Research On Multimode Motion Control And Path Planning Of Underwater Glider

As a novel underwater observing platform,underwater gliders are characterized with low energy consumption,long range,and low cost.Therefore,underwater glider is playing an increasingly significant role in both civilian and military purpose,including ocean environment monitoring,resources prospecting,and undersea investigation.As a consequence,studies on glider equipment,theories,and applications are of crucial importances.Viewing the existing theoretical studies of the underwater glider,the stability analysis,controller design,and path planning are based on the precise and complete dynamic model.In this study,the dynamic modeling process is deeply studied,the relationships between the steady states and control inputs are analyzed based on the glider dynamics.Two linear controllers are designed based on the linearized dynamic model at equilibrium state.According to Lyapunov Stability Theory,a nonlinear adaptive backstepping controller(ABC)is designed based on the nonlinear glider dynamics.The kinematic model of an underwater glider in three-dimensional(3D)Dubins motion is established,which is developed from the planar Dubins Path and the features in basic glider motion,and a genetic algorithm(GA)-based path planner is designed to search for the optimal motion trajectory.Field trials and simulation experiments are carried on the prototype of the Seagull glider,validated the theoretical studies in this dissertation preliminarily.The main research points are concluded as follows:Based on Newton-Euler equations,the full dynamic model of an underwater glider with coupled roll-pitch attitude adjusting mechanism is derived.The stability of dynamic equilibrium is analyzed,including sawtooth motion and spiral motion.A high-accuracy numerical solution in steady spiral motion is established,which is obtained by solving the equations in the form of velocity after reasonable simplification and approximation of glider dynamics.Through simulating methods,the glider dynamics and the accuracy of the proposed equilibrium solver are preliminary verified.The nonlinear glider equations of motion are linearized at the equilibrium state,the stability is analyzed in time domain;with complete controllability of the system,two linear controllers,PID and LQR,are used to modulate the attitude and velocity of glider.An adaptive backstepping controller based on Lyapunov stability criterion is designed on account of the nonlinear glider dynamics.The proposed method is able to realize multiple-input-multiple-output(MIMO)motion control,so the glider can switch among multimode motion;therefore,path planning and routing optimization can be realized.The planar Dubins problem is exteded into three dimensional space by combining with the features of gliders' motion.The energy consumptions in each subsystem during the glider motion are thoroughly analyzed,and a genetic algorithm-based path planner is used to search for the optimal trajectory.The searching strategy is verified through simulations of single glider,and then applied to the formation optimization problem of multiple gliders.The feasibility and applicability are discussed through simulation cases such as optimization of final heading angle and final destination,and provide theoretical basis for practical applications.Taken theoretical analysis results as reference,the 500-meter-depth prototype ‘Seagull' has been developed.The subsystems of Seagull have been tested individually,simulations and trial tests have been carried to verify the theoretical studies.The trial tests have demonstrated the stability of the prototype;the expected goal has been achieved,and Seagull is preliminarily capable of serving as a sensor platform for ocean sampling.