Construction of a Parallel Training Environment Model for Multi-Agent Deep Reinforcement Learning in Far-Sea Aerial Confrontation

The quality of the environment model determines whether the deep reinforcement learning system can efficiently and accurately learn and train to make good decisions. Aiming at the problems of idealized air combat environment construction and task scenarios in the context of far-sea and remote combat...

Full description

Saved in:
Bibliographic Details
Main Author: Zhang Yuan, Wang Jiangnan, Wang Wei, Li Xuan
Format: Article
Language:zho
Published: Editorial Office of Aero Weaponry 2025-06-01
Series:Hangkong bingqi
Subjects:
Online Access:https://www.aeroweaponry.avic.com/fileup/1673-5048/PDF/2025-0020.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850220224795639808
author Zhang Yuan, Wang Jiangnan, Wang Wei, Li Xuan
author_facet Zhang Yuan, Wang Jiangnan, Wang Wei, Li Xuan
author_sort Zhang Yuan, Wang Jiangnan, Wang Wei, Li Xuan
collection DOAJ
description The quality of the environment model determines whether the deep reinforcement learning system can efficiently and accurately learn and train to make good decisions. Aiming at the problems of idealized air combat environment construction and task scenarios in the context of far-sea and remote combat, this paper constructs a parallel training environment for multi-agent deep reinforcement learning in far-sea air combat. Among them, based on JSBSim and scalable radar and weapon system models, an agent model is built that takes into account both actual combat and simulation performance. This study selects 18-dimensional state space and 7-dimensional action space, and constructs a multi-reward system with the main line and 10 sub-objectives. This approach solves the problems of algorithm difficulty in convergence caused by poor guidance of sparse rewards and high dimensional space. The compliance of the environment, the effectiveness of classic deep reinforcement learning algorithms and compatibility with mainstream training frameworks are verified through simulation.
format Article
id doaj-art-5be2c9830e174902ad1e72ad10e767ee
institution OA Journals
issn 1673-5048
language zho
publishDate 2025-06-01
publisher Editorial Office of Aero Weaponry
record_format Article
series Hangkong bingqi
spelling doaj-art-5be2c9830e174902ad1e72ad10e767ee2025-08-20T02:07:09ZzhoEditorial Office of Aero WeaponryHangkong bingqi1673-50482025-06-01323485610.12132/ISSN.1673-5048.2025.0020Construction of a Parallel Training Environment Model for Multi-Agent Deep Reinforcement Learning in Far-Sea Aerial ConfrontationZhang Yuan, Wang Jiangnan, Wang Wei, Li Xuan01. Aviation Basic College, Naval Aviation University, Yantai 264001, China;2. Unit 91475 of PLA, Huludao 125001, ChinaThe quality of the environment model determines whether the deep reinforcement learning system can efficiently and accurately learn and train to make good decisions. Aiming at the problems of idealized air combat environment construction and task scenarios in the context of far-sea and remote combat, this paper constructs a parallel training environment for multi-agent deep reinforcement learning in far-sea air combat. Among them, based on JSBSim and scalable radar and weapon system models, an agent model is built that takes into account both actual combat and simulation performance. This study selects 18-dimensional state space and 7-dimensional action space, and constructs a multi-reward system with the main line and 10 sub-objectives. This approach solves the problems of algorithm difficulty in convergence caused by poor guidance of sparse rewards and high dimensional space. The compliance of the environment, the effectiveness of classic deep reinforcement learning algorithms and compatibility with mainstream training frameworks are verified through simulation.https://www.aeroweaponry.avic.com/fileup/1673-5048/PDF/2025-0020.pdf|far-sea region|aerial confrontation|multi-agent|deep reinforcement learning|jsbsim|training environment model
spellingShingle Zhang Yuan, Wang Jiangnan, Wang Wei, Li Xuan
Construction of a Parallel Training Environment Model for Multi-Agent Deep Reinforcement Learning in Far-Sea Aerial Confrontation
Hangkong bingqi
|far-sea region|aerial confrontation|multi-agent|deep reinforcement learning|jsbsim|training environment model
title Construction of a Parallel Training Environment Model for Multi-Agent Deep Reinforcement Learning in Far-Sea Aerial Confrontation
title_full Construction of a Parallel Training Environment Model for Multi-Agent Deep Reinforcement Learning in Far-Sea Aerial Confrontation
title_fullStr Construction of a Parallel Training Environment Model for Multi-Agent Deep Reinforcement Learning in Far-Sea Aerial Confrontation
title_full_unstemmed Construction of a Parallel Training Environment Model for Multi-Agent Deep Reinforcement Learning in Far-Sea Aerial Confrontation
title_short Construction of a Parallel Training Environment Model for Multi-Agent Deep Reinforcement Learning in Far-Sea Aerial Confrontation
title_sort construction of a parallel training environment model for multi agent deep reinforcement learning in far sea aerial confrontation
topic |far-sea region|aerial confrontation|multi-agent|deep reinforcement learning|jsbsim|training environment model
url https://www.aeroweaponry.avic.com/fileup/1673-5048/PDF/2025-0020.pdf
work_keys_str_mv AT zhangyuanwangjiangnanwangweilixuan constructionofaparalleltrainingenvironmentmodelformultiagentdeepreinforcementlearninginfarseaaerialconfrontation