中文摘要：

起源于移动机器人定位的同步构图定位（simultaneous localization and mapping，SLAM）算法可以克服基于先验图的水下导航算法的不足，实现水下载体的自主导航。本文给出了基于SLAM算法的水下导航算法的系统构建和执行过程，建立了状态方程和观测方程，讨论了新特征加入的问题，并利用模拟数据进行了仿真。仿真结果表明，无论是采用航渡还是区域探索航行方式，该导航算法的定位精度均高于单纯推位方法，具有广阔的应用前景。

英文摘要：

The simultaneous localization and mapping algorithm, which is originated from mobile robot localization, can overcome the disadvantages of the underwater navigation methods based on prior maps, hence realize true autonomous navigation of the underwater vehicles. The construction and execution of the underwater navigation system based on SLAM are introduced, and the state equation and observation equation are established. Furthermore, the issue of adding new features is discussed, and simulation data are used to test the system performance. Simulation results show that the algorithm can achieve better precision than pure dead-reckoning in both straight sailing and local searching patterns, and has a wide application prospect.