| With the development and widely using of Autonomous Underwater Vehicle (AUV), there aremore highly maneuverability requirements to AUV. In many tasks, AUV could not only run alongpredefined track, but also closely investigate the target under environment disturbance, whichmeans AUV must have the ability to maintain position to target which is also called dynamicpositioning (DP) ability subject to environment disturbance.The dynamic positioning of Autonomous Underwater Vehicle is a complicated process. Firstly,Because of AUV size and payload limitation, there're a little sensors and other measure unitsaboard, which result in less position and disturbance information gathering; Secondly, AutonomousUnderwater Vehicle must resist high frequency disturbance because the little size and mass of AUVresult in highly response to disturbance; Thirdly, AUV dynamic positioning system must controlmuch more degree of freedom.The motivation of this paper is to achieve AUV dynamic positioning by the study of AUVdynamic modeling, environmental disturbance modeling and prediction, sensitivity of motionresponse to hydrodynamic parameters analyses, sliding mode controller implementation issues withfinite bandwidth actuator with position and rate limitation.The main contribution of this dissertation is summarized as following:1. In order to maintain such precise control like dynamic positioning, AUV must be accuratelymodeled, which means AUV equation of motion should be modeled precisely including allenvironmental disturbance. A new generalized method to the modeling of a small autonomousunderwater vehicle subjected to wave and current effect is provided. Using appropriatemodeling formulation, the wave and current effect are directly incorporated into the equation ofmotion of AUV.2. In all environmental disturbances, wave disturbance is a rather complicated random processand cannot be simulated and predicted precisely. While in order for dynamic positioning, thedisturbances must be fed back and forward into controller. A discrete, digital filter based onwave spectrum for forward prediction of random wave disturbances is presented, and it isuseful in simulation studies when the target wave spectrum is known. However, when thetarget wave spectrum is not known which is typically the case of a deployed vehicle, alternatetechnique must be used. Then one technique which using Recursive Least Squares Algorithmbased on auto-regressive model allow vehicle to estimate wave spectrum on-line is presented.3. Outline the method currently used wave buoy to estimate wave directionality from which thetheory to estimate wave directional distribution from 3-dimension fluid velocity information gathered by an AUV is derived, also the method by which to calculate a heading commandbased on the dominant wave energy direction is describes....4. Because of the wave disturbance, the hydrodynamic parameters of AUV equation of motion arenot constant, but the function of wave frequency, so the maneuverability of AUV are effectedby wave disturbance. The sensitivity of AUV motion response to hydrodynamic parameters isproposed, and a numerical study was made of the sensitivity of AUV motion response measurein turning circle and zigzag maneuver to hydrodynamic parameters variations. The result notonly have specific implication for the design of AUV, but also can help to determine howaccurately does one to know the hydrodynamic parameters on order to ensure acceptableperformance.5. Due to the strong nonlinearity, couple and hydrodynamic parameters uncertainty of AUVequation of motion, the sliding mode controller was chosen. The influence of finite bandwidthactuator position and rate limitation to robustness and stability of sliding mode controller wasanalyzed, and the method of prefilter design with dynamic boundary layer is presented. Themain idea is to dynamically adjust the boundary layer thickness to keep the controller operatingin the linear region of a boundary layer saturation element. When the disturbance andhydrodynamic parameters varies, by increasing the boundary layer thickness, the trackingperformance is degraded but the state remain within the boundary layer, stability is maintained,actuator limits are not violated, and integrator wind-up is avoided.6. Using the methods discussed before, the dynamic position controllers including tracking andstation-keeping for a certain AUV are designed, and the final simulation result show thatthrough reasonable AUV dynamic modeling, environmental disturbance predicting andappropriate control strategy selection, the performance of AUV dynamic position is verified... |
PDF Full Text Request