Experimental Results of Supply Ship Autopilot Optimisation Using Genetic Algorithms

: The optimisation by Genetic Algorithm (GA) of a non-linear autopilot for ship course changing is studied in this paper. The theory of the optimisation technique is presented from its basis in Darwinian evolution to the development of the form of optimisation algorithm used in the application. This application involves a Sliding Mode (SM) controller for use as a course changing autopilot for a scale model of a supply ship. The GA optimised simulation results are presented and post-optimisation evaluation is carried out through further simulations and trials in a model ship testing tank. From the evaluation of the optimised autopilot the GA and SM control law are shown to be effective design methodologies for control implementation. Index Terms- optimisation, genetic algorithms, sliding mode control, ship steering autopilot, model ship testing tank. 1. Introduction Issues associated with the design of autopilots for marine surface vessels have already been extensively researched [15..