CN109465831B - Method for improving calibration precision of tool coordinate system of industrial robot - Google Patents

Abstract

The invention discloses a method for improving the calibration precision of an industrial robot tool coordinate system, and belongs to the field of industrial robot application control. In the method, in the process of calibrating a common four-point method tool coordinate system, precision evaluation is given to each point to be calibrated, when the precision meets the requirement, the next calibration point can be calibrated, and after the calibration points are recorded, a total precision evaluation is given to the final calibration result. The method is convenient for a user to teach out a reasonable calibration point, eliminates human errors caused by the user in the calibration process to obtain a calibration result of a high-precision tool coordinate system, and further eliminates the motion problems of low track tracking precision and the like caused by inaccurate calibration of the tool coordinate system. The tool coordinate system calibration lifting algorithm provided by the invention does not need to add any calibration device, and is simple and convenient to operate.

Description

Method for improving calibration precision of tool coordinate system of industrial robot

Technical Field

The invention relates to a method for calibrating an industrial robot tool coordinate system, in particular to a method for improving the calibration precision of the industrial robot tool coordinate system.

Background

In the application of an industrial robot, the robot needs to install corresponding tools, such as a welding gun, a gripper, a sucker and the like, at the tail end to complete a specific work task. For teaching and programming convenience, it is often necessary to calibrate the tool coordinate system, moving the end points of the robot from the flange center to the end points of the tool.

In the industrial robot industry, a common tool coordinate system calibration method is a four-point method, a mobile robot aligns a tail end point of a tool at the tip of a fixed needle-shaped sharp object (calibration device), records poses of the robot at four different postures when the tail end point of the tool is aligned with the tip, and calibrates a tool coordinate system position parameter by using a least square method, wherein the position parameter is shown as a formula (1). (reference: Liu Cheng, Liwenguang, et al. A robotic tool coordinate system calibration method [ J ]. Shandong science, 2012,25(1): 69-74).

In the formula, p_x，p_y，p_zFor tool coordinate system position parameter calibration results, R₁～R₄The rotation matrix corresponding to four calibration points, (e)_x1,e_y1,e_z1)～(e_x4,e_y4,e_z4) The position coordinates corresponding to the four calibration points.

The calibration precision of the tool coordinate system directly influences the track precision of the robot during operation. The calibration method of the four-point method needs to manually align the tips of the calibration device, and needs to ensure that the four points are all aligned at the same position of the tips, so that certain errors are inevitable, and the final calibration precision is difficult to ensure.

Disclosure of Invention

The invention aims to solve the technical problem that the defects in the prior art are overcome, and provides a method for improving the calibration precision of an industrial robot tool coordinate system. The method aims to obtain a high-precision tool coordinate system calibration result and further solve the problem of low track tracking precision caused by inaccurate tool coordinate system calibration.

The basic idea of the invention is as follows: the mobile robot enables the tail end point of the tool to be aligned with the tip of the calibration device, and the first point is recorded in a teaching mode; and then changing the posture of the robot, and still keeping the tail end point of the tool aligned with the tip of the calibration device, wherein in the process, a calibration lifting algorithm Part1 gives a current position in real time as precision evaluation of the calibration point, when the evaluation score is higher than 60 minutes, teaching and recording the calibration point can be successful, otherwise, the robot fails, and the robot needs to be continuously moved until the evaluation score meets the requirement, the calibration point can be taught and recorded, and the teaching of the next calibration point is carried out until the four calibration points are finished. After the teaching of the four calibration points is finished, a calibration lifting algorithm Part2 gives out a final calibration evaluation, if the precision is met, parameters of a tool coordinate system are given out, and the calibration is finished; if the precision is not satisfied, the calibration of the tool coordinate system fails, and the calibration by the four-point method needs to be carried out again.

It can be seen from equation (1) that if the rotation matrices of the four calibration points are similar, and in the extreme case, if the same rotation matrix exists, the denominator matrix has 0 entries and no solution. Based on the above, the calibration lifting algorithm Part1 adopts a quaternion description method to evaluate the similarity of the rotation matrixes of the calibration points, and gives an accuracy evaluation score.

The invention adopts the technical scheme for realizing the purpose of the invention: a method for improving the calibration precision of an industrial robot tool coordinate system comprises the following steps:

step 1, recording the attitude of the first index point (A)₁,B₁,C₁) Is converted into a rotation matrix R₁Is then converted into quaternion Q₁Description of (1):

(A₁,B₁,C₁)→R₁→Q₁ (2)

step 2, recording the current gesture of the point to be calibrated (A)₂,B₂,C₂) Is converted into a rotation matrix R₂Is then converted into quaternion Q₂Description of (1):

(A₂,B₂,C₂)→R₂→Q₂(3)

and 3, calculating the rotation angle theta of the current point to be calibrated relative to the last calibration point, wherein the range is 0-180 degrees:

θ＝f(Q₁,Q₂) (4)

step 4, calculating the precision evaluation score gamma:

the accuracy evaluation score is a piecewise function of a rotation angle theta, where theta₀、θ₁、θ₂This value is a set limit value for the angle of rotation, which is dependent on the model.

And judging whether the calibration precision of the point meets the requirement according to whether the precision evaluation score gamma meets a given standard value, and returning to the step 2 if the calibration precision of the point does not meet the requirement. If the accuracy requirement is met, the next step is performed.

And 5, selecting the next point to be calibrated, and circularly executing the steps 2, 3, 4 and 5 until the operation of all four calibration points is finished.

