A Track Keeping Neural Network Controller for Ship Guidance

Abstract Initial investigations by the authors applied neural control theory to a mathematical model in simulation with the objective of maintaining course. Research presented in this paper concentrates on the track keeping problem and a case for applying neural techniques is argued. In order to test these methods in the real environment a scale model containing a data collection and integrated navigation system has been constructed. A novel track keeping control system, its implementation, and verification by trials are presented.