基于抗差EKF的INS/GNSS紧组合算法应用研究

摘要

Tightly coupled INS/GNSS based on Robust Extended Kalman Filter algorithm is researched aiming at the observation outliers of GNSS in the INS/GNSS integrated navigation.Firstly,the INS error equations resolved into the Local Lever Frame as well as the INS/GNSS tightly coupled measurement equations are given.Secondly,the resolved model based on least-square algorithm is constructed according to the measurement equations and the equal weight covariance is constructed by the least square residual statistic,then calculation steps are given.Finally,measured data are processed to verify the algorithm.Results show that:when the outliers exist in GNSS observations,the tightly coupled INS/GNSS integration based on Robust Extended Kalman Filter can weaken the impact of the observation outliers effectively,and can improve the precision of integrated navigation system.%针对INS/GNSS组合导航中GNSS观测值存在粗差的问题,研究了基于抗差EKF的紧组合导航算法.给出了INS在当地水平坐标系下的误差方程和INS/GNSS紧组合量测方程,根据量测方程建立了最小二乘解算模型,利用最小二乘残差统计量构造了等价协方差矩阵,给出了相应的计算过程,并利用实测数据对算法进行了验证.结果表明:当GNSS观测值中存在粗差时,基于抗差EKF的紧组合导航能够较好地削弱粗差观测值的影响,提高了组合导航系统的精度.