容错联邦强跟踪卡尔曼滤波算法在组合导航中的应用研究

提出了一种改进的联邦强跟踪卡尔曼滤波算法，并将其应用于MSINS／GPS浅组合中。采用了三层容错结构，能够有效地检测出数据粗大误差、GPS速度跳变以及滤波器误差估计的故障，及时进行故障隔离，从而使系统具有良好的容错能力。仿真结果表明，联邦强跟踪卡尔曼滤波算法解决了量测相关、GPS速度跳变和初值选取敏感性等问题，可以提高系统的实时性、鲁棒性和容错性能。

Application of Fault Tolerant Federated Strong Tracking Kalman Algorithm in Integrated Navigation System

An improved federated strong tracking kalman filter has been proposed creatively and has been used in loose cou- pling MSINS/GPS integrated navigation system. Adopting three--layer fault--tolerant structure, it ean effectively check the errors of data outlier, GPS＇s velocity jump and the malfunction of the filter state estimate. When the system fault has been diagnosed, it can be insulated in time, which can make this system have better ability of toleraming the faults. The simulation results show that this approach can resolve the problem of the correlation of GPS＇s position and velocity measurement data, GPS＇s velocity jump and the system＇s sensitivity to initial values. Thus, the real time character, robust and the ability of tolerating faults of tile system has been improved.