基于VB-AUKF的GNSS/SINS组合导航信息融合算法  

Information Fusion Algorithm of GNSS/SINS Integrated Navigation System Based on VB-AUKF

作　　者：孙玉梅 潘新龙 王美春 SUN Yumei;PAN Xinlong;WANG Meichun(School of Computer Science,Weifang University of Science and Technology,Weifang 262700,Shandong,China;Naval Aviation University,Yantai 264001,Shandong,China)

机构地区：[1]潍坊科技学院计算机学院,山东潍坊262700 [2]海军航空大学,山东烟台264001

出　　处：《弹箭与制导学报》2023年第3期1-8,共8页Journal of Projectiles,Rockets,Missiles and Guidance

基　　金：山东省重点研发计划(2020CXGC010701);山东省自然基金(ZR2020MF154,ZR2016FM28);高层次人才科研启动项目(KJRC2022009)资助。

摘　　要：针对GNSS/SINS组合导航系统的标准无迹卡尔曼滤波UKF算法不能自适应调节量测噪声均方差的缺陷,首先提出了该组合导航系统的变分贝叶斯自适应UKF算法(variational bayesian adaptive UKF,VB-AUKF),然后建立了VB-AUKF算法的量测噪声均方差迭代估计模型;最后针对影响变分贝叶斯估计精度的3个参数及基于VB-AUKF算法的组合导航系统进行理论仿真计算。结果表明:VB-AUKF算法的调节因子是决定量测噪声均方差估计精度的关键因素,其他2个参数的选择仅仅影响量测噪声均方差估计初值;当量测噪声均方差发生变化时,VB-AUKF算法可以自适应估计量测噪声均方差;相对于标准UKF算法,VB-AUKF算法可以提高位置及速度精度约为12%及51%,同时大幅提高姿态估计精度。This paper firstly proposes a variational Bayesian adaptive UKF algorithm(VB-AUKF) for this integrated navigation system,aiming at the defect that the standard UKF algorithm of GNSS/SINS integrated navigation system cannot adaptively adjust the root mean square(RMS) of measurement noise.Then,an iterative estimation method for RMS of measurement noise is proposed in VB-UKF algorithm.Finally,a lot of theoretical simulation experiments are carried out for the three parameters that affect the accuracy of the variable Bayesian estimation and the integrated navigation system based on VB-AUKF algorithm.The experimental results show that the moderating factor of VB-AUKF algorithm is the key factor to determine the estimation accuracy of the RMS of the measurement noise,while the other two parameters only affect the initial value of the RMS estimation of the measurement noise;the VB-AUKF algorithm can adaptively estimate this RMS when the RMS of measurement noise changes;and compared with the standard UKF algorithm,the VB-AUKF algorithm can improve the accuracy of position and velocity by about 12% and 51%,and greatly improve the accuracy of attitude estimation.

关 键 词：变分贝叶斯 调节因子 组合导航 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]