Advances in six-degree-of-freedom dynamics and control of underwater vehicles

This thesis addresses the topics of (1) high level control of underwater vehicles; (2) identification of a class of finite-dimensional multiple degree of freedom (DOF) coupled dynamical plant models for underwater vehicles; and (3) nonlinear control of underwater vehicles based upon fully coupled dynamical plant models.;The problem of high-level control of underwater vehicles is addressed with the development of a new supervisory control system. The reported supervisory control system is currently deployed on three underwater vehicles. Results with a field-deployed autonomous underwater vehicle (AUV) were given as working examples to demonstrate capabilities of our mission controller.;The second problem---identification of a class of coupled dynamical plant models---is addressed by a review of these models and their experimental validation on the JHUROV underwater vehicle. First, we review the derivation of the general coupled nonlinear 6-DOF finite-dimensional approximate dynamical model for a submerged body. The dynamical model of underwater vehicles is then expressed in three classes of nonlinear dynamical plants: (1) decoupled 1-DOF, (2) coupled 3-DOF, and (3) coupled 6-DOF. Second, we review the parameter estimation methodologies of total least squares (TLS), ordinary least squares (OLS), and their underdetermined variants. Third, we compare the experimental identification of decoupled 1-DOF finite-dimensional plant model parameters for underwater vehicles using OLS and TLS techniques. This analysis corroborates identification results previously reported in [94]. Finally, we compare the experimental identification of the coupled nonlinear 3-DOF and 6-DOF finite-dimensional plant model parameters for an underwater vehicles using OLS and TLS techniques. This is the first reported 3-DOF and 6-DOF coupled plant model identification of a low-speed, fully actuated, neutrally buoyant underwater vehicle and the first reported use of TLS to identify the plant parameters of an underwater vehicle.;The final problem addressed by this thesis is the development of algorithms for model-based trajectory tracking of underwater vehicles employing a coupled dynamical plant model. Two types of coupled finite-dimensional plants are addressed: (1) a 3-DOF plant in the XY Plane and (2) a 6-DOF plant. First, we report two underactuated fixed model-based controllers and one underactuated non-model-based controller designed to enable an underwater vehicle to perform exact position and velocity tracking in the XY plane. This study provides one of the few comparative experimental evaluations underactuated controllers for underwater vehicles for exact position and velocity tracking in the XY plane. This study also provides the first comparative experimental evaluation that includes both the identification of the 3-DOF coupled nonlinear plant of a low-speed, fully actuated, neutrally buoyant underwater vehicle and the experimental evaluation of the underactuated controllers based upon the identified model. Second, this methodology is repeated for a 3-DOF fully-actuated plant. Finally, we experimentally compare two 6-DOF fixed model-based controllers and one non-model-based 6-DOF controller designed to enable an underwater vehicle to perform 6-DOF exact position and velocity tracking of a low-speed, fully actuated, neutrally buoyant underwater vehicles is reported. This is the first comparative experimental evaluation of both the identification 6-DOF coupled nonlinear plant model for a low speed, fully actuated, neutrally buoyant underwater vehicle and the fixed model-based and non-model-based 6-DOF controllers based upon the identified models.