Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
AUTOMATIC DRIVING VEHICLE NAVIGATION CONTROL METHOD AND SYSTEM
Document Type and Number:
WIPO Patent Application WO/2024/087654
Kind Code:
A1
Abstract:
The present invention relates to an automatic driving vehicle navigation control method and system. The method comprises performing navigation by means of an optimized navigation model according to vehicle and environment state data. A training process comprises: determining a vehicle control quantity by means of a navigation control algorithm according to point cloud data of an obstacle; constructing a navigation model by means of a DRL decision network according to the vehicle control quantity and the vehicle and environment state data, wherein the DRL decision network comprises: a DRL Actor network for outputting a first vehicle control quantity according to the vehicle and environment state data and the vehicle control quantity, and a DRL Critic network for outputting, according to the first vehicle control quantity, the vehicle control quantity, and the vehicle and environment state data, expected returns corresponding to the two groups of vehicle control quantities, and determining and outputting a final control quantity; and optimizing the navigation model by using a reward and punishment mechanism. The present invention can improve the precision of automatic driving navigation control of a DRL mode, increase the convergence speed in neural network training, and reduce the convergence difficulty in neural network training.

Inventors:
WU YAN (CN)
GAO LONGFEI (CN)
WANG LIFANG (CN)
GOU JINFANG (CN)
Application Number:
PCT/CN2023/100154
Publication Date:
May 02, 2024
Filing Date:
June 14, 2023
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
INST ELECTRICAL ENG CAS (CN)
International Classes:
G05D1/02
Foreign References:
CN115494849A2022-12-20
CN114355897A2022-04-15
CN109991987A2019-07-09
CN113335291A2021-09-03
US20190113919A12019-04-18
Other References:
LINHAI XIE ET AL.: "Learning With Stochastic Guidance for Robot Navigation", IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, vol. 32, no. 1, 31 January 2021 (2021-01-31), XP011829589, ISSN: 2162-237X, DOI: 10.1109/TNNLS.2020.2977924
Attorney, Agent or Firm:
BEIJING GAOWO LAW FIRM (CN)
Download PDF: