Research On Underwater Navigation Method Based On Terrain And Environmental Features

It is well-known that the errors of Inertial Navigation System are increasing with time. To achieve the goal of long-time and high-precision navigation for an underwater vehicle, an underwater navigation method which combines terrain and environmental features is discussed. The total terrain is divided into matching areas and unmatchable areas according to terrain information, and TAN is used to correct the error of INS in matching areas while SLAM is used to correct the error of INS in unmatchable areas. The main work is as follows:(1) Feature parameters which can measure terrain information of navigation area are introduced; a new matching area selecting method based on entropy weighted gray relational decision-making is provided to overcome the shortage of traditional selecting methods; finally, the effectiveness and advantages are verified in the simulation.(2) The basic principle of Terrain Aided Navigation based on ICCP algorithm is introduced; key technologies in ICCP algorithm are analyzed in detail; a system model of SINS/TAN is constructed and finally the effectiveness of the algorithm is verified.(3) The basic principle of Simultaneous Localization and Mapping is introduced; basic problems in SLAM, such as map representation, system model building, property of SLAM and data association, are analyzed in focus; and the Nearest Neighbor Data Association method is improved; finally the different performance characteristics are compared by the simulation.(4) Considering the particularity of underwater environment, an EKF-SLAM algorithm is adopted to correct the error of INS in unmatchable areas; the specific implementation process of EKF-SLAM of underwater vehicle, including prediction, observation and extension/updating, are presented in detail; finally, the effectiveness of the algorithm is verified.