基于DR/USBL的HOV组合导航算法

DR/USBL Integrated Navigation Algorithm for HOV

针对近海底作业的深海载人潜水器(HOV),提出了一种经济可行的组合导航算法。该算法利用HOV搭载的多普勒测速仪、捷联罗经构成航位推算(DR)系统,结合超短基线(USBL)水声定位系统提供的位置信息设计了一个四维卡尔曼滤波器构成DR/USBL组合导航系统。考虑到USBL实测定位数据中包含大量野值和少量缺失值,设计了一种利用DR辅助USBL的在线数据清洗方案,该方案具有较强的鲁棒性,能够为DR/USBL组合导航系统提供连续一致的量测数据。基于海试数据的半物理仿真结果表明,DR/USBL组合导航系统能够有效地融合2个系统的优点,输出更平滑精度更高的位置信息,定位误差不超过25 m。

An economically feasible integrated navigation algorithm is designed for deep-sea human-occupied vehicles(HOV)operating near the seafloor.The algorithm utilizes the Doppler velocity log and a strapped-down compass carried by the HOV to form a dead reckoning(DR)system.A four-dimensional Kalman filter is designed to integrate DR with the position information provided by an ultra-short baseline(USBL)acoustic positioning system to constitute a DR/USBL integrated navigation system.Considering that the actual measurements of the USBL usually contain a large number of outliers with few missing values,an online data-cleaning scheme using DR is designed.This design is robust and can provide continuous and consistent measurements for DR/USBL.Finally,semi-physical simulation results based on sea trial data show that the DR/USBL integrated navigation system can effectively integrate the advantages of the two systems to provide smoother and more accurate positioning information;the positioning error is within 25 m.