Controller and machine learning device
AssigneeFanuc Corporation
InventorKouichirou Hayashi
Life of this patent
Estimating the term…
Abstract
A machine learning device is provided in a versatile controller capable of inferring command data to be issued to each axis of a robot. The device includes an axis angle conversion unit calculating, from the trajectory data, an amount of change of an axis angle of an axis of the robot, a state observation unit observing axis angle data relating to the amount of change of the axis angle of the axis of the robot as a state variable representing a current state of an environment, a label data acquisition unit acquiring axis angle command data relating to command data for the axis of the robot as label data, and a learning unit learning the amount of change of the axis angle of the axis of the robot and the command data for the axis in association with each other by using the state variable and the label data.
Claims9
A controller that infers command data to be issued to an axis of a robot on the basis of trajectory data including positions of an end position of an arm of the robot, the controller, comprising:
- a processor configured to:
- receive the trajectory data from a trajectory measurement instrument that directly measures the positions of the end position of an arm of the robot along the trajectory;
- calculate, from the trajectory data, an actual amount of change of an axis angle of the axis of the robot when the end position of the arm of the robot moves between the positions in the trajectory data;
- determine axis angle data relating to the actual amount of change of the axis angle of the axis of the robot as a state variable representing a current state of an environment;
- acquire the command data indicating a commanded amount of change in the axis angle executed when moving the end position of the arm of the robot between the positions in the trajectory data for the axis of the robot as label data;
- learn an amount of change of the axis angle of the axis of the robot associated with executing the command data based on a correlation between the commanded amount of change to the actual amount of change;
- and
- modify the commanded amount of change in the axis angle in future command data based on the learned amount of change to control the end position of an arm of the robot to achieve a target trajectory to be followed by the end of the robot.
A controller that infers command data to be issued to an axis of a robot on the basis of trajectory data including positions of an end position of an arm of the robot, the controller, comprising:
- a processor configured to:
- receive the trajectory data from a trajectory measurement instrument that directly measures the positions of the end position of an arm of the robot along the trajectory;
- calculate, from the trajectory data, an actual amount of change of an axis angle of the axis of the robot when the end position of the arm of the robot moves between the positions in the trajectory data;
- determine axis angle data relating to the actual amount of change of the axis angle of the axis of the robot as a state variable representing a current state of an environment;
- acquire the command data indicating a commanded amount of change in the axis angle executed when moving the end position of the arm of the robot between the positions in the trajectory data for the axis of the robot as label data;
- learn an amount of change of the axis angle of the axis of the robot associated with executing the command data based on a correlation between the commanded amount of change to the actual amount of change;
- infer and output command data for the axis of the robot on the basis of the state variable observed and a result of the learning;
- and
- modify the commanded amount of change in the axis angle in future command data based on the learned amount of change to control the end position of an arm of the robot to achieve a target trajectory to be followed by the end of the robot.
A machine learning device that learns command data to be issued to an axis of a robot, the command data corresponding to trajectory data including positions of an end position of an arm of the robot, the machine learning device comprising:
- a processor configured to:
- receive the trajectory data from a trajectory measurement instrument that directly measures the positions of the end position of an arm of the robot along the trajectory;
- calculate, from the trajectory data, an actual amount of change of an axis angle of the axis of the robot when the end position of the arm of the robot moves between the positions in the trajectory data;
- determine axis angle data relating to the actual amount of change of the axis angle of the axis of the robot as a state variable representing a current state of an environment;
- acquire the command data indicating a commanded amount of change in the axis angle executed when moving the end position of the arm of the robot between the positions in the trajectory data for the axis of the robot as label data;
- and
- learn an amount of change of the axis angle of the axis of the robot associated with executing the command data based on a correlation between the commanded amount of change to the actual amount of change.
A machine learning device that has learned inference of command data to be issued to an axis of a robot, the command data corresponding to trajectory data including positions of an end position of an arm of the robot, the machine learning device comprising:
- a processor configured to:
- receive the trajectory data from a trajectory measurement instrument that directly measures the positions of the end position of an arm of the robot along the trajectory;
- calculate, from the trajectory data, an actual amount of change of the axis angle of the axis of the robot when the end position of the arm of the robot moves between the positions in the trajectory data;
- determine axis angle data relating to the actual amount of change of the axis angle of the axis of the robot as a state variable representing a current state of an environment;
- acquire the command data indicating a commanded amount of change in the axis angle executed when moving the end position of the arm of the robot between the positions in the trajectory data for the axis of the robot as label data;
- learn an amount of change of the axis angle of the axis of the robot associated with executing command data based on a correlation between the commanded amount of change to the actual amount of change;
- and
- infer and output command data for the axis of the robot on the basis of the state variable observed and a result of the learning.
Dependencies are inferred from the claim text.