中文摘要：

摘要：针对未知海域中水下机器人难以实现自主导航的问题，提出了一种基于偏振光／多普勒测速／SINS组合的自主导航方法。依据水下偏振光相机测向原理建立了偏振光／SINS姿态组合系统，依据多普勒测速原理建立了多普勒／SINS速度组合系统，并基于联邦Kalman滤波器实现了组合导航系统的信息融合。通过仿真实验验证，本文提出的组合导航方法相比于现有组合方法具有更好的定位精度和容错能力，能够满足水下机器人进行无源自主导航时的精度和可靠性的要求。

英文摘要：

Aiming at the problem that the underwater robot is difficult to realize autonomous naviga- tion in unknown waters, an integrated navigation method based on Polarized Light （PL）/Doppler Velocity （DVL）/Strapdown Inertial Navigation System （SINS） was proposed. Based on the principle of polarized light underwater camera direction, a PL/SINS attitude determination integrated system was established. Based on the principle of Doppler velocity measurement, a DVL/SINS speed determination integrated sys- tem was established. Moreover, based on federal Kalman filter, information fusion of integrated naviga- tion system was realized. The method proposed in this paper has better positioning accuracy and fault tol- erance compared to the existing combination method, realizing the accuracy and reliability when underwa- ter robot would implement autonomous navigation.