海底地形匹配导航方法，不需要外部声学基阵支持或定期上浮修正累积误差，对AUV长期水下潜航和作业具有积极意义。本项目拟在青年科学基金的基础上，针对声学高度计获取水下“点、线”地形的匹配方法在实时性和导航精度上的不足，采用多波束声纳瞬时获取水下“面地形”的方法，研究AUV的海底“面地形”快速匹配导航的机理与算法。建立真实海洋环境下海底地形的高精度基准数字地图，探讨大批量测深数据的三维关联滤波和水下“面地形”信息的特征提取机理，研究高置信度的海底“面地形”快速匹配方法和理论。进行多种导航模式的数据融合，开展仿真、水池和海中实验研究，为解决AUV“长期、隐蔽、全天候、高精度”导航奠定理论和技术基础。

Seabed terrain matching method does not need the arrangement of acoustic arrays or periodic updating the accumulated error. It is significant to long-range underwater navigation of Autonomous Underwater Vehicle (AUV). To improve real-time function and precision of “point-line” terrain matching method (supported by National Natural Science Foundation for Young Scholar), we will adopt the multi-beam sonar to get the instantaneous “seabed surface terrain”, research the mechanics of rapid “seabed surface terrain matching” method. The accurate digital seabed map will be build. Three-dimensional correlative filter, feature extraction of “surface terrain” and rapid “seabed surface terrain matching” method will be discussed. Data fusion of multi-navigation mode is to be carried out. The method will be tested in simulation, pool and sea trial, which can be developed to the theory and technology of “long-range, hidden, weatherproof, accurate” underwater navigation for AUV.