Title:
MOTION PLANNING METHOD AND APPARATUS
Document Type and Number:
WIPO Patent Application WO/2021/238303
Kind Code:
A1
Abstract:
The present application relates to the field of artificial intelligence, and in particular to the field of autonomous driving. Provided are a motion planning method and apparatus. The method comprises: acquiring driving environment information, wherein the driving environment information comprises position information of a dynamic obstacle; inputting a state representation of the driving environment information into a trained reinforcement learning network model, and acquiring a prediction horizon that is output by the reinforcement learning network model, wherein the prediction horizon represents a duration or the number of steps for performing motion trajectory prediction on the dynamic obstacle; and using the prediction horizon to perform motion planning. A prediction horizon is obtained by means of reinforcement learning, so that the prediction horizon can be dynamically changed along with a change in a driving environment, and an autonomous vehicle can flexibly cope with a dynamic obstacle during the process of interaction between the autonomous vehicle and the dynamic obstacle.
Inventors:
WANG ZHITAO (CN)
ZHUANG YUZHENG (CN)
LIU WULONG (CN)
GU QIANG (CN)
ZHUANG YUZHENG (CN)
LIU WULONG (CN)
GU QIANG (CN)
Application Number:
PCT/CN2021/075925
Publication Date:
December 02, 2021
Filing Date:
February 08, 2021
Export Citation:
Assignee:
HUAWEI TECH CO LTD (CN)
International Classes:
G05D1/02
Foreign References:
| CN109855639A | 2019-06-07 | |||
| CN110456634A | 2019-11-15 | |||
| CN110293968A | 2019-10-01 | |||
| JP2002162242A | 2002-06-07 | |||
| US20190283742A1 | 2019-09-19 |
Other References:
CHEN SENGPENG, WU JIA;CHEN XIU-YUN: "Hyperparameter Optimization Method Based on Reinforcement Learning", JOURNAL OF CHINESE COMPUTER SYSTEMS, GAI-KAN BIANJIBU , SHENYANG, CN, vol. 41, no. 4, 30 April 2020 (2020-04-30), CN , pages 679 - 684, XP055872007, ISSN: 1000-1220
Attorney, Agent or Firm:
LONGSUN LEAD IP LTD. (CN)
Download PDF: