Research On Trajectory Tracking Control For Redundant Underwater Vehicles

Recent trends in ocean engineering,such as geological sampling,search,and undersea cable inspection,have led to extensive applications of underwater vehicles.When performing tasks,underwater vehicles are frequently required to track a pre-defined trajectory,and thereby the trajectory tracking control is the fundamental.This thesis analyzes the redundancy of a specific configuration of underactuated underwater vehicles and fully-actuated underwater vehicles in the standard three-dimensional trajectory tracking problem.1.Design and modeling of underwater vehiclesA quadrotor-like underactuated underwater vehicle and a fully-actuated underwater vehicle have been designed and manufactured.Firstly,the hardware and software systems of the two vehicles are introduced,then a general kinematics and dynamics model is investiaged.Control allocation problem is addressed by analyzing the actuation systems of two vehicles.2.Orientation-sensitive trajectory tracking control for redundant underactuated underwater vehiclesThe redundancy of a class of underactuated vehicles represented by quadrotor-like underactuated underwater vehicle for the trajectory tracking problem is analyzed,to introduce the orientation-sensitive problem,where both position and orientation of the vehicles are regulated to the reference.In the kinematic stage,these vehicles are capable of moving sideways by exploiting the non-zero angle of the roll,so that the position tracking can be handled while the steering tracks the referenced yaw angle.As for the kinetic stage,a robust non-smooth controller is implemented to guarantee the exponential convergence of velocity tracking errors.Trajectories are planned for the docking task and bridge pier inspection task.It is shown that the position tracking error can converge to a neighborhood about zero,and the yaw angle tracking error converges to zero exponentially as well.The effectiveness of the proposed method is illustrated by both simulation results and experimental results.3.Optimal and constrained control for redundant fully-actuated underwater vehiclesThe proposed strategy offers vehicles more control flexibility for addressing the position tracking problem under the attitudes and velocities constraints.In kinematics,position tracking is formulated as a convex optimization problem where the velocity constraints are considered as a feasible region,and the attitudes are bounded via a barrier function.Then,a conservative boundary is determined in the context of stability analysis of the dynamic controller with a disturbance observer and an auxiliary system so that kinematic variables lie in the constraints even with the velocity tracking error.Also,it is proven that the position tracking error can converge to a neighborhood about zero,which can be made arbitrarily small.The simulations and experiments verify the efficiency of the proposed strategy.4.Fault diagnosis and fault-tolerant control for redundant fully-actuated underwater vehiclesWith the motion redundancy,this thesis proposed a control framework that can be automatically reconfigured based on the inputs malfunctions in kinetics,with specific application to trajectory tracking control from fully-actuated to underactuated underwater vehicles.A diagnostic component is introduced to generate fault signals that are robust to model uncertainties and unknown time-varying disturbances by analyzing the boundedness of tracking errors.The diagnostic signals are then utilized by the kinematic controller to replan the referenced velocities.The position tracking error can converge to a neighborhood about zero,which can be made arbitrarily small.Simulation and experiments verify the efficiency of the proposed strategy.5.Experimental platform for motion control of underwater vehiclesTo fulfill the demand of testing control algorithms,an underwater vehicle motion control experimental platform has been designed.The visual positioning system of the experimental platform combines the depth measurement information from the underwater vehicle to locate the underwater vehicle by the monocular camera.The accuracy of the positioning system is analyzed via both reprojection method and real robot experiment,which proves that the experimental platform meets the requirements of the motion control experiment.