复杂环境下空地无人系统智能协同定位与导航方法研究

Autonomous unmanned systems (AUS) possess outstanding characteristics of flexibility, low-cost, strong adaptability and so on. Nowadays, AUS has been a growing and active research area with many application prospects in both military and civilian fields. Generally, AUS has been regarded as a new engine to promote military reform, economic development and social progress in the world. To meet the development needs of national civil-military integration strategy and military intelligence, this project intends to conduct in-depth research on theory and methods for intelligent cooperative positioning and navigation of unmanned air/ground systems in complex environments, which will promote the development of AUS intelligence, cooperation and swarm. First, we put forward a multi-sensor intelligent fusion positioning method with variable structure for satisfying intelligent combination of navigation sensors according to their sensing ability, and guaranteeing the continuity and robustness of navigation. Second, in order to make full use of the advantages of AUS multi-domain cooperation, we propose a new method of air/ground relative navigation, distributed fast transfer alignment and cooperative positioning for compensating low-cost sensor errors, and enhancing the ability of fault detection, diagnosis and recovery. Finally, inspired by human perception, memory, decision-making and behavioral mechanism, we focus on the technology breakthrough of intelligent navigation in unmanned air/ground cooperative systems, and make great efforts to form a relatively complete theoretical system of air/ground intelligent cooperative navigation. The research of this project will evidently improve the ability of complex task organization and precise cooperative coordination of air/ground autonomous unmanned systems.

自主无人系统具有机动灵活、低成本和适应性强等特点，是推动军事变革、经济发展和社会进步的新引擎，在军事和民用领域都具有广阔的应用前景。面向国家军民融合发展战略和军事智能化发展的重大需求，本项目开展复杂环境下空地无人系统智能协同定位与导航方法研究，促进自主无人系统智能化、协同化和集群化发展。研究内容包括：研究多传感器智能融合定位方法，实现特定场景下根据传感器感知能力进行优选组合，确保融合定位的连续性和稳健性；充分发挥自主无人系统多域协同的优势，研究分布式快速传递对准与协同定位方法，弥补无人机载荷有限、搭载传感器精度低的不足，并对可能出现的故障进行识别、隔离和修复；借鉴人类的认知、记忆、决策和行为机理，设计空地智能协同定位与导航系统，并搭建试验平台加以验证。本项目将形成较为完整的空地无人系统智能协同定位与导航方法，提升智能无人装备复杂任务组织和精确协同配合的能力。

面向自主无人系统智能化、协同化和集群化的发展需求，项目以无人车和无人机组成的空地无人系统为研究对象，聚焦于复杂环境下空地无人系统智能协同定位与导航方法研究，取得了以下主要研究成果：1）针对复杂环境下多传感器信息评价与智能化融合的问题，提出了异构多信息源可观测度量化分析与评价准则，设计了适用于跨场景的多源导航信息可信性分析方法和多传感器智能化融合定位方法；2）针对空地无人系统跨域信息共享与协同导航定位的问题，建立了无人车和无人机相对导航模型，设计了鲁棒自适应卡尔曼滤波算法，形成了空地无人系统相对导航与协同导航方法；3）针对空地无人系统智能协同定位与导航的仿真与验证需求，设计了空地无人系统智能协同导航架构及系统集成方法，验证了复杂环境下空地无人系统智能协同定位与导航方法的可行性和有效性；4）发表高水平学术论文9篇（其中国际顶级和重要期刊论文4篇、国内期刊论文2篇、会议论文3篇），授权国家发明专利3项，培养硕士研究生3人。