中文摘要：

动基座传递对准具有非线性非高斯的特征，运用卡尔曼滤波（EKF）这样基于局部线性化和高斯假设的滤波方法很难取得理想的滤波精度。超球面单形无迹粒子滤波（SSUPF）是一种基于SSUT变换的无迹粒子滤波，SSUT变换利用超球面分布的sigma点采样方式，减少了sigma点数量，对于高维系统，计算量大幅度减少。在滤波算法中，通过SSUKF产生重要性概率密度函数，引入了最新的观测数据，因此更接近于系统状态的后验概率。仿真结果表明：SSUPF相比EKF和UKF提高了传递对准的精度，达到与标准UPF相当的程度，并可以获得高于UPF的计算效率。

英文摘要：

With the nonlinear and non-gaassian features, moving-base tranfer alignment is hard to achieve ideal filter precision with EKF, which is based on partial linearization and gaussian hypothesis. SSUPF is a unscented particle filter based on SSUT which decreases the sigma points and thus cuts clown the calculation in high dimensional system with the hyperspherically distributed sigma points. The filter algorithm creates importance probability density function through SSUKF, and the newest observation data is introduced to better approximate the posterior probability of the system state. The simulation result indicates that campared with EKF and UKF, SSUPF enhances the filter precision to the degree similar with the standard UPF, and acquires the calculating efficiency higher than UPF.