基于新息正交原理的抗差UKF及其在INS/GPS组合导航中的应用

为提高标准UKF对异常的量测噪声统计的鲁棒性,提出了一种基于新息正交原理的抗差UKF算法.该算法根据新息序列的正交性确定最优的抗差因子,而后通过对新息协方差阵引入抗差因子在线调整滤波增益,进而抑制异常量测对滤波解的影响.将提出的算法应用于INS/GPS组合导航系统进行仿真验证,并与标准UKF和现有的抗差UKF进行比较,结果表明,当量测噪声统计不准确时,提出的基于新息正交原理的抗差UKF滤波性能明显优于上述两种算法,提高了组合导航系统的定位精度.

Innovation Orthogonality Based Robust UKF and Its Application In INS/GPS Integration Navigation

An innovation orthogonality based robust UKF is presented to improve the robustness of the standard UKF against the abnormal measurement noise statistics.The proposed algorithm determines the optimal robust factor based on the orthogonality of innovation sequence.Subsequently,the robust factor is introduced into the innovation covariance to adjust the Kalman gain matrix online,and further resist the impact of the abnormal measurement on the filtering solution.The proposed algorithm is applied to the INS/GPS integrated navigation system for simulation in comparison with the standard UKF and existing robust UKF.The simulation results demonstrate that the filtering performance of the proposed innovation orthogonality based robust UKF is significant superior to the above two algorithms,thus the positioning precision of the integrated navigation system is improved.