引用本文:吴蔚楠,崔乃刚,郭继峰.基于目标信息估计的分布式局部协调任务分配方法[J].控制理论与应用,2018,35(4):566~576.[点击复制]
WU Wei-nan,CUI Nai-gang,GUO Ji-feng.Distributed task assignment method based on local information consensus and target estimation[J].Control Theory and Technology,2018,35(4):566~576.[点击复制]
基于目标信息估计的分布式局部协调任务分配方法
Distributed task assignment method based on local information consensus and target estimation
摘要点击 2458  全文点击 1579  投稿时间：2017-03-19  修订日期：2017-12-04
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2017.70172
  2018,35(4):566-576
中文关键词  分布式决策  贝叶斯定理  一致性协调  局部链式通信  在线协同
英文关键词  distributed decision-making  Bayes law  consensus algorithm  sparse connected network  cooperation online
基金项目  国家自然科学基金
作者单位E-mail
吴蔚楠* 哈尔滨工业大学 wuweinan123@126.com 
崔乃刚 哈尔滨工业大学  
郭继峰 哈尔滨工业大学  
中文摘要
      分布式决策是提高群体自主性的关键技术之一. 以侦查类无人机(unmanned search aerial vehicles, USAV)和 打击类无人机(unmanned combat aerial vehicles, UCAV)执行协同搜索、攻击灰色目标区域问题为背景, 建立了一种 考虑局部链式通信、无人机飞行性能和任务执行能力等多约束的分布式任务分配模型, 基于贝叶斯定理将任务空 间的连续/离散不确定量用任务收益值量化描述. 然后, 提出了一种基于一致性协调算法的在线协同策略, 并利用一 致协调理论建立了一种冲突调解规则, 在此基础上, 设计了一种分布式任务分配求解算法, 能够实现多USAV, UCAV的协同多任务快速分配. 最后, 通过数值仿真, 验证了本文算法求解不确定空间任务分配问题的可行性和快 速性.
英文摘要
      Distributed decision-making is one of the key components to improve the autonomy of robotic agents. Based on the problem of robust task assignment for a fleet of unmanned combat aerial vehicles (UCAV) and unmanned search aerial vehicles (USAV) to targets under environmental uncertainty, a distributed task assignment model which considering multi-constraints such as sparse connected network, UAV performance and task execution ability is given firstly. Bayes law is adopted to handle both continuous uncertainties and discrete uncertainties in mission space, then an on-line cooperative approach based on consensus algorithm is proposed for the given problem, and the decision rules for conflict resolution is given afterwards, and then a distributed task assignment algorithm which can simultaneously allocates UCAVs and USAVs fastly, is proposed. Finally, aspects of feasibility and rapidity of the proposed algorithm is verified according to the simulation results.