基于联邦滤波算法的组合导航信息融合方法设计  

Design of Combined Navigation Information Fusion Method Based on Federated Filter Algorithm

作　　者：万晓帆 吕耀文[1] 刘伟[1] WAN Xiaofan;LV Yaowen;LIU Wei(School of Optoelectronic Engineering,Changchun University of Science and Technology,Changchun 130022)

机构地区：[1]长春理工大学光电工程学院

出　　处：《长春理工大学学报（自然科学版）》2019年第3期57-60,64,共5页Journal of Changchun University of Science and Technology(Natural Science Edition)

基　　金：吉林省重点科技攻关项目(20170204048GX)

摘　　要：为了满足飞行器高精度的导航需求,针对于传统的联邦滤波器中信息分配因子不能够实现动态的自适应系统的问题,设计了基于最小二乘加权的混合联邦滤波器,提出混合联邦滤波器中的主滤波器采用最小方差准则,根据最优系数加权的方法对子滤波器的最优估计进行融合,进而主滤波器实现实时、动态的信息分配因子更新,得到全局最优估计值。MATLAB仿真结果表明:相比于传统的联邦滤波器,基于最小二乘加权的混合联邦滤波器的角度误差、速度误差、位置误差均明显降低,且算法稳定、容错性强,对于实际工程具有参考意义。The traditional federal filter cannot realize the dynamic factors in the problem of adaptive system, so the mixed federal filter based on the weighted least-squares for the high-precision of aircraft navigation was designed. Minimum variance criterion was adopted in senior filter of mixed federal filter;the optimal estimation of the sub filter was fused according to the method of optimal coefficient weighting;then the optimal estimate was acquired with senior filter can update the information. The results of MATLAB simulation showed that the angle error,speed error and position error of the hybrid federated filter based on least square weighting were significantly reduced,and the algorithm was stable and fault-tolerant,which was reference significance for practical engineering.

关 键 词：组合导航 信息融合 卡尔曼滤波 联邦滤波器 

分 类 号：V249.3[航空宇航科学与技术—飞行器设计]