Adaptive Speed Controller for the SeaFox Autonomous Surface Vessel.

摘要

The thesis addressed the control system development for a high-speed surface vessel. In particular, the work utilized modern adaptive control techniques to design a speed following controller for the SeaFox ASV; the vehicle features three distinct speed, regimes including the displacement, rapid transition and planing regimes. The study started with the collection of experimental data required to characterize the operating modes and the inherent nonlinear phenomena of the high-speed ASV. Then, it proceeded to system identification study with an objective to develop a mathematical model of the vehicle thus aiming to represent the ASV's speed dynamics at various regimes and to facilitate control system development. After completing the model development, three speed following controllers were designed: A classical Proportional-Integral-Derivative (PID), a nonlinear Model Reference Adaptive (MRAC) and a L1 Adaptive Controller. The motivation behind the choice of three different controllers is two-fold. First, comparison of the linear and nonlinear control approaches is desired to better illustrate the achievable control architecture limitations. Second, comparing two types of nonlinear adaptive control architectures allowed the selection of the best control algorithm for operating the ASV speed in the presence of highly non-linear dynamics and significant disturbances acting on it. Furthermore, each controller is integrated with the SeaFox mathematical model and implemented with and without realistic operational disturbances. This provided a basis for objective comparison among the controllers and gave a means to demonstrate their relative robustness and performance characteristics. Finally, the MRAC and the PID controller were implemented onboard the actual SeaFox ASV and tested in numerous sea-trials under natural conditions to once again demonstrate the advantages and limitations of the chosen control architectures.