基于抗差EKF的GNSS/INS紧组合算法研究

GNSS/INS Tightly Coupled Navigation Model Based on Robust EKF

提出了GNSS/INS紧组合导航的抗差EKF算法,采用21状态GNSS/INS紧组合状态方程,根据多余观测分量及预测残差统计构造抗差等价增益矩阵,建立抗差EKF算法,通过迭代给出GNSS/INS组合导航的抗差解,并开发GNSS/INS紧组合导航模拟平台,通过对观测值加入单粗差、多粗差及缓慢增长三类误差,测试本文算法对不同粗差的抑制能力。分析表明,抗差EKF可以将三类粗差抑制在相应观测值的残差中,达到削弱其对状态参数估计的影响。本文算例证明,抗差EKF算法可将导航解的误差精度从dm级提高为cm级甚至mm级,导航精度及可靠性得到明显提高。

A robust extended Kalman filter（EKF） algorithm is proposed and applied to tightly coupled GNSS/INS integrated navigation system.The algorithm is constructed by adopting a 21-demensional state equation and the redundancy and predicted residuals are used to build equivalent Kalman gain matrix.The iterative scheme is suggested for solving the GNSS/INS robust EKF solution.A simulation platform of GNSS/INS tightly integrated navigation system was developed and used to test the ability of the algorithm to reject three kinds of gross errors（single-outlier,multi-outlier,slow growth error）.The analysis indicates that the robust EKF algorithm can handle gross error by control the corresponding residual to reduce the influence of gross error on state parameter estimation.In the example,robust EKF algorithm can improve the precision of navigation solution from decimeter to centimeter or even millimeter and the reliability and precision are obviously improved.