前后向迭代滤波在SINS/OD组合导航中的应用  

Application of Forward-backward Iterative Filter in SINS/OD Integrated Navigation

作　　者：黄湘远[1] 汤霞清[1] 武萌[1] 蔺博[1] 

机构地区：[1]装甲兵工程学院控制工程系,北京100072

出　　处：《导弹与航天运载技术》2016年第3期63-67,共5页Missiles and Space Vehicles

基　　金：军队计划项目资助课题

摘　　要：为了提高SINS/OD组合导航的精度,推导了载体坐标系下单位时间内位移增量之差的观测模型,利用残差x^2检测对OD信号进行两层故障检测和隔离,使用前后向迭代滤波完成误差参数估计和校正。利用实车实验分别对Kalman滤波、1次迭代滤波、2次迭代滤波进行对比分析,结果表明2次迭代滤波精度最高,但相对于1次迭代滤波精度提高效果不够明显。从滤波精度和计算量上综合考虑,SINS/OD组合导航中1次迭代滤波效果最好。To improve the navigation precision of SINS/OD integrated navigation system, the observation model of the deference between displacement increments per unit time in the body coordinate system is derived. The paper conducts the two-layer fault detection and isolation for the OD＇s signal using residual χ2 detection, and estimates and corrects of the error parameters by using forward-backward iterative filter. Vehicle test is conducted to comparatively analyze Kalman filter, one-iteration and two-iteration filters. The results show that the filtering precision of two-iteration filter is the best, which doesn＇t have an obvious improvement compared with one-iteration filter. For the comprehensive consideration of filtering precision and computational complexity, the filtering effect of one-iteration filer in SINS/OD integrated navigation system is the best.

关 键 词：故障检测 后向平滑 里程计 组合导航 

分 类 号：U666.1[交通运输工程—船舶及航道工程]