-
公开(公告)号：US09597797B2
公开(公告)日：2017-03-21
申请号：US14102410
申请日：2013-12-10
Applicant: BRAIN CORPORATION
Inventor： Filip Ponulak , Moslem Kazemi , Patryk Laurent , Oleg Sinyavskiy , Eugene Izhikevich
CPC classification number: B25J9/163 , B25J9/161 , G05B2219/36418 , G05B2219/36425 , G05B2219/40499 , G05D1/005 , G05D1/0088 , G05D1/0221 , G05D2201/02 , G06N3/008 , G06N3/049 , G06N99/005
Abstract: Robotic devices may be trained by a trainer guiding the robot along a target trajectory using physical contact with the robot. The robot may comprise an adaptive controller configured to generate control commands based on one or more of the trainer input, sensory input, and/or performance measure. The trainer may observe task execution by the robot. Responsive to observing a discrepancy between the target behavior and the actual behavior, the trainer may provide a teaching input via a haptic action. The robot may execute the action based on a combination of the internal control signal produced by a learning process of the robot and the training input. The robot may infer the teaching input based on a comparison of a predicted state and actual state of the robot. The robot's learning process may be adjusted in accordance with the teaching input so as to reduce the discrepancy during a subsequent trial.
-
Patent Agency Ranking
公开(公告)号：US10717191B2
公开(公告)日：2020-07-21
申请号：US15845832
申请日：2017-12-18
Applicant: Brain Corporation
Inventor： Filip Ponulak , Moslem Kazemi , Patryk Laurent , Oleg Sinyavskiy , Eugene Izhikevich
Abstract: Robotic devices may be trained by a trainer guiding the robot along a target trajectory using physical contact with the robot. The robot may comprise an adaptive controller configured to generate control commands based on one or more of the trainer input, sensory input, and/or performance measure. The trainer may observe task execution by the robot. Responsive to observing a discrepancy between the target behavior and the actual behavior, the trainer may provide a teaching input via a haptic action. The robot may execute the action based on a combination of the internal control signal produced by a learning process of the robot and the training input. The robot may infer the teaching input based on a comparison of a predicted state and actual state of the robot. The robot's learning process may be adjusted in accordance with the teaching input so as to reduce the discrepancy during a subsequent trial.
-
公开(公告)号：US20180272529A1
公开(公告)日：2018-09-27
申请号：US15845832
申请日：2017-12-18
Applicant: Brain Corporation
Inventor： Filip Ponulak , Moslem Kazemi , Patryk Laurent , Oleg Sinyavskiy , Eugene Izhikevich
Abstract: Robotic devices may be trained by a trainer guiding the robot along a target trajectory using physical contact with the robot. The robot may comprise an adaptive controller configured to generate control commands based on one or more of the trainer input, sensory input, and/or performance measure. The trainer may observe task execution by the robot. Responsive to observing a discrepancy between the target behavior and the actual behavior, the trainer may provide a teaching input via a haptic action. The robot may execute the action based on a combination of the internal control signal produced by a learning process of the robot and the training input. The robot may infer the teaching input based on a comparison of a predicted state and actual state of the robot. The robot's learning process may be adjusted in accordance with the teaching input so as to reduce the discrepancy during a subsequent trial.
-
公开(公告)号：US20170232613A1
公开(公告)日：2017-08-17
申请号：US15464104
申请日：2017-03-20
Applicant: Brain Corporation
Inventor： Filip Ponulak , Moslem Kazemi , Patryk Laurent , Oleg Sinyavskiy , Eugene zhikevich
CPC classification number: B25J9/163 , B25J9/161 , G05B2219/36418 , G05B2219/36425 , G05B2219/40499 , G05D1/005 , G05D1/0088 , G05D1/0221 , G05D2201/02 , G06N3/008 , G06N3/049 , G06N99/005
Abstract: Robotic devices may be trained by a trainer guiding the robot along a target trajectory using physical contact with the robot. The robot may comprise an adaptive controller configured to generate control commands based on one or more of the trainer input, sensory input, and/or performance measure. The trainer may observe task execution by the robot. Responsive to observing a discrepancy between the target behavior and the actual behavior, the trainer may provide a teaching input via a haptic action. The robot may execute the action based on a combination of the internal control signal produced by a learning process of the robot and the training input. The robot may infer the teaching input based on a comparison of a predicted state and actual state of the robot. The robot's learning process may be adjusted in accordance with the teaching input so as to reduce the discrepancy during a subsequent trial.
