Abstract:
In the emerging field of intelligent transportation systems (ITS), ''TS for the sea''
refers to the area of maritime traffic. Automated vehicle control systems are a key
technology for ITS. An autonomous surface vessel (ASV) can be defined as a vehicle
controlling its own steering and speed for Navigation, dynamic positioning, motion
stabilization and obstacle detection and avoidance.
The scope of the research is defined by two main objectives viz. developing complete
mathematical model of a surface vessel by analyzing hydrodynamic forces and main
other effects arising when manoeuvring in the ocean, and design online-learning
adaptive controller for path tracking and speed control using real control inputs;
propeller thrust and steering angle. The vessel moves in a hydrodynamic environment
where many uncertainties, non-linear and non-predictive behaviours always appear.
The ocean vehicle is modelled mathematically using first principles and derivations
wherever possible.
In this work, the problem of control with guaranteed sway and yaw stability for
automated surface vessel operation is addressed with special emphasis on speed
control. A control scheme to solve this problem without simplifying the dynamics is
proposed and extensively studied using formative mathematical analyses and
simulations.
The main academic motivation of this research was to study and synthesis the power
of artificial intelligence techniques in controlling of non-linear dynamical systems
with online-learning and adaptive capabilities. A model-based neural network
adaptive controller is developed blending a self adaptive neural network module and a
classical Proportional plus Derivative (PD)-like control to obtain optimum control
performance by complementing each other. The adaptive neural module counteracts
for inherent model discrepancies, strong nonlinearities and coupling effects.
