Adaptive ship course-keeping system

The article presents an original adaptive ship course-keeping system. The control algorithm is based on a knowledge base. Its integral part - a computer ship movement dynamical model - uses a set of signals obtained from the object’s input and output. In this way we avoid problems arising while designing classical control algorithms for a complex, non-linear ship model. Approximators utilizing base functions of the model are used for model adaptation. The presented methodology is general as it can also be applied in other ship control tasks or to other dynamic objects. The proposed intelligent course-keeping system has been verified by simulation.