-
公开(公告)号：US09844873B2
公开(公告)日：2017-12-19
申请号：US15464104
申请日：2017-03-20
Applicant: Brain Corporation
Inventor： Filip Ponulak , Moslem Kazemi , Patryk Laurent , Oleg Sinyavskiy , Eugene Izhikevich
CPC classification number: B25J9/163 , B25J9/161 , G05B2219/36418 , G05B2219/36425 , G05B2219/40499 , G05D1/005 , G05D1/0088 , G05D1/0221 , G05D2201/02 , G06N3/008 , G06N3/049 , G06N99/005
Abstract: Robotic devices may be trained by a trainer guiding the robot along a target trajectory using physical contact with the robot. The robot may comprise an adaptive controller configured to generate control commands based on one or more of the trainer input, sensory input, and/or performance measure. The trainer may observe task execution by the robot. Responsive to observing a discrepancy between the target behavior and the actual behavior, the trainer may provide a teaching input via a haptic action. The robot may execute the action based on a combination of the internal control signal produced by a learning process of the robot and the training input. The robot may infer the teaching input based on a comparison of a predicted state and actual state of the robot. The robot's learning process may be adjusted in accordance with the teaching input so as to reduce the discrepancy during a subsequent trial.
-
公开(公告)号：US20150127150A1
公开(公告)日：2015-05-07
申请号：US14102410
申请日：2013-12-10
Applicant: BRAIN CORPORATION
Inventor： Filip Ponulak , Moslem Kazemi , Patryk Laurent , Oleg Sinyavskiy , Eugene Izhikevich
IPC: B25J9/16
CPC classification number: B25J9/163 , B25J9/161 , G05B2219/36418 , G05B2219/36425 , G05B2219/40499 , G05D1/005 , G05D1/0088 , G05D1/0221 , G05D2201/02 , G06N3/008 , G06N3/049 , G06N99/005
Abstract: Robotic devices may be trained by a trainer guiding the robot along a target trajectory using physical contact with the robot. The robot may comprise an adaptive controller configured to generate control commands based on one or more of the trainer input, sensory input, and/or performance measure. The trainer may observe task execution by the robot. Responsive to observing a discrepancy between the target behavior and the actual behavior, the trainer may provide a teaching input via a haptic action. The robot may execute the action based on a combination of the internal control signal produced by a learning process of the robot and the training input. The robot may infer the teaching input based on a comparison of a predicted state and actual state of the robot. The robot's learning process may be adjusted in accordance with the teaching input so as to reduce the discrepancy during a subsequent trial.
Abstract translation: 机器人设备可以由训练者训练，该训练者使用与机器人的物理接触沿着目标轨迹引导机器人。 机器人可以包括自适应控制器，其被配置为基于训练者输入，感觉输入和/或性能测量中的一个或多个来产生控制命令。 训练者可以观察机器人执行任务。 响应于观察目标行为与实际行为之间的差异，培训者可以通过触觉动作提供教学输入。 机器人可以基于由机器人的学习过程产生的内部控制信号与训练输入的组合来执行动作。 机器人可以基于预测状态与机器人的实际状态的比较来推断教学输入。 机器人的学习过程可以根据教学输入进行调整，以减少后续试验期间的差异。
-
-
-
-
-
Search Results
6
records
(took 0.014 seconds)
