基于UKF的INS/GNSS/CNS组合导航最优数据融合方法  

UKF-based optimal data fusion method for integrated INS/GNSS/CNS

作　　者：孟阳[1] 高社生[1] 高兵兵[1] 王维[1] MENG Yang GAO She-sheng GAO Bing-bing WANG Wei(School of Automatics, Northwestern Polytechnical University, Xi＇an 710072, China)

机构地区：[1]西北工业大学自动化学院,西安710072

出　　处：《中国惯性技术学报》2016年第6期746-751,共6页Journal of Chinese Inertial Technology

基　　金：国家自然科学基金(61174193);航天科技创新基金(2014-HT-XGD)资助课题

摘　　要：为了提高INS/GNSS/CNS组合系统的导航精度,提出了一种基于UKF的多传感器最优数据融合方法。该方法具有两层融合结构,第一层中,GNSS和CNS分别通过两个局部UKF滤波器与INS组合,以并行的方式获得INS/GNSS和INS/CNS子系统的局部最优状态估值;第二层中,根据线性最小方差准则推导出一种矩阵加权数据融合算法,对局部状态估值进行融合,获取系统状态的全局最优估计。提出的方法无需采用方差上界技术对局部状态进行去相关处理,克服了联邦卡尔曼滤波(FKF)及其优化形式存在的缺陷。仿真结果表明,相比于FKF,提出方法的导航精度可至少提高36.4%;相比于UKF-FKF,其导航精度也可至少提高21.0%。An UKF-based multi-sensor optimal data fusion algorithm is presented to improve the navigation accuracy of integrated INS/GNSS/CNS system.It＇s fusion structure has two levels：on the first level, GNSS and CNS are respectively integrated with INS by two local UKFs to obtain the local optimal state estimates in parallel; on the second level, a matrix weighted data fusion algorithm is derived based on the rule of linear minimum variance to fuse the local state estimates for generating the global optimal state estimation.Since the proposed method refrains using the upper bound technique to eliminate the correlation between local states, it overcomes the limitations of the federated Kalman filter（FKF） and its improved forms.Simulation results demonstrate that the navigation accuracy achieved by the proposed method is at least 36.4% higher than that by FKF and 21.0% higher than that by UKF-FKF.

关 键 词：INS/GNSS/CNS组合系统 数据融合 联邦卡尔曼滤波 无迹卡尔曼滤波 

分 类 号：V249.3[航空宇航科学与技术—飞行器设计]