抗差自适应UKF的INS/GNSS组合导航算法

摘要

为了进一步提高滤波精度和稳定性,针对INS/GNSS组合导航系统的特点建立系统非线性误差模型,并提出一种抗差自适应UKF算法:基于UKF滤波中新息序列动态调整观测噪声数学统计特性,并结合残差序列构造自适应因子,改善初始协方差误差对标准UKF的影响,提高标准UKF自适应能力,减少存在观测粗差情况下滤波精度下降问题.仿真结果表明,该算法能够较好抑制观测粗差对定位精度及稳定性的影响.%In order to further improve the filtering accuracy and stability, the paper established the systematic nonlinear error model according to the characteristics of INS/GNSS integrated navigation system, and proposed a robust adaptive unscented Kalman filter algorithm: the statistical characteristics of the observation noise were dynamically estimated and adjusted based on innovation vectors of UKF, and the adaptive factors based on residual sequences were construct, the influence of the values of initial parameters on standard UKF was reduced, the adaptive capability of the standard UKF was enhanced, and the decreased filter accuracy when gross errors existed in the measurement values was improved.Simulation results showed that the method could effectively restrain the influence of observed gross errors on the positioning accuracy and stability of the integrated navigation.