中文摘要：

自主水下航行器（Autonomous underwater vehicle,AUV）的协同导航是解决水下导航定位问题的重要方法,其中导航系统的定位误差增长特性是衡量其定位性能的关键指标.本文针对单领航者相对位置测量的多AUV协同导航系统,利用扩展卡尔曼滤波方法建立了导航系统的整体定位误差关于相对位置量测误差的传递方程.在此基础上,通过求解系统定位误差随时间演化的代数黎卡提方程,得到了其在稳态情形下的方差上界估计.理论分析表明,单领航AUV协同导航系统的整体定位误差有界收敛且与初始化滤波方差无关,具有良好的综合性能.最后,仿真实例验证了文中理论分析结果的正确性.

英文摘要：

The cooperative navigation of autonomous underwater vehicles（AUVs）is an important approach to solve the underwater navigation problems in which the localization error growth property is a key indicator.This paper deals with the performance analysis of the multi-AUV navigation system based on relative position measurements with a single leader.First,propagation equations are derived by using an extended Kalman filter.These equations can describe the system’s global positioning error related to the measurement errors.Then,the upper bound of the covariance estimation in the steady state is obtained by solving an algebraic Riccati equation which characterizes the evolution of the system’s positioning error.Theoretical analysis shows that the global positioning error of this navigation system can converge to an upper bound.Furthermore,the convergence is independent of the initial filter covariance,that is,the navigation system has a good comprehensive performance.Finally,simulation results validate the usefulness of the theoretical analysis.