Controller and machine learning device
A machine learning device of a controller observes, as state variables expressing a current state of an environment, teaching position compensation amount data indicating a compensation amount of a teaching position in control of a robot according to the teaching position and data indicating a disturbance value of each of the motors of the robot in the control of the robot, and acquires determination data indicating an appropriateness determination result of the disturbance value of each of the motors of the robot in the control of the robot. Then, the machine learning device learns the compensation amount of the teaching position of the robot in association with the motor disturbance value data by using the observed state variables and the determination data.
1. A controller that determines a compensation amount of a teaching position in control of a robot according to the teaching position included in teaching data, the controller comprising:
a machine learning device that learns a compensation amount of the teaching position in the control of the robot according to the teaching position, wherein
the machine learning device has
a state observation section that observes, as state variables expressing a current state of an environment, teaching position compensation amount data indicating the compensation amount of the teaching position in the control of the robot according to the teaching position, motor disturbance value data indicating a disturbance value of each of motors of the robot in the control of the robot, and teaching position data including the teaching position of the teaching data,
a determination data acquisition section that acquires determination data indicating an appropriateness determination result of the disturbance value of each of the motors of the robot in the control of the robot, and
a learning section that learns the compensation amount of the teaching position of the robot in association with the motor disturbance value data and the teaching position by using the state variables and the determination data.
2. The controller according to claim 1 , wherein
the determination data includes, besides the appropriateness determination result of the disturbance value of each of the motors of the robot in the control of the robot, at least any of an appropriate determination result of a teaching position to which the robot finally moves, an appropriateness determination result of a value detected by a sensor, and an appropriateness determination result of cycle time in the control of the robot according to the teaching position after compensation.
3. The controller according to claim 1 , wherein
the learning section has
a reward calculation section that calculates a reward associated with the appropriateness determination result, and
a value function update section that updates by using the reward a function expressing a value of the compensation amount of the teaching position relative to the disturbance value of each of the motors of the robot in the control of the robot.
4. The controller according to claim 1 , wherein
the learning section performs calculation of the state variables and the determination data on the basis of a multilayer structure.
5. The controller according to claim 1 , further comprising:
a decision-making section that outputs a command value based on the compensation amount of the teaching position in the control of the robot according to the teaching position on a basis of a learning result of the learning section.
6. The controller according to claim 1 , wherein
the learning section learns the compensation amount of the teaching position in the control of the robot according to the teaching position in each of a plurality of robots by using the state variables and the determination data obtained for each of the plurality of robots.
7. The control according to claim 1 , wherein
the machine learning device exists in a cloud server.
8. A machine learning device that learns a compensation amount of a teaching position in control of a robot according to the teaching position included in teaching data,
the machine learning device comprising:
a state observation section that observes, as state variables expressing a current state of an environment, teaching position compensation amount data indicating the compensation amount of the teaching position in the control of the robot according to the teaching position, motor disturbance value data indicating a disturbance value of each of motors of the robot in the control of the robot, and teaching position data including the teaching position of the teaching data;
a determination data acquisition section that acquires determination data indicating an appropriateness determination result of the disturbance value of each of the motors of the robot in the control of the robot; and
a learning section that learns the compensation amount of the teaching position of the robot in association with the motor disturbance value data and the teaching position by using the state variables and the determination data.