基于时空梯度耦合虚拟目标的欠驱动AUV航迹控制研究

Autonomous Underwater Vehicle (AUV) is the important tool for undersea exploration and exploitation. Path tracking control is one of the key techniques which guarantee the successful implementation of various underwater operations and missions. However, the path tracking control design for underactuated AUV is far more challenging due to the inherently nonlinearity and second-order nonholonomic characteristics. Traditional trajectory tracking control scheme suffers from the stringent time limitation, easily rendering the saturation of actuators and the non-smooth actual trajectory of the vehicle. On the other hand, path following related to spatial domain cannot fulfill the mission with tight temporal requirements. Herein, a novel path tracking control design based on passivity theory and dynamics backstepping design, is proposed by introducing non-singular and collaborative virtual target which moves along the path with temporal-spatial gradient coupling characteristics. Thus, the novel controller can combine the advantages of both the trajectory tracking and path following methods, and achieve smooth tracking trajectory along with guaranteed temporal performance. Furthermore, robust adaptive control scheme is adopted for online estimation of the boundedness of unknown ocean currents. All the control schemes are developed under the framework of Lyapunov synthesis approach, such that the robustness and the stability of the whole closed loop path tracking control system is rigorous. Finally, the hardware-in-loop simulation, practical experiments in the tank and open water will be performed, in order to validate the effectiveness of the proposed path tracking control of underactuated AUV based on temporal-spatial gradient coupled virtual target.

智能水下机器人是海洋探测和资源开发的重要工具，航迹控制是其关键技术之一，也是水下机器人能否顺利完成各种作业任务的前提，而欠驱动水下机器人的本质非线性和二阶非完整特性使其控制设计更具挑战性。传统的轨迹跟踪控制受到时间域的限制，容易导致驱动器饱和以及产生非平滑的收敛轨迹；而空间域上的路径跟随控制又难以满足时间紧迫任务的要求。本项目拟通过引入非奇异的协作式虚拟运动目标，基于耗散理论、反演设计等非线性控制技术，以梯度最优的方式在时空域融合路径跟随和轨迹跟踪的优点，实现机器人在航迹空间域上的平滑收敛，同时保证系统在时域上的性能指标。通过采用鲁棒自适应技术对未知海流的上界进行在线自适应估计，并结合Lyapunov合成技术，保证机器人闭环控制系统的鲁棒性和稳定性。最后，分别通过半实物仿真、水池和开放水域试验，对所提出的基于时空梯度耦合虚拟目标的欠驱动水下机器人航迹控制算法进行验证和性能评估。

智能水下机器人是海洋探测和资源开发的重要工具，航迹控制是其关键技术之一，也是水下机器人能否顺利完成各种作业任务的前提，而欠驱动水下机器人的本质非线性和二阶非完整特性使其控制设计更具挑战性。本课题基于欠驱动水下机器人的动力学模型特点，结合时空耦合虚拟目标特性，采用非线性控制技术完成了航迹跟踪控制算法设计工作，为欠驱动水下机器人的运动控制技术提供了新的研究思路和方法。研究工作及其创新性主要体现为：1)完成了基于AUV本体特性和执行机构模型的欠驱动AUV动力学建模工作；2)在考虑欠驱动AUV测漂角特性的基础上，结合时空耦合虚拟目标的协作式运动特性，采用Lyapunov直接法和Backstepping反演控制技术完成了非线性航迹跟踪运动控制算法设计；3)最后，通过全数字仿真实验、半实物仿真实验和水池试验，对欠驱动AUV系统模型和运动控制算法进行了验证和性能评估。上述系统建模-算法设计-仿真试验等一体化的研究流程，验证了本项目研究工作的有效性，为欠驱动AUV航迹跟踪控制提供了重要的理论依据和实验经验，具有较高的理论价值和应用前景。