组合导航系统UKF滤波算法设计

Design of Filter in Micro-Integrated Navigation System

将平淡Kalman滤波(UKF)方法应用于低成本SINS/GPS/MC组合导航系统,建立了系统的非线性误差方程,并设计了相应的UKF算法与滤波器;针对组合导航中捷联惯性导航子系统的反馈修正,提出了一种根据误差角对四元数进行修正的方法.试验结果表明,本文设计的UKF效果优于集中Kalman滤波,且采用四元数反馈修正方法后较大地提高了滤波精度.

The unscented Kalman filter （UKF） with feedback was applied in an integrated-navigation system which composed of inertial navigation system （INS）, global positioning system （GPS） and magnetic compass （MC）. The design of the UKF was described based on the nonlinear model of the system. A novel method by modifying the quaternion according to the feedback UKF filtering results was presented in order to reduce the attitude error. The experimental results show that the presented nonlinear UKF obtains a better performance compared to the centralized Kalman filter. Furthermore, the feedback of the filtering results can apparently increase the precision of the navigation results.