Method of determining control position of robot and robot system

1. A method for determining a control position of a robot, comprising:

(a) acquiring N real reference positions in a real space and N control reference positions in a robot control coordinate system when a control point of the robot is located in the respective N real reference positions, N being an integer equal to or larger than three;

(b) setting M figures having vertices in a plurality of real reference positions of the N real reference positions within the real space, and obtaining a transform function expressing a correspondence relationship between a real position and a control position within each figure, M being an integer equal to or larger than one;

(c) receiving input of a target position of the control point in the real space;

(d) selecting an object figure for calculation of a control position for the target position from the M figures; and

(e) calculating a target control position as a control position in the robot control coordinate system for the target position using the transform function with respect to the object figure, wherein

(d) includes selecting a figure closest to the target position of the M figures as the object figure when the target position is not contained in any of the M figures.

2. The method according to claim 1 , wherein

M is an integer equal to or larger than two,

(d) includes selecting two or more figures closest to the target position as the object figures when the target position is not contained in any of the M figures, and

(e) includes:

(e1) calculating a corresponding target control position as a control position for a corresponding target position corresponding to the target position in each of the two or more object figures; and

(e2) calculating the target control position for the target position by interpolation using the corresponding target control positions calculated in the two or more object figures.

3. The method according to claim 1 , wherein

(d) includes: when the target position is not contained in any of the M figures,

(d1) setting a virtual real reference position in which a real position and a control position coincide;

(d2) setting a virtual figure containing the target position using the virtual real reference position and a plurality of real reference positions selected from the N real reference positions, and selecting the virtual figure as the object figure; and

(d3) obtaining a transform function expressing a correspondence relationship between a real position and a control position within the virtual figure.

4. The method according to claim 1 , wherein

M is an integer equal to or larger than two, and

the M figures are M solid figures closely placed in the real space.

5. The method according to claim 4 , wherein

(d) includes selecting one or more figures closest to the target position as the object figures when the target position is not contained in any of the M figures, and

(e) includes:

(e1) calculating a corresponding target control position as a control position for a corresponding target position corresponding to the target position in each of the one or more object figures; and

(e2) calculating the target control position for the target position by extrapolation using the corresponding target control positions calculated in the one or more object figures.

6. The method according to claim 1 , further comprising:

(i) obtaining a position control error when controlling the robot using the target control position;

(ii) acquiring (N+1) or more real reference positions and (N+1) or more control reference positions by adding one or more new real reference positions and one or more new control reference positions when the position control error exceeds a predetermined threshold; and

(iii) increasing the M by executing (b) again using the (N+1) or more real reference positions and the (N+1) or more control reference positions.

7. The method according to claim 1 , wherein

one of the real reference position and the control reference position is selected by a worker.

8. The method according to claim 1 , wherein

(a) includes obtaining N real reference postures as postures of the robot in the N real reference positions and N target control postures as control postures in the N target control positions,

at (b), the transform function is generated to also express a relationship between a real posture and a control posture within each figure of the M figures,

(c) includes receiving a target posture with the target position, and

(e) includes calculating a target control posture as a control posture in the robot control coordinate system for the target posture using the transform function.

9. A robot system determining a control position of a robot, comprising:

a robot; and

a control unit that controls the robot, wherein

the control unit executes

(a) processing of acquiring N real reference positions in a real space and N control reference positions in a robot control coordinate system when a control point of the robot is located in the respective N real reference positions, N being an integer equal to or larger than three;

(b) processing of setting M figures having vertices in a plurality of real reference positions of the N real reference positions within the real space, and obtaining a transform function expressing a correspondence relationship between a real position and a control position within each figure, M being an integer equal to or larger than one;

(c) processing of receiving input of a target position of the control point in the real space;

(d) processing of selecting an object figure for calculation of a control position for the target position from the M figures; and

(e) processing of calculating a target control position as a control position in the robot control coordinate system for the target position using the transform function with respect to the object figure, wherein

(d) includes selecting a figure closest to the target position of the M figures as the object figure when the target position is not contained in any of the M figures.

10. A method for determining a control position of a robot, comprising:

(a) acquiring N real reference positions in a real space and N control reference positions in a robot control coordinate system, each of the N real reference positions corresponding to a control point of the robot where the robot is actually located while the robot moves in the real space, the N control reference positions corresponding to designed positions through which the robot is supposed to move, N being an integer equal to or larger than three;

(b) setting M figures having vertices in a plurality of real reference positions of the N real reference positions within the real space, and obtaining a transform function expressing a correspondence relationship between a real position and a control position within each figure, each of the M figures being in a triangular shape, the M figures being obtained by dividing a rectangular figure having the vertices, M being an integer equal to or larger than one;

(c) receiving input of a target position of the control point in the real space, the target position being a position to which the robot is supposed to reach;

(d) selecting an object figure for calculation of a control position for the target position from the M figures;

(e) calculating a target control position as a control position in the robot control coordinate system for the target position using the transform function with respect to the object figure; and

(f) describing a motion command for the robot based on the calculated target control position.