网络约束下集群运动体组网姿态确定与协同估计技术研究

Cooperative attitude estimation based on wireless network architecture is an important approach for the cluster mobile agents to break through the bottleneck between efficiency and cost, thus is a promising direction for the deepening and expansion of networked navigation theory. The innovation of this research is to apply networked navigation theory into the attitude determination problem of the cluster mobile agents, and the study starts from the configuration of the baseline of the mobile agents, on which we research the key issues in networked attitude determination theory item-by-item. Finally, the research can form a basic theory architecture for networked attitude determination, which is composed of baseline configuration and source selection strategies, networked attitude derivation and cooperative estimation algorithms, QoS and optimization strategies. The main contents of the study include: (1) As to the attitude measurement requirement of mobile agents, a set of baseline configuration and source selection strategies will be researched based on networked navigation theory. (2) Based on the distinguishing features of network constraints and induced errors, a set of networked attitude determination schemes will be presented, and the related cooperative state estimation algorithms will be deduced with full consideration of network constraints and network-induced errors. (3) QoS scheme will be established based on information theories, and optimization strategy for networked attitude estimation algorithms will be designed based on control theories. Finally, according to a specific application, the performance of networked attitude estimation algorithms will be verified by based on the UWB-NNS principle prototype. The results of this research can provide some theoretical supports for military fields of weapon platforms’ cooperative attitude determination in the scenarios of GNSS-challenge or system failure, and for civil fields in terms of networked attitude estimation and cooperative formation control of clustered mobile agents.

网络信源环境下的协同姿态估计是集群运动体突破“效能与成本”瓶颈的一个重要途径，也是组网导航理论深化与拓展的一个新兴方向。本项目拟将组网导航理论引入集群运动体的测姿问题，从配置组网运动体的测姿基线入手，对组网测姿理论及其关键科学问题开展研究，形成涵盖基线配置与信源优选策略、组网姿态确定与协同估计方案、QoS组合优化与联调机制的技术基础。具体研究内容为：⑴针对集群运动体的姿态测量需求，提出组网导航机制下的测姿基线配置与信源筛选方案；⑵分析网络约束和诱导误差的影响，设计组网测姿方案与协同估计算法；⑶针对信源可靠性与容错性需求，研究QoS优化与联调联试方案；最后，将结合具体应用背景，依托组网导航开发系统，对组网测姿的相关方案与算法进行试验验证。项目研究成果能够为卫星拒止甚至惯导失效时的集群作战平台的协同姿态测量等军事用途，以及集群无人平台的组网测姿与队形控制等民用领域的实用化提供理论与技术支撑。

本项目针对集群运动平台对新体制测姿技术的特殊要求，开展了基于导航传感器网络（NC-NET）的多运动体组网测姿技术研究，项目从分析运动体的集群应用对姿态测量和相对姿态确定的需求着手，创新性地提出多运动体组网测姿理论，构建了适用于集群运动体协同的组网测姿理论体系框架，并对其关键技术进行了系统地研究。. 项目在执行过程中首先面向集群测姿需求对NC-NET的通信协议进行了优化和改进，进而完成了组网测姿所需的基线配置策略和信源筛选方案；其次，基于NC-NET辅助信息的获取特点，提出了一系列组网测姿新方案和新模式；此后，针对多种导航场景下的网络约束，推导并验证了组网测姿系统的多种融合估计算法；此外，课题中还初步探索了集群智能体的组网导航基控制反馈问题；然后，针对组网测姿系统的QoS机制和容错控制问题，提出了针对线性/非线性组网测姿系统的多种故障检测、容错控制和最优估计方案；最后，基于上述研究成果研发完成了组网测姿技术的原理样机，并进行了系统性地试验验证。. 项目在研究过程中，在国内外著名期刊及会议发表论文34篇，其中SCI检索论文1篇，EI检索论文24篇，中文核心期刊论文27篇；待出版英文学术专著1部；培养了博士后1名，博士研究生5名，硕士研究生2名。圆满完成了项目的各项研究计划。. 项目按计划完成了预期研究内容，达到了预期研究目标，获得了预期研究成果，其成果可为我国军用/民用运动平台的集群运动与协同、组网测姿与协同控制等任务提供理论与技术支撑。