tool end relative to the robot joint end;

[0015] Step (3) Establish the error function E of robot tool calibration, including the distance error E d and coplanar error E p ; The distance error E d It is the difference between the actual distance and the calculated distance of two adjacent key points. The formula is as follows:

[0016]

[0017] Among them, d ij is the true distance between two adjacent key points;

[0018] The coplanar error is the sum of the squares of the distances from each key point to the plane, and the formula is as follows:

[0019]

[0020] Then we have:

[0021] E=E d +λE p

[0022] Among them, λ is the weight factor used to balance the distance error and the coplanarity error;

[0023] Step (4) Based on the recorded robot joint end data True distance d ij And the error function E = E d +λE p , using the LM method to optimize and solve, we get

[0024] Furthermore, the calibration plate is one of a checkerboard calibration plate, a dot calibration plate or a grating glass Chart.

[0025] Furthermore, in the coplanar error, it is assumed that all key points are located in the same horizontal plane, and the plane equation is ax+by+cz+d=0.

[0026] The present invention provides a robust and low-cost robot tool calibration method, which can calibrate the tool TCP using a common calibration plate without the need to customize a special target. At the same time, in the process of solving the tool TCP, an optimized error function is established by using the principle that the distances of all contact points are known and coplanar. The solution is solved in an optimized manner, which is simple to operate, high in accuracy, and robust to noise; in the presence of noise, this solution can still obtain an approximate optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a flow chart of a robust and low-cost robot tool calibration method according to the present invention. DETAILED DESCRIPTION

[0028] The present invention is further described in detail below based on the accompanying drawings and embodiments. The described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] The present solution provides a robust, low-cost robot tool calibration method, including a horizontally arranged calibration plate for the end of the robot tool to touch, and a number of key points are arranged on the calibration plate; it should be noted that in robot welding or grasping scenarios guided by vision, because hand-eye calibration is required, calibration plates are generally available on site, so they are easy to obtain and low in cost; and the calibration plate in this embodiment selects one of the common checkerboard calibration plates, dot calibration plates or grating glass Chart.

[0030] like Figure 1 As shown, this scheme includes the following steps:

[0031] Step (1) Move the tool end of the robot to the key point position on the plane calibration plate in different postures in turn, and obtain the relative position transformation of the robot joint end relative to the robot base coordinate system through the robot controller according to the contact between the robot tool end and the key point

[0032] Where i∈(1,n); n is the number of key points, n>5, and the positions of key points are not collinear;

[0033] Step (2) Establish the coordinates of each key point on the calibration plate in the robot base coordinate system As follows:

[0034]

[0035] Then we have:

[0036]

[0037] in, is the translation of the tool end relative to the robot joint end, that is, the TCP position, which is the quantity to be determined;

[0038] Step (3) Based on the fact that the actual distances between key points on the calibration plate are known and all points are located in the same plane, the error function E of the robot tool calibration is established, including the distance error E d and coplanar error E p ;

[0039] Specifically, the distance error E d It is the difference between the actual distance and the calculated distance of two adjacent key points. The formula is as follows:

[0040]

[0041] Among them, d ij is the real distance between two adjacent key points, which can be obtained according to the point index and the specifications of the calibration plate;

[0042] Assuming that all key points are located on the same horizontal plane, the plane equation is ax+by+cz+d=0. For each key point, the plane equation should be satisfied, and the coplanar error is the sum of the squares of the distances from each key point to the plane, and the formula is as follows:

[0043]

[0044] Then we have:

[0045] E=E d +λE p

[0046] Among them, λ is the weight factor used to balance the distance error and the coplanarity error;

[0047] Step (4) Based on the robot joint end data recorded in the above steps True distance d ij And the error function E = E d +λE p , using the LM method to optimize and solve, we get

[0048] It should be noted that the LM (Levenberg-Marquard) method used in this embodiment is an optimization algorithm for nonlinear least squares problems, which combines the advantages of the gradient descent method and the Gauss-Newton method, and introduces a parameter to balance the trade-off between the two, thereby providing better convergence performance. In the early stage of the algorithm, more emphasis is placed on the gradient descent method to explore the solution space; in the later stage, more emphasis is placed on the Gauss-Newton method for rapid convergence. The steps include: initializing parameter values; calculating the Jacobian matrix and Hessian matrix of the error function; calculating the next update step size according to the current Hessian matrix and the method of dynamically adjusting the error function of the algorithm; judging the termination condition, such as reaching the maximum number of iterations or the residual sum of squares is less than the threshold, the algorithm terminates.

[0049] Therefore, in the solution process, by establishing an error function and using the LM iterative optimization method, an approximate optimal solution can be obtained in the presence of noise, so this method is robust.

[0050] The above description is only for illustrating the implementation mode of the present invention and is not intended to limit the present invention. For those skilled in the art, any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A robust and low-cost robot tool calibration method, characterized in that The method comprises a horizontally arranged calibration plate for the end of the robot tool to touch, wherein a plurality of key points are arranged on the calibration plate; and the method comprises the following steps: Step (1) According to the contact between the robot tool end and the key point, the relative position transformation of the robot joint end relative to the robot base coordinate system is obtained through the robot controller Where i∈(1,n); n is the number of key points, n>5, and the positions of key points are not collinear; Step (2) Establish the coordinates of each key point on the calibration plate in the robot base coordinate system As follows: Then we have: in, is the translation of the tool end relative to the robot joint end; Step (3) Establish the error function E of robot tool calibration, including the distance error E d and coplanar error E p ; The distance error E d It is the difference between the actual distance and the calculated distance of two adjacent key points. The formula is as follows: Among them, d ij is the true distance between two adjacent key points; The coplanar error is the sum of the squares of the distances from each key point to the plane, and the formula is as follows: Then we have: And=And d +λE p Among them, λ is the weight factor used to balance the distance error and the coplanarity error; Step (4) Based on the recorded robot joint end data True distance d ij And the error function E = E d +λE p , using the LM method to optimize and solve, we get 2. A robust and low-cost robot tool calibration method according to claim 1, characterized in that: The calibration plate is one of a checkerboard calibration plate, a dot calibration plate or a grating glass Chart.

3. A robust and low-cost robot tool calibration method according to claim 1, characterized in that: In the coplanar error, it is assumed that all key points are located in the same horizontal plane, and the plane equation is ax+by+cz+d=0.

Images