The paper presents a description of the evaluation phase of the Evolutionary Sets of Safe Ship Trajectories method. In general, the Evolutionary Sets of Safe Ship Trajectories method combines some of the assumptions of game theory with evolutionary programming and finds an optimal set of cooperating trajectories of all ships involved in an encounter situation. While developing a new version of this method, the authors decided to use real maps instead of a simplified polygon modelling and also to focus on better handling of COLREGS. The upgrade to the method enforced redesigning the evaluation phase of the evolutionary process. The new evaluation is thoroughly described and it is shown how evaluation affects final solutions returned by the method.
2
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
This paper presents the idea of using machine learning techniques to simulate and demonstrate learning behavior in ship maneuvering. Simulated helmsman is treated as an individual in population, which through environmental sensing learns itself to navigate through restricted waters selecting an optimum trajectory. Learning phase of the task is to observe current situation and choose one of the available actions. The individual improves his fitness function with reaching destination and decreases its value for hitting an obstacle. Neuroevolutionary approach is used to solve this task. Speciation of population is proposed as a method to secure innovative solutions.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.