Online System Identification and Control of an Autonomous Underwater Vehicle

2015 DECLARATION I hereby declare that this thesis is my original work and it has been written by me in its entirety. I have duly acknowledged all the sources of information which have been used in the thesis. This thesis has also not been submitted for any degree in any university previously. Summary More and more autonomous underwater vehicles (AUVs) are designed to be modular, where their payload configuration can be changed frequently depending on the mission requirements. When an AUV is reconfigured with different payloads, its dynamic characteristic is affected. Since the dynamic model underlies the design of its navigation, guidance and control systems, any deviation from the nominal model would potentially degrade the performance or in the worst case, cause critical safety issues. In this thesis, an online method is developed to identify and validate the dynamics of a newly configured AUV. The AUV is programmed to perform a compact set of maneuvers where the vehicle's response is measured under known excitation. The method is composed of two stages. In the training stage, a State Variable Filter and Recursive Least Square (SVF-RLS) estimator is used to estimate the model parameters. In the validation stage, the prediction capability of the identified model is checked using a fresh data set. Compared to the conventional offline identification method, the SVF-RLS estimator is better in terms of prediction accuracy, computational cost, and training duration. We illustrate how the identified model can be used to estimate the turning radius of an AUV at different speeds and to design a gain-scheduled controller. In order to meet the decoupling assumption, the roll angle of an AUV has to be kept small. To tackle this problem, we develop an internal rolling mass mechanism to actively stabilize the roll motion. We rotate a custom-made electronics tray, which has an off-centric center of gravity, to produce the required torque to stabilize the roll motion. The mechanical design of such a mechanism and its dynamic model and control are discussed in detail. The effectiveness of the mechanism in regulating the roll motion is shown in both tank tests and field experiments. As the dynamic model is scheduled according to the vehicle's forward speed, the operating range of the speed need to be known. The minimum speed is not zero, but a certain speed at which the AUV must travel for depth keeping. When the fins lose their effectiveness at …