Modelling and control of lightweight underwater vehicle-manipulator systems

Abstract

This thesis studies the mathematical description and the low-level control structures for underwater robotic systems performing motion and interaction tasks. The main focus is on the study of lightweight underwater-vehicle manipulator systems. A description of the dynamic and hydrodynamic modelling of the underwater vehicle-manipulator system (UVMS) is presented and a study of the coupling effects between the vehicle and manipulator is given. Through simulation results it is shown that the vehicle’s capabilities are degraded by the motion of the manipulator, when it has a considerable mass with respect to the vehicle. Understanding the interaction effects between the two subsystems is beneficial in developing new control architectures that can improve the performance of the system. A control strategy is proposed for reducing the coupling effects between the two subsystems when motion tasks are required. The method is developed based on the mathematical model of the UVMS and the estimated interaction effects. Simulation results show the validity of the proposed control structure even in the presence of uncertainties in the dynamic model. The problem of autonomous interaction with the underwater environment is further addressed. The thesis proposes a parallel position/force control structure for lightweight underwater vehicle-manipulator systems. Two different strategies for integrating this control law on the vehicle-manipulator structure are proposed. The first strategy uses the parallel control law for the manipulator while a different control law, the Proportional Integral Limited control structure, is used for the vehicle. The second strategy treats the underwater vehicle-manipulator system as a single system and the parallel position/force law is used for the overall system. The low level parallel position/force control law is validated through practical experiments using the HDT-MK3-M electric manipulator. The Proportional Integral Limited control structure is tested using a 5 degrees-of-freedom underwater vehicle in a wave-tank facility. Furthermore, an adaptive tuning method based on interaction theory is proposed for adjusting the gains of the controller. The experimental results show that the method is advantageous as it decreases the complexity of the manual tuning otherwise required and reduces the energy consumption. The main objectives of this thesis are to understand and accurately represent the behaviour of an underwater vehiclemanipulator system, to evaluate this system when in contact with the environment and to design informed low-level control structures based on the observations made through the mathematical study of the system. The concepts presented in this thesis are not restricted to only vehicle-manipulator systems but can be applied to different other multibody robotic systems.