论文部分内容阅读
初始对准是实现惯性导航高精度的一项关键技术。无迹滤波(UKF)在SINS系统静基座大方位失准角初始对准中计算量大,在不精确或错误的噪声统计情况下,收敛速度变慢,估计精度下降,甚至滤波发散。针对这一问题,将超球体采样与强跟踪无迹滤波(STFUKF)算法相结合,提高了运算速度和对准精度。利用SINS的非线性误差模型,通过数字仿真将卡尔曼滤波、UKF和STFUKF的性能进行比较,证明该方法具有精度高、抗干扰性好、跟踪能力强的特点。
Initial alignment is one of the key technologies to achieve high accuracy of inertial navigation. Unscented filtering (UKF) is computationally intensive in the initial alignment of large azimuth misalignment of static base of SINS system. In the case of inaccurate or incorrect noise statistics, the convergence speed is slow, the estimation accuracy is decreased and even the filtering is divergent. In response to this problem, the combination of hyperspherical sampling and STFUKF algorithm improves the speed of operation and the alignment accuracy. The SINS nonlinear error model is used to compare the performance of Kalman filter, UKF and STFUKF through digital simulation. The results show that this method has the characteristics of high precision, good anti-interference and strong tracking ability.