Research On Trajectory Tracking And Cooperative Fencing Control Of A Quadrotor Underwater Vehicle

As a new intelligent oceanic equipment,autonomous underwater vehicles(AUVs)have been playing an increasingly important role in many applications such as marine resources exploration,environment monitoring,maritime investigation,and so on.On one hand,accurate trajectory tracking is the key guarantee for an AUV to perform the given mission.On the other hand,with the deepening of ocean development,the cooperation of multiple AUVs has gradually become a hot topic of current research.Compared with an individual AUV,multiple AUVs can achieve largescale underwater operations such as mine clearance,search,and rescue.However,it is challenging to achieve the trajectory tracking and cooperative control of AUVs considering the underactuated,highly nonlinear and strongly-coupled AUV dynamics,as well as the external disturbances induced by waves and currents.This thesis takes a novel quadrotor underwater vehicle as the research object,and conducts a series of researches on two key issues for the three-dimension trajectory tracking control and cooperative fencing control.Firstly,the kinematic and dynamic models of the quadrotor underwater vehicle are investigated,respectively.In particular,the “X” actuation system deployed in the vehicle is introduced,and the complete affine nonlinear model is derived.The nonlinear control theory and geometric control approach are used to analyze the nonholonomic property,stabilization,and small-time local controllability of the quadrotor underwater vehicle,laying a foundation for further control design.The second part is devoted to the 3-dimension trajectory tracking problem of the quadrotor underwater vehicle in the presence of input saturation constraints and external time-varying disturbances.A coordinate bias approach is used to design the kinematic controller based on input-output feedback linearization.The backstepping-based dynamic controller is then designed using the disturbance observer,the auxiliary system,and the dynamic surface control method.The dynamic surface control method avoids the differential operation of the traditional backstepping approach,which simplifies the control design.It is proved that all the signals of the trajectory tracking closed-loop system are uniformly ultimately bounded.Then,to address the cooperative moving-target-fencing problem of multiple quadrotor underwater vehicles,a coordinate bias approach is used to design the kinematic controller based on input-output feedback linearization.In addition,the adaptive backstepping controller and the robust backstepping controller are designed to deal with the constant and time-varying disturbances,respectively.The proposed controllers need neither the predefined stand-off distance/formation nor the global position information.Theoretical analysis shows that both controllers simultaneously assure target fencing,collision avoidance,and velocity estimation.Finally,considering the over-actuation of the quadrotor underwater vehicle subsystem in the cooperative moving-target-fencing task,we investigate the constrained control allocation problem with input saturation constraints.A null-space projection-based direct allocation(NDA)algorithm is proposed.In the NDA,if the desired moment vector belongs to the attainable moment set,a control allocation strategy is designed based on minimizing thruster energy consumption;Otherwise,a scaling factor is computed to mapping the desired moment vector into the boundary of the attainable moment set,such that the resulting allocated vector and the desired moment vector are in the same direction.The results are proved by rigorous theoretical analysis and verified by numerical simulation.