火力与指挥控制2024，Vol.49Issue(9)：58-63,6.DOI:10.3969/j.issn.1002-0640.2024.09.008
基于EC-MAVEN算法的多智能体多约束作战任务分配方法
Multi-Agent and Multi-Constraint Combat Task Allocation Method Based on EC-MAVEN Algorithm
摘要
Abstract
To solve the problems of slow convergence and low exploration efficiency of existing heu-ristic algorithms in the face of multi-constraint combat task allocation.Based on the combat simulation environment,an improved multi-intelligent agent variational exploration method(EC-MAVEN)is pro-posed to model the combat task assignment problems with multiple constraints,and to improve the explora-tion efficiency of intelligent agents in complex tasks,and the control of scenes is used to improve the utili-zation rate of good samples and to speed up the training speed,and to facilitate the jumping out of the local optimum.The simulation example is verified by the combat strike task allocation environment,and the results show that the EC-MAVEN algorithm can reduce falling into local sub-optimality,and can shorten the training time,and to improve the mission success rate.关键词
多智能体/强化学习/任务分配/协作任务Key words
multi-intelligent agent/reinforcement learning/task allocation/collaborative tasks分类
信息技术与安全科学引用本文复制引用
钟孙健,张德平..基于EC-MAVEN算法的多智能体多约束作战任务分配方法[J].火力与指挥控制,2024,49(9):58-63,6.基金项目
国防基础科研基金(JCKY2020605C003) （JCKY2020605C003）
国防基础科研基金资助项目(JCKY2022605C006) （JCKY2022605C006）
