有人/无人作战智能体分布式协同目标分配方法  

Distributed cooperative target assignment method of manned combat agents and unmanned combat agents

作　　者：万路军[1,2] 姚佩阳[1] 周翔翔[3] 孙鹏[1] 

机构地区：[1]空军工程大学信息与导航学院 [2]空军工程大学空管领航学院 [3]中国人民解放军95616部队

出　　处：《系统工程与电子技术》2014年第2期278-287,共10页Systems Engineering and Electronics

基　　金：国家自然科学基金(61273048);空军工程大学信息与导航学院博士创新基金(KDY2011-002)资助课题

摘　　要：有人作战智能体(manned combat agent,MCA)控制无人作战智能体(unmanned combat agent,UCA)的分布式目标分配是MCA/UCA协同作战的重要内容之一。结合设想的MCA/UCA混合编队结构和功能,建立了有限控制下的分布式协同目标分配模型。提出通信结构、信息结构和协同机制三维一体的模型求解策略,以合同网协议作为协同机制基本框架,分别对合同网的目标选择策略、招标策略、投标策略、中标策略进行扩展,可有效在多个编组之间进行目标到UCA及其所隶属的MCA的协同目标分配。针对想定案例进行了仿真计算,结果表明该方法能以牺牲少量计算时间为代价,以较少的通信量得到了较高质量的解。The distributed target assignment of unmanned combat agents （UCA） controlled by manned combat agents （MCA） is very important in MCA/UCA cooperative combat. Under the hypothesis of MCA/ UCA hybrid aerial team＇s architecture and function, the distributed cooperative target assignment model with limited control is established. The integrative model strategy of communication structure, information structure and cooperative mechanism is proposed. Utilizing the contract net protocol as the basic frame, the target selec- tion strategy, announcement strategy, bidding strategy and winner determination strategy of the contract net are extended. Through the improved mechanism of the contract net algorithm, the target can be assigned to UCA and MCA effectively in multiple group. Simulations results based on combat scenario indicate that the algorithm can attain high quality results with low traffic and limited computing time.

关 键 词：有人 无人作战智能体 分布式目标分配 通信结构 信息结构 协同机制 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP237.5[自动化与计算机技术—计算机科学与技术]