一种改进的强跟踪UKF算法及其在SINS大方位失准角初始对准中的应用

An Improved Strong Tracking UKF Algorithm and Its Application in SINS Initial Alignment Under Large Azimuth Misalignment Angles

针对现有的强跟踪无迹卡尔曼滤波(UKF)算法存在理论依据不足和滤波性能欠佳等问题,从正交性原理出发,通过严谨的推导得到强跟踪UKF成立的充分条件,在此基础上提出一种改进的强跟踪UKF算法。该算法无需求解雅可比矩阵且计算量较小,渐消因子的作用位置以及求解公式均不同于原始的强跟踪滤波器。给出了该算法的流程和渐消因子的求解方法,证明了该算法满足强跟踪滤波器的充分条件,并分析了其渐消因子的作用机理。进行了捷联惯性导航系统(SINS)大方位失准角初始对准仿真,结果验证了所提强跟踪UKF算法的正确性和有效性。

Against the lack of theoretical basis and poor filtering performance of the existing strong tracking unscented Kal-man filter （UKF） algorithms, the sufficient condition of a strong tracking UKF is rigorously derived in this paper from the or-thogonality principle, based on which an improved strong tracking UKF algorithm is proposed. This new algorithm requires less computation than the existing strong tracking UKF since it does not have to calculate the Jacobian matrix, and it is differ-ent from the original strong tracking filter in both the position and solution of the fading factor. This study presents the algo-rithm flow and the solution to the fading factor, and proves that the improved algorithm satisfies the sufficient condition of strong tracking UKF. Furthermore, the action mechanism of the fading factor is analyzed. The results of the strapdown iner-tial navigation system （SINS） initial alignment simulation under large azimuth misalignment angles verify the validity and ef-fectiveness of the improved strong tracking UKF algorithm.