欠驱动水下机器人(AUV)由于在某些自由度上没有直接驱动力，减少了控制输入量，其运动控制较全驱动AUV困难得多，且受海流扰动影响非常严重。本项目拟对复杂流场运动及欠驱动AUV三维空间下路径跟踪控制方法进行深入研究。根据海流速度测量数据，探明定常流和奇异流的关系，辨识海流奇点位置和奇异流形态，揭示复杂流场运动规律，获得复杂流场模型。依据级联系统理论将跟踪控制系统分解为运动学控制和动力学控制两个相互级联的子系统，独立设计控制率，能够快速跟踪参考路径的同时对海流扰动和模型不确定有强鲁棒性。建立海流观测器估计完全未知的海流速度，为控制系统提供更为有效的反馈参量。通过WL-2欠驱动AUV的试验研究验证方法的可行性和有效性，为解决海流扰动下欠驱动AUV三维空间路径跟踪控制问题提供理论依据，为智能化、远程化AUV的研究和应用提供技术支持。

Path tracking control for underactuated AUVs is more difficult than full drive AUVs because there are not direct driving forces in some degrees of freedom, which reduces the control inputs. It is influenced seriously by ocean current disturbance. Therefore, motion of complex flow field and three-dimensional path tracking control for underactuated AUVs will be discussed deeply in this program. According to the measure data of current velocity, we prove the relationship between steady flow and singular flow, and identify the position of singular point and the form of singular flow, and reveal the motion rule of flow field. According to cascade system theory, we decompose the tracking control system into two cascade subsystems which are kinematic control system and dynamic control system and design control laws independently. The system can not only track the reference path quickly, but also be robust strongly to parametric modeling uncertainty. An current observer is also added to estimate unknown current velocity, which can supply more effective feedback valves for the control system. Experiments are carried out based on WL-2 underactuated AUV to verify the feasibility and validity. The work can provide the theoretical basis to path tracking control for underactuated AUVs aiming at ocean current disturbance, and provide the technical support to the research and application for intelligence and long-distance of AUVs.