基于盒粒子滤波的低成本皮纳卫星高效姿态确定算法

针对低成本皮纳卫星姿态确定系统在质量、体积、计算量以及能耗等方面的限制问题，本文基于区间分析理论提出了卫星姿态区间化描述方法并建立了运动学区间化方程，提出了基于盒粒子滤波(BPF)的皮纳卫星姿态确定算法。该算法首先采用双矢量算法对太阳敏感器和磁强计得到的量测进行姿态解算，并将解算出的姿态四元数作为伪量测值输入传递给BPF，从而降低敏感器噪声对估计精度的影响。仿真实验表明，相比于传统粒子滤波的姿态确定算法，本文所提出的BPF姿态确定算法能够在保证姿态确定精度的同时大幅缩短算法运行时间。

Computationally Efficient Attitude Determination for Low Cost Pico /Nano Satellites Using Box Particle Filter#br#

In this paper, a box particle filter for attitude determination of low cost pico-/nano-satellites is proposed using interval based satellite attitude and kinematic equations. The proposed approach firstly invokes a three-axis attitude determination algorithm (TRIAD) to calculate the satellite’s attitude quaternion, which is then used in the box particle filter (BPF) as the attitude measurement to compress the impact of sensor noise on the estimation accuracy. The simulation study is conducted, and the results demonstrate that compared with the conventional particle filter attitude determination approach, the proposed BPF attitude determination approach significantly reduces the computational time while maintaining the accuracy of attitude determination.