联邦滤波器信息分配因子优选问题

Study on Optimized Algorithm for Information-Sharing Factors of Federated Filter

在剖析联邦滤波核心理论的基础上,对当前学术热点——联邦滤波器信息分配因子优选问题进行探讨。分析了文献中几种常见的优选算法,指出其中的理论缺陷。从全局估计、局部估计、容错性等三方面质疑了信息分配因子优选的必要性和可行性,提出一些值得商榷之处,并得出结论。通过S IN S/GPS/TAN/SAR四组合导航系统联邦滤波器仿真分析,验证了观点的有效性。

The federated filter theory and information sharing principle is discussed. Current optimized selection methods for information-sharing factors are reviewed, and some underlying assumptions and flaws are pointed out. In terms of global estimation accuracy, local estimation accuracy and fault tolerance, the necessity and feasibility of optimized algorithms for information-sharing factors are oppugned, which draws some reverse conclusions to previous literature. Simulation results are presented for the SINS/GPS/TAN/SAR integrated navigation system, which demonstrated the validity of the opinion of this paper.