Machine learning device, machine learning method, and controller for robot
A machine learning device and a machine learning method are provided. A controller, which serves as the machine learning device and the machine learning method, is provided to perform machine learning of algorithms for arranging a flexible wire-like workpiece to a predetermined state using an industrial robot. In an embodiment, the machine learning device includes: an acquisition unit that acquires, as state variables, a state of the workpiece before arrangement starts and a state of the workpiece during arrangement; and a learning unit that performs machine learning of an algorithm for arranging the workpiece based on the state variables acquired by the acquisition unit.
1. A machine learning device that performs machine learning of an algorithm for arranging a flexible wire-like workpiece to a predetermined state using a robot, the machine learning device comprising:
an acquisition unit configured to acquire state variables showing a state of the workpiece before a start of arrangement of the workpiece which is to be performed by the robot, and acquire repeatedly the state variables during the arrangement of the workpiece which is performed by the robot to the predetermined state at which the workpiece is finally arranged, the workpiece having a connector, the state variables for each of the states including a position of the workpiece, a force applied from the workpiece when the connector of the workpiece is gripped, and a posture of the robot;
a storage unit configured to pre-store a plurality of training data for arranging the workpiece to the predetermined state; and
a learning unit configured to perform machine learning of the algorithm for arranging the workpiece based on the state variables acquired by the acquisition unit, by classifying values of the state variables acquired by the acquisition unit according to the training data pre-stored in the storage unit.
2. A machine learning method for performing machine learning of an algorithm for arranging a flexible wire-like workpiece to a predetermined state using a robot, the machine learning method comprising the steps of:
acquiring state variables showing a state of the workpiece before a start of arrangement of the workpiece which is to be performed by the robot and acquiring repeatedly the state variables during the arrangement of the workpiece which is performed by the robot to the predetermined state at which the workpiece is finally arranged, the workpiece having a connector, the state variables for each of the states including a position of the workpiece, a force applied from the workpiece when the connector of the workpiece is gripped, and a posture of the robot; and
performing machine learning of the algorithm for arranging the workpiece based on the state variables acquired by the acquiring step, by classifying values of the acquired state variables according to training data pre-stored in a storage unit.
3. A controller for a robot, the controller comprising:
a control unit configured to control an operation of the robot based on an algorithm learned by a machine learning device, the machine learning device including:
an acquisition unit configured to acquire state variables showing a state of a flexible wire-like workpiece provided before a start of arrangement of the workpiece which is to be performed by the robot, and acquire the state variables during the arrangement of the workpiece which is performed by the robot to a predetermined state at which the workpiece is finally arranged, the workpiece having a connector, the state variables for each of the states including a position of the workpiece, a force applied from the connector of the workpiece when the workpiece is gripped, and a posture of the robot;
a storage unit configured to pre-store a plurality of training data for arranging the workpiece to the predetermined state; and
a learning unit configured to perform machine learning of the algorithm for arranging the workpiece based on the state variables acquired by the acquisition unit, by classifying values of the state variables acquired by the acquisition unit according to the training data pre-stored in the storage unit.
4. The machine learning device according to claim 1 , wherein the flexible wire-like workpiece is a wire harness, a power cable, an electric wire, a vinyl-made cord, or a rubber hose.
5. The machine learning method according to claim 2 , wherein the flexible wire-like workpiece is a wire harness, a power cable, an electric wire, a vinyl-made cord, or a rubber hose.