Sutulo, S. and Guedes Soares, C. (2018) , “Comparative testing of an identification method based on the genetic and ABC algorithms”, Proceedings of The International Marine Simulator Forum (MARSIM 2018), 12-16 August, Halifax, Canada

A method for offline identification of ship manoeuvring mathematical models developed earlier by the authors was modified and tested on artificial training records. The method allows application of several different metrics but as the superiority of the Hausdorff metric had been demonstrated earlier, only this metric is used in this study. In addition to the classic genetic algorithm used for the internal global optimization, the recently proposed ABC (Artificial Bee Colony) algorithm was suggested as an option. The testing was performed on clean training records and those heavily polluted with the Gauss distributed white noise. In all cases, perfect results were obtained from clean records while no one method performed well when the noise was added. At the same time, contrary to optimistic conclusions that can be found in the literature, the ABC algorithm not only did not show definite advantages but even performed slightly worse than the classic genetic algorithm on the problem considered.

You can download this paper from:

If you did not manage to obtain a copy of this paper: Request a copy of this article

For information about all CENTEC publications you can download: Download the Complete List of CENTEC Publications