基于预测滤波的卫星交会对接相对位姿确定方法

Predictive Filter-Based Relative Attitude and Position Determination for Rendezvous and Docking between Spacecrafts

基于视频制导系统提供的视线测量,引入一种高精度、高鲁棒性的、实时的预测滤波算法,建立系统的误差模型,对角速度和位移速度进行一步预测,从而精确地估计卫星的相对位置和相对姿态。此外提出修正方案,通过卡尔曼增益对相对姿态和相对位移进行修正和更新,估计性能得到改进优化。通过卫星交会对接仿真实验验证算法的有效性,结果表明,预测滤波算法高精确估计相对姿态和相对位置,修正后算法得到了更高的精度结果,满足卫星交会对接要求,对在轨实验有重要参考价值。

A predictive filtering algorithm for real-time attitude and position determination with high accuracy and strong robustness is proposed in this paper.Based the line-of-sight measurement of Vision-based Navigation,a system errors model is established to predict the angular rate and linear rate one step and then the relative attitude and position between spacecrafts are accurately estimated.What is more,the predicted attitude and position are modified and updated by using Kalman gain and performance of algorithm is improved.Results of simulation based rendezvous and docking show that predictive filtering can accurately estimate the relative attitude and position,and the modified algorithm give a better result.The accuracy meets the requirement for rendezvous and docking.It has an important reference value for in-orbit experiment.