Title:
CONTROL DEVICE, CONTROL METHOD, AND STORAGE MEDIUM
Document Type and Number:
WIPO Patent Application WO/2022/013933
Kind Code:
A1
Abstract:
A control device 200X has, in terms of function, an action policy acquisition means 21X and a policy synthesis means 23X. The action policy acquisition means 21X acquires an action policy pertaining to the action of a robot. The policy synthesis means 23X synthesizes at least two action policies so as to generate a control command for the robot.
Inventors:
ITOU TAKEHIRO (JP)
OYAMA HIROYUKI (JP)
OYAMA HIROYUKI (JP)
Application Number:
PCT/JP2020/027311
Publication Date:
January 20, 2022
Filing Date:
July 14, 2020
Export Citation:
Assignee:
NEC CORP (JP)
International Classes:
B25J13/00
Domestic Patent References:
| WO2020058669A1 | 2020-03-26 |
Foreign References:
| JP2016196079A | 2016-11-24 |
Other References:
UCHIBE EIJI: "Forward and Inverse Reinforcement Learning Based on Linearly Solvable Markov Decision Processes", JAPANESE NEURAL NETWORK SOCIETY, 5 March 2016 (2016-03-05), pages 2 - 13, XP055898081, Retrieved from the Internet DOI: 10.3902/jnns.23.2
Attorney, Agent or Firm:
NAKAMURA, Toshinobu et al. (JP)
Download PDF:
Previous Patent: CENTRAL DEVICE, MAP GENERATION SYSTEM, AND MAP GENERATION METHOD
Next Patent: X-RAY FLUORESCENCE ANALYZER
Next Patent: X-RAY FLUORESCENCE ANALYZER