中文摘要：

为了提高组合导航定位系统的定位精度和可靠性，分别对扩展卡尔曼滤波（EKF）、粒子滤波（PF）和U卡尔曼滤波（UKF）3种算法进行了分析。通过分析3种算法各自的特点，将PF算法和UKF算法的优点相结合，提出了一种新的粒子滤波算法——U粒子滤波（UPF）算法，并将其应用于GPS／DR组合导航系统中。通过对UPF算法与PF算法在GPS／DR组合导航系统中的仿真研究比较，进一步证实了UPF算法的可行性及计算的精确性。

英文摘要：

In order to improve the positioning precision and reliability, extended Kalman filter （EKF）, particle filter （PF） and unscented Kalman filter （UKF） algorithms were analyzed respectively. A new particle filter algorithm called unscented particle filter （UPF） algorithm was given based on the analysis of each algorithm＇s characteristic and the combination of PF＇s and UKF＇s advantages, and it was applied to GPS/DR integrated navigation system. Finally, the feasibility and calculation accuracy of UPF algorithm were proved with a simulation and comparison of UPF and PF algorithms in GPS/DR integrated navigation system.