JPH01205988A - Robot control method - Google Patents

Abstract

PURPOSE:To reduce the burden of an operator during teaching by applying a dead-weight compensating torque to an articulation subjected to affection of the dead-weight from an actuator and applying a weak positional servo as a present position desired value, and by changing the desired value toward an actual articulation position when the desired value and the actual articulation position exceed a certain value. CONSTITUTION:A sample hold circuit 2 holds a value from an instant shaft position sensor 3 when it receives a teaching operation initiating signal 1. A position control section 4 constitutes a weak position servo system using the hold value as a desired value and the value from the sensor 3 as a feed-back value. Further, a dead-weight compensation computing section 5 receives the value from the sensor 3, and calculates and delivers a present dead-weight compensating value 6 for the shaft with the use of the value from the sensor 3, and an another shaft position signal 10. The sum of the output value and an output value of the control section 4 is outputted as an instruction value for a motor driver 9. Further, when the deviation between the desired value and the sensor 4 exceeds a predetermined value, a circuit 14 uses a present position as a new desired value.

Description

[Detailed Description of the Invention] [Objective of the Invention] (Industrial Application Field) The present invention is directed to the teaching time of a robot device, such as an industrial robot, which has joints that are affected by its own weight and has a teaching playback function. This is a robot control method that reduces the burden on people.

(Prior Art) A robot device has joints like an industrial robot and has a teaching playback function.

Conventionally, during teaching, the servo system of the actuator is turned off (servo free), and the operator holds the tip of the robot and backdrives the robot to the desired position and posture.

Since each axis of the robot is affected by its own weight, except for the rotation axes of horizontally placed Cartesian robots and square robots, the operator must support this weight during teaching.

As a means to solve this problem, as shown in FIG.
This detection signal is used as the self-weight compensation calculation section 5 and the teaching position output 13. The output of this self-weight compensation calculating section 5 is input to a motor drive 9 to drive a motor 11. In other words, the influence of gravity acting on each joint is calculated using the signal of the position detection sensor of each joint, or if there is no influence of other axes and no influence of posture change, it is calculated in advance, and the actuator for each joint calculates this effect. Outputs torque to compensate for However, with this method, there is a possibility of drifting in one direction due to changes in mechanical conditions, identification errors, and calculation errors associated with numerical calculations.

(Problem to be solved by the invention) As mentioned above, in the conventional control method during teaching,
Operators have to work while supporting the weight of the joints, resulting in poor operability. Furthermore, compensation by the actuator also has the possibility of drift.

This invention was made in view of the above-mentioned problems, and uses the self-weight calculated in advance for each joint based on the signal of the position detection sensor, or when there is no influence from other axes and no influence from changes in posture. In addition to compensating with the actuator of
Drifts caused by changes in mechanical conditions, identification errors, and self-weight calculation errors associated with numerical calculations are suppressed by configuring a weak position servo system that uses the current position as the target value, and furthermore, it If the deviation from the position exceeds a certain value, it is determined that the operation is in that direction, and by changing the new target value in that direction, the influence of the spring system due to the position servo is removed, and the operator's teaching work is improved. The purpose is to reduce the burden.

[Structure of the invention]

(Means for Solving the Problems) The robot control method of this application includes joints that are affected by the weight of at least one axis, actuators for driving these joints, sensors for detecting the positions of the joints, and teaching. In a robot device consisting of a control device with a playback function, an actuator for an optical intersection outputs torque to compensate for the weight of the joints that are affected by the weight due to the position of the joint during teaching, and
A weak position servo is applied with the current position of the optical internode as the target value, and when the deviation between the target value and the actual joint position exceeds a certain value, the target value is changed to the actual joint position side, and during teaching. This reduces the burden on the robot and suppresses drift caused by mechanical state changes, identification errors, and self-weight calculation errors associated with numerical calculations.

(Function) In the robot control method configured in this way, the self-weight calculated in advance is calculated based on the signal from the position detection sensor, or when there is no influence from other axes and no influence from changes in posture, and the self-weight calculated in advance is applied to the actuator for each joint. In addition to compensating for drift caused by self-weight calculation errors and changes in mechanical conditions, by configuring a weak position servo system that targets the current position, the burden on the operator's teaching work can be reduced. .

(Example) The invention of this application will be described in detail below using examples shown in the drawings. FIG. 1 is a block diagram showing one embodiment of the present invention. Upon receiving the signal 1 to start the teaching operation, the sample hold circuit 2 holds the value of the position detection sensor 3 of the axis. The position control unit 4 constitutes a weak position servo system that uses this held value as a target value and uses the value of the position detection sensor 3 as a feedback value.

Further, the value of the position detection sensor 3 is also taken into the self-weight compensation calculation section 5, and the self-weight compensation calculation section 5 uses the value of the position detection sensor 3 and the position signal 10 from the other axis to calculate the current value of the axis. Calculate and output the self-weight compensation amount 6. The sum of this output value and the output value 7 of the position control section 4 is output as a command value to the motor driver 9 of the axis.

Here, when the load 12 is moved by the teaching work and the deviation between the position target value and the position detection sensor 3 which is the feedback value exceeds a certain value, the position deviation scene judgment circuit 14 causes the sample hold circuit 2 to Outputs a command to set the current position as the new target value again.

In a robot with a control system configured in this way, during teaching, the actuator for the joint outputs a torque that compensates for the weight of the joint that is affected by its own weight based on the position of the joint, and the current position of the joint is set as the target value. By applying a weak position servo to reduce the burden during teaching, it is possible to suppress drift caused by mechanical state changes, identification errors, and self-weight calculation errors associated with numerical calculations.

Further, the decrease in operability caused by the position servo can be eliminated by changing the target value using the position deviation amount determination circuit.

〔Effect of the invention〕

As described above, this invention comprises joints that are affected by the weight of at least one axis, actuators for driving these joints, sensors for detecting the positions of the joints, and a control device with a teaching playback function. In robotic equipment.

During teaching, the actuator for the joint outputs a torque that compensates for the weight of the joint that is affected by the weight due to the position of the joint, and a weak position servo is applied to set the current position target value of the joint. When the joint position exceeds a certain value, the target value is changed to the actual joint position to reduce the burden during teaching, and to reduce drift due to mechanical state changes, identification errors, and self-weight calculation errors associated with numerical calculations. It has the characteristic of suppressing

[Brief explanation of the drawing]

FIG. 1 is a block diagram of an embodiment of the present invention, and FIG. 2 is a block diagram when only self-weight compensation is performed. 1... Teaching start signal 2... Sample hold circuit 3... Position detection sensor 4... Position control section 5...
・Self-weight compensation calculation unit 6...Self-weight compensation value 7...Position servo output value 8...Motor output command value 9...Motor driver 10...Other axis position signal 11...Motor 121...・Load 13...Teaching position output 14...Position deviation amount judgment circuit agent Patent attorney Noriyuki Chika Yudo Mitsuru Matsuyama

Claims (1)

[Claims] A robot equipped with joints that are affected by its own weight on at least one axis, actuators for driving these joints, sensors for detecting the positions of the joints, and a control device with a teaching playback function. In the device, during teaching, the actuator for the joint outputs a torque that compensates for the weight of the joint that is affected by the weight of the joint, and also applies a weak position servo that uses the current position of the joint as the target value. When the deviation between the target value and the actual joint position exceeds a certain value, the target value is changed to the actual joint position side, reducing the burden during teaching, and reducing mechanical state changes and identification errors.
A method for controlling a robot, characterized in that a robot is subject to drift due to self-weight calculation errors associated with numerical calculations.