Multi-Target Encirclement with Collision Avoidance via Deep Reinforcement Learning using Relational Graphs
文献类型:会议论文
作者 | Zhang TL(张天乐)1,2; Liu Z(刘振)1,2; Pu ZQ(蒲志强)1,2; Yi JQ(易建强)1,2 |
出版日期 | 2022 |
会议日期 | May 23-27, 2022 |
会议地点 | Philadelphia, PA, USA |
英文摘要 | In this paper, we propose a novel decentralized method based on deep reinforcement learning using robot-level and target-level relational graphs, to solve the problem of multi-target encirclement with collision avoidance (MECA). Specifically, the robot-level relational graphs, composed of three heterogeneous relational graphs between each robot and other robots, targets and obstacles, are modeled and learned through using graph attention networks (GATs) for extracting different spatial relational representations. Moreover, for each target within the observation of each robot, a target-level relational graph is built with GAT to construct spatial relations from the robot. Furthermore, the movement of each target is modeled by the target-level relational graph and learned through supervised learning for predicting the trajectory of the target. In addition, a knowledge-embedded compound reward function is defined to solve the multi-objective problem in MECA, and guide the policy learning for deriving the behavior of MECA. An actor-critic training algorithm based on the centralized training and decentralized execution framework is adopted to train the policy network. Simulation and real-world experiment results demonstrate the effectiveness and generalization of our method. |
会议录出版者 | IEEE |
语种 | 英语 |
源URL | [http://ir.ia.ac.cn/handle/173211/51959] |
专题 | 综合信息系统研究中心_飞行器智能技术 |
通讯作者 | Liu Z(刘振) |
作者单位 | 1.中国科学院自动化研究所 2.中国科学院大学人工智能学院 |
推荐引用方式 GB/T 7714 | Zhang TL,Liu Z,Pu ZQ,et al. Multi-Target Encirclement with Collision Avoidance via Deep Reinforcement Learning using Relational Graphs[C]. 见:. Philadelphia, PA, USA. May 23-27, 2022. |
入库方式: OAI收割
来源:自动化研究所
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。