State-space modeling and optimal control of ship motions in a seastate

Abstract

In this thesis, a new state-space model and motion control algorithm are developed from first principles for the improvement of the seakeeping performance of high-speed vessels equipped with lifting appendages that are actively controlled in regular and random waves. A ship at sea can experience all the translational and rotational modes of motion that are undesirable, yet unavoidable. These motions have been of great concern to the navies and other organizations engaged in shipping for decades and need to be dealt with through the use of a control system. In this work, a new general purpose state-space control-oriented time domain model for the ship motions is introduced. A discrete auto-regressive state-space model is developed using the state-of-the-art linear seakeeping simulation method SWAN. Novel features of this state-space model are its ability to capture all free-surface memory effects present in the seakeeping problem, its coupling with the theoretical framework of Linear Quadratic (LQ) controllers and its efficient implementation.

(cont.) The development from first principles of a reliable ship motion control simulation method based on SWAN and its coupling with LQ controllers used to actively regulate the angle of attack of lifting appendages, circumvents the need to perform sea trials or model experiments that are harder, time-consuming and expensive to carry out. The performance of the method is illustrated for a catamaran vessel fitted with bow and stern hydrofoils. Simulations of the vessel motions were performed with and without the effect of the controller in regular and random waves. It is concluded that the combination of the proposed state-space model with the LQ controller was very effective in reducing the undesired motions of the vessel in waves over a wide range of wave frequencies and ship speeds.