JPH04259010A - Adverse conversion method for robot - Google Patents

Abstract

PURPOSE:To precisely decide the position of a tool of a robot having an offset wrist with the adverse conversion calculation. CONSTITUTION:The position and the attitude of a desired tool are set (S1), and the adverse conversion calculation A is carried out (S2) based on the position and the attitude of the tool. The calculation A is identical with a repetitive solution, and the joint angles J4, J5 and J6 of a wrist axis are obtained (S3 and S4) through the convergence as the approximate solutions. Then the adverse conversion calculation B is carried out based on the angles J4-J6 and the preceding position of the tool. Thus, the joint angles J1, J3 and J5 of a basic axis of a robot are obtained (S7) through the calculation B.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an inverse transformation method for a robot for determining the joint angle of each joint axis from the position and orientation of a robot tool, and more particularly to an inverse transformation method for a robot having an offset wrist.

0002

2. Description of the Related Art Inverse transformation calculation is a method for determining the joint angle of each joint axis of the robot from the desired position and orientation of a tool given the position and orientation of the robot. This inverse transformation calculation is performed by iterative calculation (iterative solution method) when the robot has an offset wrist.
The joint angles found are approximate solutions.

0003

For this reason, when the position and orientation of the tool are determined from the joint angles, the position and orientation do not accurately realize the desired position and orientation of the tool. On the other hand, when a robot is used to perform work, it is necessary that at least the position of the tool does not deviate from the desired position. The present invention has been made in view of these points, and it is an object of the present invention to provide a robot inverse conversion method that can accurately realize the position of a given tool even in the case of a robot with an offset wrist. shall be.

0004

[Means for Solving the Problems] In order to solve the above problems, the present invention provides a robot inverse transformation method for determining the joint angle of each joint axis of a robot having an offset wrist from the position and orientation of a tool of the robot. , perform a first inverse transformation calculation based on the position and orientation of the tool, converge the first inverse transformation calculation by an iterative solution method, obtain the joint angle of the wrist axis of the robot, and calculate the joint angle of the wrist axis of the robot. Provided is a robot inverse transformation method, characterized in that a second inverse transformation calculation is performed based on the angle and the position of the tool, and a joint angle of a fundamental axis of the robot is determined by the second inverse transformation calculation. be done.

[0005]

[Operation] A desired tool position and orientation are given, and the first inverse transformation calculation is performed based on the tool position and orientation. This first inverse transformation calculation is an iterative solution method, and by convergence, the joint angle of the wrist axis is obtained as an approximate solution. Next, a second inverse transformation calculation is performed based on the joint angle of the wrist axis and the initially given tool position. Through this second inverse transformation calculation, the joint angle of the basic axis of the robot is determined as an analytical solution. The position of the tool is accurately realized by the joint angle of the wrist axis and the joint angle of the basic axis.

[0006]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 2 is a diagram showing the appearance of a robot in which the present invention is implemented. In the figure, a robot 1 is a six-axis articulated robot, and is composed of a basic axis 2 and a wrist axis 3. The basic axis 2 consists of three joint axes: θ, W, and U. The wrist axis 3 is
As shown in the figure, it is an offset wrist, and there are three
Consists of joint axes. A tool 4 is attached to the α axis located at the tip of the wrist axis. The robot control device 10 has a microprocessor configuration and controls the robot 1 by driving a servo motor of the robot 1. Robot control device 1
0 is provided with a teaching operation panel (not shown). When data on the position and orientation of the tool 4 desired by the operator is input from this teaching operation panel, the robot control device 1
0 performs an inverse transformation calculation, which will be described later, based on the position and orientation data, and calculates each joint angle of the robot 1. The robot 1 realizes the position of the tool 4 by its joint angles.

Next, a method for determining each joint angle will be explained. FIG. 1 is a diagram showing a flowchart of the robot inverse conversion method of the present invention. In the figure, the number following S indicates the step number. [S1] Data regarding the position and orientation of the tool 4 is input. [S2] Based on the position and orientation data of the tool 4,
Perform inverse transformation calculation A. This inverse transformation calculation A is performed by an iterative solution method. [S3] Find the joint angle J4 of the γ-axis by inverse transformation calculation A,
It is determined whether the joint angle J4 has converged to an approximate solution. If it has not converged, return to S2, and if it has converged, return to S4
Proceed to. [S4] The β-axis joint angle J5 and the α-axis joint angle J6 are sequentially determined as approximate solutions. [S5] Joint angles J4, J5, and J6 are output once. [S6] Joint angles J4, J5, and J6 are input. Also,
The data on the position of the tool 4 input in S1 is input again. [S7] Inverse transformation calculation B is performed based on the joint angles J4, J5, J6 and the position data of the tool 4. This inverse transformation calculation B is analytically possible. [S8] Through inverse transformation calculation B, each joint angle J1, J2, J3 of the θ-axis, W-axis, and U-axis is determined as an analytical solution, and the joint angles J1, J2, and J3 are output. In this way, first, the joint angle J4 of the wrist axis 3 is determined as an approximate solution by the inverse transformation calculation A, and the joint angles J5 and J6 of the wrist axis 3 are determined based on the approximate solution J4. Next, inverse transformation calculation B is performed based on the joint angles J4, J5, J6 of the wrist axis and the position of the tool 4. Through the inverse transformation calculation B, the joint angles J1, J2, and J3 of the basic axis 2 are obtained as analytical solutions. Therefore, the position of the tool 4 does not deviate from the desired position. That is, the robot 1 with an offset wrist
Even in this case, the given position of the tool 4 can be realized accurately. In the above description, the invention was applied to a 6-axis articulated robot, but it can be similarly applied to articulated robots other than 6-axes.

[0008]

[Effects of the Invention] As explained above, in the present invention, first,
The joint angle of the wrist axis is obtained as an approximate solution by the first inverse transformation calculation, and then the second inverse transformation calculation is performed based on the joint angle of the wrist axis and the position of the tool. The joint angle of the basic axis is determined as an analytical solution by calculation. Therefore, the position of the tool does not deviate from the desired position. That is, even in the case of a robot with an offset wrist, the position of a given tool can be realized accurately.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a flowchart of a robot inverse conversion method according to the present invention.

FIG. 2 is a diagram showing the appearance of a robot in which the present invention is implemented.

[Explanation of symbols]

1 Robot 2 Basic axis 3 Wrist axis 4 Tools 10 Robot control device

Claims (4)

[Claims] 1. A robot inverse transformation method for determining joint angles of respective joint axes of the robot from the position and orientation of a tool of a robot having an offset wrist, comprising: performing a first inverse transformation based on the position and orientation of the tool; do the calculations,
converging the first inverse transformation calculation by an iterative solution method to obtain a joint angle of the wrist axis of the robot, and performing a second inverse transformation calculation based on the joint angle of the wrist axis and the position of the tool; A robot inverse transformation method, characterized in that a joint angle of a fundamental axis of the robot is determined by the second inverse transformation calculation. 2. The joint angle of the wrist axis is obtained as an approximate solution of the first inversion calculation, and the joint angle of the basic axis is obtained as an analytical solution of the second inversion calculation. The robot inverse transformation method according to claim 1. 3. The robot inverse conversion method according to claim 1, wherein the wrist axes are an α-axis, a β-axis, and a γ-axis, and the basic axes are a θ-axis, a W-axis, and a U-axis. 4. The joint angle of the wrist axis is determined by first determining the γ-axis by a convergence calculation, and determining the β-axis and the α-axis based on an approximate solution of the γ-axis determined by the convergence calculation. The robot inverse transformation method according to claim 2, characterized in that: