多敏感器联邦SSUKF融合姿态确定算法

针对四元数姿态估计问题,提出了一种分布式非线性滤波融合结构.通过引入基于超球面分布采样点变换(SSUT)技术的无迹卡尔曼滤波算法(SSUKF),以较低的计算量实现了高数据更新率、高精度的非线性滤波,并通过融合重构,保障系统无间断可靠工作,不受敏感器故障、视场盲区等因素影响.应用该算法对陀螺、磁强计、太阳敏感器、星敏感器构成的系统进行了具体设计并开展仿真研究,验证了算法的有效性.

Federated SSUKF for Multi-sensor Spacecraft Attitude Estimation

A federated version of spherical simplex unscented Kalman filter (SSUKF) was presented for space-borne orbit debris surveillance mission. The nonlinear transformation filtering method brought the algorithm a high precision, while the reduced sigma point filtering technique endowed the algorithm with a high computing speed, and the decentralized structure ensured the algorithm a good capability of reconstruction in case of sensor fault or vision limit. A filtering system composed of gyroscopes, magnetometer, sun sensor and star tracker was designed using this algorithm, and was validated with numerical simulation.