基于模型优化的GNSS/INS矢量深组合跟踪环路研究

As the development of GNSS/INSS integration systems, deep integration technique has been the focus of related research institutes and scholars. In contrast to the low-level deep integration structures that have been well studied, the investigation of high-complexity vector-based integration systems is still in the process of simulation or even in the startup stage. The complex structures and involved advanced algorithms prohibit the system from being analyzed in depth from the mathematical model perspective and developing further. The project mainly focuses on the typical cascade/non-coherent deep GNSS/INS integration system. Firstly, the pros and cons of the Kalman based pre-filter is investigated. Then, the Kalman based pre-filter is analyzed in depth exploiting its control model in S zone, leading to the mathematic descriptions of the key parameters’ properties and effects. Furthermore, a new advanced pre-filter is proposed by optimizing the previous control model. Up to the system level, the S-zone models of key components are constructed collectively, forming the whole tracking loop model in S zone, which is exploited to analyze the feedback delay effect and develop the optimized frequency discriminator . Finally, the related model analysis theory and the proposed optimized vector tracking loop in the deep integration system are tested and verified both on software and hardware platforms. The work in this project provides the theory basis for the further in-depth study and development of the vector-based GNSS/INS deep integration systems.

随着GNSS/INS组合导航的日益发展，深组合导航技术在军民领域已成为国内外研究和发展的焦点。相比于趋于成熟的低复杂度标量深组合系统技术，高复杂度矢量深组合系统的研究基本处于理论仿真实现、甚至起步阶段。结构与算法的高复杂度成为了矢量深组合技术在系统模型深入分析与进一步发展的瓶颈。本项目基于典型级联/非相干式矢量深组合系统结构，分析了以Kalman为前置滤波器的典型跟踪环路的优劣势。在此基础上，通过在S域建立Kalman前置滤波器的控制模型，得到系统参数配置机理及其对系统影响的数学描述，进而完成低复杂度的新型环路前置滤波器设计。在系统层次，通过环路各个组件的S域建模，建立深组合系统的跟踪环路模型，用于分析组合导航滤波器反馈延迟影响及鉴频器的改进优化。最后通过软/硬平台实现整个优化后的矢量深组合系统评估。本项目为矢量深组合系统的模型深入研究、优化及进一步发展提供了理论支撑。

GNSS/INS矢量深组合系统是GNSS与INS组合系统中最优、最难的解决方案，特别在弹载等高动态场景中，将是可靠性极高的系统。目前对于该方案的研究，主要的瓶颈是理论和实际工程无法对应。因此对于该解决方案的深层次理论剖析是十分必要的。项目基于GNSS/INS深组合背景，分别从理论建模、分析及软硬件平台的验证，完成了系统的理论架构，为后续的技术落地提供了理论指导。主要完成的研究内容包括：Kalman跟踪环路的理论模型建立与分析；Kalman跟踪环路与传统跟踪环路的对比；Kalman+VDFLL架构的矢量跟踪环路模型建立与分析；INS精度、时钟精度、闭环回路延时对系统的影响分析；基于“深组合”理论的技术拓展及应用。项目清晰、有效地解决了GNSS/INS矢量深组合系统的瓶颈问题，并提出了系统配置理论方案及工程实现的理论优化策略，为后续的民用、军用领域的技术落地提供了完整的指导。依托本项目发表SCI期刊论文5篇，国际会议2篇，国际报告3次，申请专利4篇，培养博士后1名，博士毕业生1名，硕士毕业生1名，在读博士生1名。依托项目的深组合导航技术，参与获得了2个省部级二等奖，并在2个重点项目平台中得到具体的应用。