Autonomous Underwater Vessels

Abstract

The control of multi-DOF autonomous underwater vessels (AUVs) exhibits particular difficulties which are due to the complicated nonlinear model of the submersible vessels, the coupling between the systems control inputs and outputs, and the uncertainty about the values of their model’s parameters. Moreover, the AUVs’ dynamic model is subject to external perturbations which are caused by variable sea conditions and sea currents. Consequently, an efficient control scheme for AUVs should not only compensate for the nonlinearities of the associated dynamic model, but should also exhibit robustness to model parameter variations and to external disturbances. To this end, the present chapter provides results on robust control of AUVs, as well as on adaptive control of such submersible vessels. Thus the control problem for autonomous underwater vessels is treated with (i) global linearization methods (ii) approximate linearization methods and (iii) Lyapunov methods. The solution of the control problem requires a more elaborated procedure when the AUVs’ dynamic model is underactuated. which means that the number of actuators included in its propulsion system is less than the number of its degrees of freedom.The methods developed in this chapter treat also the case of underactuated AUVs. Moreover, advanced estimation methods are used to identify in real time the unknown dynamics of the underwater vessels or disturbance forces and torques that affect them. This allows for the implementation of indirect adaptive control schemes for the AUVs. Additionally, for the precise localization of the AUVs and their safe navigation elaborated nonlinear filtering methods are developed. These permit to solve problems of multi-sensor fusion as well as problems of decentralized state estimation with the use of spatially distributed nonlinear filters that track the AUVs motion. In particular the chapter treats the following topics: (a) Global linearization-based control of autonomous underwater vessels, (b) Flatness-based adaptive fuzzy control of autonomous submarines, and (c) Nonlinear optimal control of autonomous submarines.