基于新息正交性自适应滤波的惯性/地磁组合导航方法  

INS/GNS integrated method based on innovation orthogonality adaptive Kalman filter

作　　者：朱占龙[1] 单友东[2,3] 杨翼 年海涛 杨功流[1,5] 

机构地区：[1]东南大学仪器科学与工程学院,南京210096 [2]天津大学机械学院 [3]天津航海仪器研究所 [4]总装备部驻天津地区军事代表室 [5]北京航空航天大学仪器科学与光电工程学院

出　　处：《中国惯性技术学报》2015年第1期66-70,共5页Journal of Chinese Inertial Technology

基　　金：国家自然科学基金(61340044)

摘　　要：针对量测值存在野值以及量测噪声统计特性发生变化导致基于滤波的惯性/地磁组合导航系统滤波精度和稳定性下降的情形,提出一种基于新息正交性自适应滤波的惯性/地磁组合导航方法。首先通过新息过程正交性是否丧失进行量测值的野值辨识,然后对含野值量测值加权以减弱野值对系统滤波的不利影响,随后设置滑动窗口实时监测滤波器新息的动态变化以修正量测噪声方差,进而调整滤波器增益,提高系统的适应能力。仿真结果表明:在量测值存在野值以及噪声统计特性发生变化时,该方法能够抑制惯导系统的累积误差,并且得到的标准差指标相比标准EKF缩减至1/4左右,提高了组合导航系统的稳定性和适应性。The precision and stability of integrated INS/GNS(inertial navigation system/geomagnetic navigation system) system is usually decreased because the observed values may contain outliers and the statistical properties of noise is varied with time. In this paper, a method based on an innovation orthogonality adaptive Kalman filter is proposed. First, the outliers in observed values are detected by judging whether the orthogonality of innovation is lost or not. Then the proper weights are assigned to outliers so as to reduce the adverse influence on the system filter. Finally, the sliding window is set to monitor the dynamic alteration of innovation in the interest of modifying the noise variance, and the filter gain is adjusted to improve its adaptive ability. Simulation results prove that this method can restrain the accumulative errors of INS even if the observed values contain outliers and the noise is time-variant, in which the errors of standard deviation are decreased to 1/4 compared to conventional EKF, showing that the proposed method can improve the stability and adaptability of integrated INS/GNS system. ©, 2015, Editorial Department of Journal of Chinese Inertial Technology. All right reserved.

关 键 词：惯性/地磁组合导航 新息正交性 自适应滤波 野值 噪声 

分 类 号：V249.32[航空宇航科学与技术—飞行器设计]