Decentralized Non-communicating Multiagent Collision Avoidance with Deep Reinforcement Learning
Finding feasible, collision-free paths for multiagent systems can be challenging, particularly in non-communicating scenarios where each agent's intent (e.g. goal) is unobservable to the others. In particular, finding time efficient paths often requires anticipating interaction with neighboring...
Main Authors: | , , , |
---|---|
Format: | Report |
Language: | unknown |
Published: |
arXiv
2016
|
Subjects: | |
Online Access: | https://dx.doi.org/10.48550/arxiv.1609.07845 https://arxiv.org/abs/1609.07845 |
id |
ftdatacite:10.48550/arxiv.1609.07845 |
---|---|
record_format |
openpolar |
spelling |
ftdatacite:10.48550/arxiv.1609.07845 2023-05-15T17:53:48+02:00 Decentralized Non-communicating Multiagent Collision Avoidance with Deep Reinforcement Learning Chen, Yu Fan Liu, Miao Everett, Michael How, Jonathan P. 2016 https://dx.doi.org/10.48550/arxiv.1609.07845 https://arxiv.org/abs/1609.07845 unknown arXiv arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ Multiagent Systems cs.MA FOS Computer and information sciences Preprint Article article CreativeWork 2016 ftdatacite https://doi.org/10.48550/arxiv.1609.07845 2022-04-01T11:14:04Z Finding feasible, collision-free paths for multiagent systems can be challenging, particularly in non-communicating scenarios where each agent's intent (e.g. goal) is unobservable to the others. In particular, finding time efficient paths often requires anticipating interaction with neighboring agents, the process of which can be computationally prohibitive. This work presents a decentralized multiagent collision avoidance algorithm based on a novel application of deep reinforcement learning, which effectively offloads the online computation (for predicting interaction patterns) to an offline learning procedure. Specifically, the proposed approach develops a value network that encodes the estimated time to the goal given an agent's joint configuration (positions and velocities) with its neighbors. Use of the value network not only admits efficient (i.e., real-time implementable) queries for finding a collision-free velocity vector, but also considers the uncertainty in the other agents' motion. Simulation results show more than 26 percent improvement in paths quality (i.e., time to reach the goal) when compared with optimal reciprocal collision avoidance (ORCA), a state-of-the-art collision avoidance strategy. : 8 pages, 10 figures Report Orca DataCite Metadata Store (German National Library of Science and Technology) |
institution |
Open Polar |
collection |
DataCite Metadata Store (German National Library of Science and Technology) |
op_collection_id |
ftdatacite |
language |
unknown |
topic |
Multiagent Systems cs.MA FOS Computer and information sciences |
spellingShingle |
Multiagent Systems cs.MA FOS Computer and information sciences Chen, Yu Fan Liu, Miao Everett, Michael How, Jonathan P. Decentralized Non-communicating Multiagent Collision Avoidance with Deep Reinforcement Learning |
topic_facet |
Multiagent Systems cs.MA FOS Computer and information sciences |
description |
Finding feasible, collision-free paths for multiagent systems can be challenging, particularly in non-communicating scenarios where each agent's intent (e.g. goal) is unobservable to the others. In particular, finding time efficient paths often requires anticipating interaction with neighboring agents, the process of which can be computationally prohibitive. This work presents a decentralized multiagent collision avoidance algorithm based on a novel application of deep reinforcement learning, which effectively offloads the online computation (for predicting interaction patterns) to an offline learning procedure. Specifically, the proposed approach develops a value network that encodes the estimated time to the goal given an agent's joint configuration (positions and velocities) with its neighbors. Use of the value network not only admits efficient (i.e., real-time implementable) queries for finding a collision-free velocity vector, but also considers the uncertainty in the other agents' motion. Simulation results show more than 26 percent improvement in paths quality (i.e., time to reach the goal) when compared with optimal reciprocal collision avoidance (ORCA), a state-of-the-art collision avoidance strategy. : 8 pages, 10 figures |
format |
Report |
author |
Chen, Yu Fan Liu, Miao Everett, Michael How, Jonathan P. |
author_facet |
Chen, Yu Fan Liu, Miao Everett, Michael How, Jonathan P. |
author_sort |
Chen, Yu Fan |
title |
Decentralized Non-communicating Multiagent Collision Avoidance with Deep Reinforcement Learning |
title_short |
Decentralized Non-communicating Multiagent Collision Avoidance with Deep Reinforcement Learning |
title_full |
Decentralized Non-communicating Multiagent Collision Avoidance with Deep Reinforcement Learning |
title_fullStr |
Decentralized Non-communicating Multiagent Collision Avoidance with Deep Reinforcement Learning |
title_full_unstemmed |
Decentralized Non-communicating Multiagent Collision Avoidance with Deep Reinforcement Learning |
title_sort |
decentralized non-communicating multiagent collision avoidance with deep reinforcement learning |
publisher |
arXiv |
publishDate |
2016 |
url |
https://dx.doi.org/10.48550/arxiv.1609.07845 https://arxiv.org/abs/1609.07845 |
genre |
Orca |
genre_facet |
Orca |
op_rights |
arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
op_doi |
https://doi.org/10.48550/arxiv.1609.07845 |
_version_ |
1766161504150749184 |