基于无迹粒子滤波的飞行器再入段传递对准方法

Transfer alignment for reentry spacecraft based on unscented particle filter in space platform

针对再入式飞行器再入返回前需要精确和快速的传递对准需求以及空间环境特殊性,提出了一种适用于再入式飞行器空间平台对准方法,建立J2000坐标系下光纤陀螺惯组的系统状态方程,利用星敏感器输出的姿态和光纤陀螺惯组输出的姿态差值作为观测量,实现飞行器的"姿态"匹配传递对准算法。针对系统状态方程和量测方程的非线性特性以及单独使用无迹卡尔曼滤波(UKF)方法存在滤波不稳定性和单独使用粒子滤波(PF)方法存在粒子退化导致估计误差变大的问题,提出了无迹粒子滤波(UPF)算法,采用卫星工具包(STK)软件模拟飞行器在轨运行进行仿真验证。仿真结果表明:该方法可以在3s内使得失准角的估计误差小于0.6',估计精度提高了30%,收敛时间缩短了70%,相比传统方法提高了对准的精确性和快速性。

A transfer alignment method for reentry spacecraft in view of the space environment＇s particularity is proposed to serve the need of rapid and precise transfer alignment. The state and measurement equations of the fiber optic gyroscope inertial measurement unit（FOG IMU） are established, and by using the output attitude difference between the outputs of FOG IMU and star sensor, a transfer alignment algorithm is achieved. Since only using UKF has the problems of filtering instability, and only using PF has the problems of particle degradation, an algorithm based on unscented particle filter（UPF） method is presented on account of the nonlinear feature of the state and measurement equations. Simulations are made by using a satellite tool kit（STK） tool to simulate the spacecraft its operations on-orbit, which show that, compared with its traditional method, the misalignment estimation can converge to 0.6 arcmin within 3s, the estimation accuracy is increased by 30%, and convergence time is shorten by 70%。