基于Agent的多无人机任务分配模型  

Task Allocation Model for Multi UAVs Based on Agents

作　　者：罗贺[1,2,3] 王国强[1,2,3] 胡笑旋[1,2,3] 夏维[1,2,3] 

机构地区：[1]合肥工业大学飞行器网络系统研究所,合肥230009 [2]合肥工业大学管理学院,合肥230009 [3]合肥工业大学过程优化与智能决策教育部重点实验室,合肥230009

出　　处：《火力与指挥控制》2014年第7期22-26,共5页Fire Control & Command Control

基　　金：国防基础科研计划基金(B1420110145);中航工业产学研专项基金(CXY2011HFGD20);中央高校基本科研业务费专项基金(2012HGZY00009);国家自然科学重点基金资助项目(71131002)

摘　　要：分析了多无人机任务分配经典模型中存在的局限性,建立了基于Agent协商的多无人机任务分配模型。构建了无人机编队中资源管理和任务管理等两种Agent角色,并给出了相应的角色模型及其形式化模型描述,分别设计了相应的资源管理策略和任务管理策略。建立了两种角色之间任务分配协商协议,该协议既有效地保证了个体Agent的自主性,又最大限度地发挥了多Agent之间的合作性。实验表明,该模型能够有效地实现对分布式任务的合理分配。The limitation of the classical model of UAV task allocation is analyzed, and a new task allocation model for multi UAVs based on agent is proposed in this paper. The two different kinds of agent roles has been discussed and defined, as well as the formal model and role model of the agent. The related policies for resource management and task management are also proposed for the agents. On the basis of this,the negotiation protocol between these two roles is also suggested so as to guarantee the autonomy of a single agent and the collaboration among different agents. The experiment results this method can improve the effective of distributed task allocation.

关 键 词：无人机 协商 任务分配 多AGENT系统 角色模型 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]