摘要
将UKF(Unscented Kalman Filter)方法用于惯性/重力组合导航系统。UKF方法设计了少量的呈高斯分布的σ点,在每个更新过程中,σ点随着非线性状态方程和测量方程传播,从而获得滤波值及较高的计算精度,而且避免了对非线性方程的线性化过程。仿真结果表明:UKF方法比传统卡尔曼滤波及其改进的滤波模型都有更高的估计精度,并能有效的克服非线性严重时出现的滤波发散问题。
This paper introduced Unscented Kalman Filter into gravity/inertial integrated navigation system. In Unscented Kalman Filtering, the updating of the filtering based on the nonlinear state space equation was realized by designing a few sigma points and calculating the propagation of these sigma points via nonlinear functions. The simulation results show that the UKF outperforms the standard KF and its improved model in accuracy and divergence performance.
出处
《测绘科学技术学报》
北大核心
2007年第5期371-373,共3页
Journal of Geomatics Science and Technology
