Title:
NON-MODEL ROBOT CONTROL METHOD FOR MULTI-SHAFT-HOLE ASSEMBLY OPTIMIZED BY ENVIRONMENTAL PREDICTION
Document Type and Number:
WIPO Patent Application WO/2020/207219
Kind Code:
A1
Abstract:
Provided is a non-model robot control method for multi-shaft-hole assembly optimized by environmental prediction. A general action value function is used for environment-related knowledge prediction, and a fuzzy logic system designed according to people's assembly experience uses the learned knowledge prediction as an input, outputs parameters for optimizing a non-model control algorithm, and when the number of iterations of a deep reinforcement learning network reaches the upper limit and the assembly meets requirements, outputs the trained deep reinforcement learning network to output assembly actions during the multi-shaft-hole assembly process. The existing non-model robot control algorithm is optimized, and the time required for a robot to carry out multi-shaft-hole assembly is shortened.
More Like This:
Inventors:
XU JING (CN)
HOU ZHIMIN (CN)
HOU ZHIMIN (CN)
Application Number:
PCT/CN2020/080017
Publication Date:
October 15, 2020
Filing Date:
March 18, 2020
Export Citation:
Assignee:
UNIV TSINGHUA (CN)
International Classes:
B25J9/16; B23P19/00
Domestic Patent References:
WO2017156243A1 | 2017-09-14 |
Foreign References:
CN110238839A | 2019-09-17 | |||
CN108196447A | 2018-06-22 | |||
CN109543823A | 2019-03-29 | |||
CN106112505A | 2016-11-16 | |||
CN108161934A | 2018-06-15 |
Attorney, Agent or Firm:
TSINGYIHUA INTELLECTUAL PROPERTY LLC (CN)
Download PDF: