论文部分内容阅读
针对轻型星敏感器、微机电系统(MEMS)陀螺、太阳敏感器和磁强计构成的轻型化卫星姿态确定系统,设计了一种分布式非线性融合滤波结构,提出一种快速采样点姿态估计算法,仅需4个采样点,即可实现完整的姿态及陀螺漂移估计,精度不低于常规采样点滤波,而运算量显著降低,与扩展Kalman滤波(EKF)算法相当。综合考虑器件特点,采用方差插值,融合子滤波器的相关信息,实现异类传感器的优势互补,从而获得高精度、高实时性、持续稳健的姿态及速率估计。仿真结果验证了算法的有效性。
Aiming at the light satellite attitude determination system composed of light star sensor, MEMS gyroscope, sun sensor and magnetometer, a distributed non-linear fusion filter structure is proposed and a fast sampling point attitude estimation Algorithm, only 4 sampling points are needed to complete the pose and gyro drift estimation. The accuracy is not less than that of the conventional sampling point filter, and the computational complexity is significantly reduced, which is equivalent to that of the extended Kalman filter (EKF) algorithm. Taking the characteristics of the device into consideration, the method of variance interpolation and sub-filter are used to realize the complementary advantages of heterogeneous sensors so as to obtain the high-accuracy, high-real-time and steady attitude and velocity estimation. Simulation results verify the effectiveness of the algorithm.