传统联邦卡尔曼滤波算法的高容错性是以牺牲精度为代价的,为此提出了改进的联邦滤波算法,采用分时融合反馈的联邦滤波信息分配方式,在提高系统容错性的同时兼顾导航精度和实时性,而且无融合反馈阶段延长了故障反应时间,提高了系统对渐变故障的检测能力,取得较为理想的结果。 The traditional federation Kalman filtering algorithm is highly fault-tolerant at the expense of accuracy. Therefore, an improved federated filtering algorithm is proposed, and the federated filtering information distribution method based on time-sharing fusion feedback is adopted to improve system fault tolerance and balance navigation accuracy And real-time, and no fusion feedback phase to extend the fault reaction time and improve the ability of the system to detect gradual failure, achieve better results.