基于联邦滤波的多源融合导航算法

在复杂多变环境下，单一导航源的定位性能和鲁棒性会受到一定的影响。针对汽车、小型飞行器在城市、峡谷、卫星信号缺失或被遮挡以及导航信息源繁多等情况，研究了基于联邦滤波的多源融合导航算法。该算法综合利用了各种不同的信息源，经过多传感器的高度集成、多信息源的数据融合，生成时空基准统一且具有抗干扰、连续、可靠的PNT服务信息。设计的联邦滤波器采用两级结构，在子滤波器中进行局部估计后，在主滤波器中进行最优合成。此外，每个子滤波器加入了故障诊断算法，且结合自适应滤波理论进行信息因子的自适应分配，有效提高了故障检测能力。最后，通过实验验证了不同信息源组合的有效性，表明所设计的基于联邦滤波的多源融合算法可以提供稳定、可靠以及高精度的多源融合定位服务，具有一定的研究意义和实际价值。

Research on Multi-source Fusion Navigation Algorithm Based on Federated Filtering

In a complex and variable environment, the positioning performance and robustness of a single navigation source will be affected. The multi-source fusion navigation algorithm based on federated filtering is studied for the situations of automobile, small aircraft in city, canyon, satellite signal missing or blocked, and many navigation information sources. The algorithm uses various information sources, through multi-sensor high integration and multi-information data fusion, to generate a unified spatio-temporal benchmark with anti-interference, continuous and reliable PNT service information. The federated filter designed in this paper adopts a two-stage structure. After local estimation in the sub-filter, the optimal synthesis is carried out in the main filter. In addition, each sub-filter is added with a fault diagnosis algorithm, and the information factor adaptive allocation is combined with the adaptive filtering theory which effectively improves the fault detection capability. Finally, the effectiveness of different information source combinations is verified by experiments. It shows that the designed multi-source fusion algorithm based on federated filtering can provide stable, reliable and high-precision multi-source fusion positioning services, which has certain research significance and practical value.