Modeling And Path Tracking Control Of Unmanned Surface Vessel

Under the interference of various uncertain marine environmental factors,the use of effective control algorithms to achieve path tracking control is the premise that Unmanned Surface Vessel(USV)can successfully complete scientific research,hydrographic surveying,search and rescue patrol and other tasks.Due to the nonlinearity,high order and uncertainty of USV,it is very difficult to establish accurate dynamic models.In the existing research,the reference input of most controllers is expected surge speed and ship course angle,defining it as a function of time may cause underactuated USV to deviate from the desired path.In order to solve the above problems,this paper integrates ship navigation system and course control algorithm,and carries out research on control algorithm of the USV.Firstly,based on the separated mechanism model,the six-degree-of-freedom motion mathematical model and disturbance model of the USV are established,which lays the foundation for the design of the subsequent controller.Guidance and control are two important modules for ship control systems to achieve complex automatic navigation tasks.In terms of guidance,aiming at the drawbacks of the current Line of Sight(LOS)guidance algorithm,such as the slow convergence speed of the USV to the desired path when the track error is large,and the automatic steering near the route point position,an improved LOS guidance algorithm is proposed.The algorithm can guide the USV to quickly converge to the route when the track error is large,and provides a reasonable navigation mechanism during the route reference point switching process.In terms of control,aiming at the problem that fixed gain PID controller can not usually provide satisfactory control performance at certain operating points,and the difficulty of tuning control parameters,this paper designs a PID-GA controller.The Genetic Algorithm(GA)is used to select the optimal gain parameters for different tracks and sea conditions,and the gain adjustment is performed online according to the changes of the track and sea conditions.The simulation results show that the designed PID-GA controller has better adaptive ability and anti-interference ability than the traditional PID controller,but still can not achieve intelligent control.Finally,aiming at the inherent nonlinearity,uncertainty and underactuation of ship maneuvering,as well as the inability to obtain an accurate mathematical model of USV motion,this paper designs a DDPG-H controller by using the Deep Deterministic Policy Gradient(DDPG)algorithm in reinforcement learning,and realizes intelligent control.The DDPG-H controller controls the USV to track the known trajectory through real-time feedback of the environment and autonomous learning.The simulation results show that the DDPG-H controller has good control performance and can complete high-precision path tracking.