大尺度欠驱动高速AUV导航系统研制

Development of navigation system for large-scale and high-speed underactuated AUV

[目的]针对某新型大尺度欠驱动高速自主式水下机器人(AUV)导航系统的研制,[方法]详细介绍AUV导航系统的硬件组成及软件构架,构建基于VxWorks实时嵌入式操作系统的AUV组合导航系统。考虑到高速AUV导航系统的非线性特征及容错性要求,将无迹卡尔曼滤波(UKF)技术用于联邦滤波器,设计基于UKF的联邦卡尔曼滤波器,并对导航系统的二维平面模型进行数值仿真。[结果]结果表明,联邦无迹卡尔曼滤波器相比传统的卡尔曼滤波方法能够获得更高的精度和更好的鲁棒性;半实物仿真联调试验、水池试验、湖上试验及艇载试验充分验证了导航系统的实时性和可靠性,[结论]证明该系统能够满足AUV航行任务的需要。

[Objectives]This paper details the development of a navigation system for a new model of large-scale and high-speed underactuated Autonomous Underwater Vehicle(AUV).[Methods]First,the hardware composition and software architecture of AUV navigation system are introduced in detail,and the integrated navigation system of the AUV is constructed on the basis of the VxWorks real-time embedded operating system.Considering the non-linear characteristics and fault-tolerance requirements of high-speed AUV navigation systems,Unscented Kalman Filter(UKF)technology is used in the federated filters,and a federated filter based on UKF is designed.[Results]A two-dimensional plane model of the navigation system is simulated,showing that the federal unscented Kalman filter can obtain higher accuracy and better robustness than the traditional Kalman filter method.Finally,the real-time operation and reliability of the navigation system are verified via a semi-physical simulation debugging experiment,pool test,lake test and boat test.[Conclusions]The results show that this navigation system can meet the needs of the AUV in completing its missions.