Abstract: The cooperative localization method fusing internal dead reckoning estimate and external measurement information of Unmanned Underwater Vehicles（UUV） is an effective means to improve the localization accuracy of the UUVs which are only equipped with low precise proprioceptive localization equipments.When the structure of the cooperative system is fixed，the choice for the filters should decide the improvement extent of the accuracy.The cooperative localization methods using the Sigma-Point Kalman Filters（SPKF） are proposed to be dead against the biggish truncation errors and the burdensome calculations when the Extended Kalman Filter（EKF） deals with the nonlinear systems.Compared with the EKF，the Unscented Kalman Filter（UKF） and the Central Difference Kalman Filter（CDKF） have better robustness and can improve the localization accuracy of a UUV further without increasing the computational complexity.The improving effects for the localization accuracy through the cooperative localization methods adopting the different filtering algorithms are compared，and the validity and the consistency of the cooperative localization methods utilizing the SPKF for UUVs are shown by the last simulation.

Key words: unmanned underwater vehicle, cooperative localization, extended Kalman filter, unscented Kalman filter, central difference Kalman filter, simulation

摘要： 融合无人水下航行器（UUV）内部航位递推估计和外部量测信息的协同定位方法是一种提高只配备低精度自定位装置的UUV定位精度的有效手段。当协同系统结构固定时，滤波器的选择就决定了精度提高的幅度。针对扩展卡尔曼滤波（EKF）在处理非线性系统时具有较大的截断误差和繁琐的计算，提出了使用sigma点卡尔曼滤波（SPKF）的协同定位方法。与EKF相比，无味卡尔曼滤波（UKF）和中心差分卡尔曼滤波（CDKF）具有更好的鲁棒性，在没有增加计算复杂度的基础上进一步提高了UUV的定位精度。仿真比较了采用不同滤波算法的协同定位方法提高定位精度的效果，验证了利用sigma点卡尔曼滤波的多UUV协同定位方法的有效性和一致性。

关键词: 无人水下航行器, 协同定位, 扩展卡尔曼滤波, 无味卡尔曼滤波, 中心差分卡尔曼滤波, 仿真