| As a typical unmanned intelligent platform for ocean transportation,unmanned surface vehicle(USV)is capable of completing diversified marine operations.It is an indispensable carrier in marine development and national defense construction.The intelligent control of USVs has been increasingly concerned and studied by many countries.Most USVs have nonlinear characteristics,such as strong coupling,underactuated and state limitation,and the motion control will be affected by control constraints and external disturbances.Considering that the finite-time control technique has the advantages of fast convergence,strong disturbance rejection and high control accuracy,this dissertation investigates the path following and cooperative formation control problems of underactuated USVs based on the finite-time control method,and the main contributions are as follows:1.Path following guided by a parameterized path is investigated for a single USV.Firstly,aiming at the problem of slow convergence of position tracking errors,based on the traditional line-of-sight(LOS)method,a finite-time LOS guidance law is proposed so that the position tracking errors can converge to the origin rapidly within a finite time.Considering that the surge and sway velocities can hardly obtained from sensors in practice,a finite-time velocity observer is designed to identify the velocity components and compensate the time-varying sideslip angle in the guidance law.Considering the physical constraints of the actuator,a novel auxiliary dynamic system with fractional power terms is designed to solve the input saturation problem.The finite-time control laws for sway and yaw directions are developed based on the sliding mode technique.Cascade system theory verifies that the closed-loop path following system is uniformly global finite-time stable with fewer parameters to be adjusted.Then,considering the ocean currents and model uncertainties,the velocities of USV will be affected and there will be the relative speed.To solve the problem of unknown sideslip angle and ocean currents,a finite-time guidance-control method is designed based on finite-time adaptive predictor,fuzzy predictor and integral LOS guidance technique,such that the unknown sideslip angle and ocean currents can be estimated and compensated,simultaneously,and the model uncertainties can be approximated using the fuzzy predictor.The closed-loop system is proved to be practical finite-time stable.Simulation results demonstrate the effectiveness of the proposed schemes.2.Based on the study of path following for a single USV,the cooperative formation control for multiple USVs guided by a single path is investigated.Two formation control strategies are developed to achieve different formation pattern.Firstly,considering model uncertainties,external disturbances,and unknown USV velocity,a third-order finite-time observer is designed to calculate the lumped uncertainties and USV velocity rapidly and accurately.To save communication resources,different from most studies on undirected communication in cooperative formation control,combining with the tracking differentiator and distributed following guidance strategy,a finite-time cooperative formation control method law based on directed communication is proposed,in which the information transmission between neighboring USVs is unidirectional.The formation pattern that each USV maintains the same distance to its neighbours while following the dynamic target can be achieved.The closed-loop formation control system is proved to be practical finite-time stable.Then,a finite-time circular cooperative formation control method based on directed communication is designed using distributed circular guidance strategy and LOS technique.Meanwhile,an auxiliary dynamic system with the smoothing function is designed to deal with the actuator saturation problem in the control loop.The dynamic target tracking and equal arc length surrounding formation pattern can be achieved,so that all USVs are evenly distributed around the target at any time.The closed-loop formation control system is proved to be practical finite-time stable.Simulation results demonstrate the effectiveness of the proposed cooperative formation control schemes.3.Based on the study of cooperative formation control for multiple USVs guided by a single path,the cooperative formation control guided by multiple paths is investigated for the USVs,and the control object is chosen as the asymmetric USVs which are more realistic.Firstly,a two-order finite-time observer is proposed to accurately identify the lumped uncertainties.Combining with distributed guidance strategy,a finite-time cooperative containment control method based on event-triggered mechanism is developed,and the leader-containment formation pattern is achieved,in which the event-triggered mechanism is applied in the sensor-to-controller channel.The dynamic surface control technique is introduced to solve the jumping problem of the virtual control laws caused by event-triggered control.Different from the existing static triggering mechanism,a dynamic triggering mechanism based on the output information of the system is designed to ensure the high efficiency of triggering while minimizing the communication burden,thereby leading to a good control performance.The closed-loop system is proved to be practical finite-time stable.Then,considering the limited communication resources between USVs,the event-triggered mechanism based on formation error is designed for communication by using the proposed information estimator for neighboring USVs,and then a finite-time cooperative containment control method with aperiodic communication is proposed,which effectively reduces the communication burden.The closed-loop system is proved to be practical finite-time stable,and the Zeno behavior can be avoided.Simulation results demonstrate the effectiveness of the proposed control schemes. |