The first step part1 is referred to above as the nominal accuracy improvement, and the second part2 is referred to below as the nominal accuracy improvement.

Step 6, calibrating the precision improving algorithm: the recorded four calibration points are comprehensively evaluated by using the standard deviation, so that the position error caused by aligning the human eyes with the tip of the calibration device is eliminated, and the final calibration result evaluation is given, and the method comprises the following steps:

step 6.1, according to the recorded four calibration points, four calibration results are calculated:

in the formula (p)_x1,p_y1,p_z1)～(p_x4,p_y4,p_z4) And calibrating results for four groups of tool coordinate systems.

Step 6.2, calculating the standard deviation of the four calibration results:

in the formula (I), the compound is shown in the specification,

as the mean value, σ, of the components of the calibration result_x,σ_y,σ_zThe standard deviation of each component of the calibration result is shown.

Step 6.3 final calibration result evaluation σ:

when the sigma is within a certain range, the calibration precision is considered to meet the requirement, the calibration is successful, and the position parameter of the tool coordinate system is given as

According to the method for improving the calibration precision of the tool coordinate system of the industrial robot, the rotation relation of the calibration points is evaluated by adopting a quaternion method in each calibration step, so that precision evaluation is given, a user can conveniently teach reasonable calibration points, precision errors caused by the calibration points in each step are eliminated, the precision of a calibration result is ensured, and further the track precision is ensured.

According to the method, after teaching of all the calibration points is finished, a total precision evaluation is added, so that the operation error of aligning the calibration tip is eliminated, and the final precision of the calibration result is ensured.

The method of the invention does not need to add any calibration system or device, only adds a calibration algorithm in the most basic calibration process of the four-point method, realizes the software part, has no difference with the common four-point method for users, and is simple and easy to operate.

Drawings

FIG. 1 is a flow chart of the method of the present invention for improving accuracy.

Fig. 2 is a schematic diagram of the accuracy evaluation.

Detailed Description

The process of the present invention will be described in further detail with reference to examples.

Example (b):

taking a certain universal six-joint as an experimental object, selecting specific accuracy evaluation sub-standards as follows, as shown in the attached figure 2.

1. Teaching records the first calibration point as follows: (655.95, -302.35,394.36,100.29,19.93,160.18);

2. and finally selecting two, three, four calibration points as follows according to the real-time evaluation score obtained by the precision evaluation standard: (600.91, -302.35,305.42,94.02,21.98,142.77), (633.46, -391.49,305.42,109.86,21.98,142.77), (655.95, -188.78,401.75,104,46.05,166.82), and the evaluation scores were 82, 80, and 100, respectively.

3. Four calibration results are calculated according to the four calibration points:

①：(99.9873,200.007,300.0368)；

②：(99.95417,200.0286,300.0437)；

③：(99.98198,200.040798,300.042966)；

④：(99.95656,200.0258699,300.0429449)。

4. calculating the standard deviation:

σ_x＝0.017082

σ_y＝0.0139089

σ_z＝0.003208

5. calculating the evaluation of the final calibration result:

σ＝0.0223

and (3) the final calibration result is far smaller than the set requirement (2mm), the precision of the calibration result is considered to be higher, the calibration is successful, and finally the position parameters of the tool coordinate system are obtained as follows: 99.9700,200.0256,300.0416.

Claims (1)

1. A method for improving the calibration precision of an industrial robot tool coordinate system comprises the following steps:

step 1, recording the attitude of the first index point (A)₁,B₁,C₁) Is converted into a rotation matrix R₁Is then converted into quaternion Q₁Description of (1):

(A₁,B₁,C₁)→R₁→Q₁

step 2, recording the current gesture of the point to be calibrated (A)₂,B₂,C₂) Is converted into a rotation matrix R₂Is then converted into quaternion Q₂Description of (1):

(A₂,B₂,C₂)→R₂→Q₂

and 3, calculating the rotation angle theta of the current point to be calibrated relative to the last calibration point, wherein the range is 0-180 degrees:

θ＝f(Q₁,Q₂)；

step 4, calculating the precision evaluation score gamma:

wherein theta is₀、θ₁、θ₂For a set boundary value of the angle of rotation

Judging whether the calibration precision of the point meets the requirement according to whether the precision evaluation score gamma meets a given standard value, and returning to the step 2 if the calibration precision of the point does not meet the requirement; if the precision requirement is met, executing the next step;

step 5, selecting the next point to be calibrated, and executing the steps 2, 3, 4 and 5 in a circulating way until the operation of all four calibration points is finished;

and 6, calibrating the precision improvement algorithm, comprising the following steps:

step 6.1, according to the recorded four calibration points, four calibration results are calculated:

in the formula (p)_x1,p_y1,p_z1)～(p_x4,p_y4,p_z4) Calibrating results for the four groups of tool coordinate systems; r₁～R₄The rotation matrix corresponding to four calibration points, (e)_x1,e_y1,e_z1)～(e_x4,e_y4,e_z4) Position coordinates corresponding to the four calibration points;

step 6.2, calculating the standard deviation of the four calibration results:

in the formula (I), the compound is shown in the specification,

as the mean value, σ, of the components of the calibration result_x,σ_y,σ_zThe standard deviation of each component of the calibration result is obtained;

step 6.3 final calibration result evaluation σ:

when the sigma is within a certain range, the calibration precision is considered to meet the requirement, the calibration is successful, and the position parameter of the tool coordinate system is given as

Images