UKF在深组合GPS/INS导航系统中的应用

采用联邦滤波的深组合GPS/INS导航系统预滤波器量测模型具有很强的非线性,导致扩展卡尔曼滤波（EKF）的预滤波器估计精度不高。Unscented卡尔曼滤波（UKF）方法是一种非线性分布近似方法,它使用有限数量的sigma点去逼近整个非线性动态系统的分布可能,从而避免了对非线性测量模型进行线性化,具有较高的精度和较好的鲁棒性。在分析深组合导航系统预滤波器模型和UKF原理的基础上,设计了基于UKF滤波算法的预滤波器,对码相位误差、载波相位误差、载波频率误差、载波频率变化率等参数进行估计,同时将UKF和EKF算法进行了仿真比较。结果表明,在深组合导航系统中使用UKF滤波比EKF有更高的导航定位精度。

Application of UKF in deeply coupled GPS/INS navigation system

A pre-filter measurement model of deeply coupled GPS/INS navigation system adopting the Federated Filter has strong nonlinearity, so the estimation accuracy of extended Kalman filter's pre-filter is low. Since the Unscented Kalman Filter (UKF) adopts an approximation method with nonlinear distribution, which uses a finite number of sigma points to propagate the probability of state distribution through the nonlinear dynamics of system, it has higher precision and reliability, and needn't linearize the nonlinear measurement model. Based on the analysis of the pre-filter model of deeply coupled navigation system and the principle of UKF, a pre-filter based on UKF is designed for estimating the parameters, such as code phase error, carrier phase error, carrier frequency error and carrier frequency change-rate. The simulation results show that the deeply integrated navigation system using UKF has higher navigation precision than that of using EKF.