Robot path generating device and robot system
AssigneeKabushiki Kaisha Yaska Wa Denki
InventorsKoji Sokabe, Masaru Adachi
Life of this patent
Estimating the term…
Abstract
To generate a more appropriate path, provided is a robot path generation device including circuitry configured to: hold a track planning module learning data set, in which a plurality of pieces of path data generated based on a motion constraint condition of a robot, and evaluation value data, which corresponds to each of the plurality of pieces path data and is a measure under a predetermined evaluation criterion, are associated with each other; and generate, based on a result of a machine learning process that is based on the track planning module learning data set, a path of the robot between a set start point and a set end point, which are freely set.
Claims11
A robot system, comprising:
- a robot;
- a robot path generation device, comprising, circuitry configured to:
- hold a data set, in which a plurality of pieces of path data, generated based on a motion constraint condition of a robot, and evaluation value data, which corresponds to each of the plurality of pieces of path data and is a measure under a predetermined evaluation criterion, are associated with each other;
- and
- generate, based on a result of a machine learning process that is based on the data set, a path of the robot between a set start point and a set end point, which are freely set, the path including a plurality of via points, and
- a robot controller configured to control a motion of the robot based on a result of generation obtained by the robot path generation device,
- wherein the predetermined evaluation criterion includes at least one of an electric power consumption, a period of time of motion, a vibration evaluation value, or a load on a specified axis,
- wherein the circuitry is further configured to:
- generate, based on content learned in the machine learning process, the plurality of via points through which the path is to next pass, such that a value of the predetermined evaluation criterion is optimized;
- and
- generate the path by starting branching and searching at the plurality of via points from the set start point, and repeating the branching and searching until the path reaches a vicinity of the set end point, wherein a distance between the vicinity and the set end point is less than a predetermined distance;
- wherein the plurality of via points are selected based on content learned in the machine learning process.
Dependencies are inferred from the claim text.