基于改进型两步卡尔曼滤波的相对导航方法

摘要

采用Clohessy-Wiltshire(C-W)方程描述的近圆轨道相对导航状态方程具有线性的形式,而以航天器相对距离和相对方位作为测量信息的观测方程是非线性的,针对近圆轨道航天器相对导航的这一特点,给出了采用两步卡尔曼滤波(Two Step Kalman Filtering,TSF)的相对导航算法,并且利用Unscented变换方法,解决了两步卡尔曼滤波的状态初值确定问题,给出了TSF的完整算法.数值仿真比较了TSF和扩展卡尔曼滤波(Extended Kalman Filtering,EKF)、无迹卡尔曼滤波(Unscented Kalman Filtering,UKF)等算法的性能,验证了采用TSF方法实现相对导航的可行性和有效性.%For the spacecraft relative navigation in the circular orbits, the linear ClohessyWiltshire (C-W) equations were used as the state equations, and the nonlinear equations for the relative distance and azimuth between two spacecraft were used as the observation equations.The improved two-step Kalman filtering (TSF) was designed for the special relative navigation problem.The unscented transformation is used for determining the initial distribution about the state vector, and the intact form of the two step Kalman filtering was obtained.The simulation results validate the availability of this navigation method, and show that the TSF can provide better performance than extended Kalman filtering (EKF) and unscented Kalman filtering (UKF).