Optimization of autonomous ship maneuvers applying swarm intelligence

This paper introduces a new possible solution to the problem of manoeuvre planning for marine vehicles. This proposal is based on the use of a new swarm algorithm called Ant Colony Extended, a new kind of ant algorithm derived from Ant Colony Optimisation. This algorithm allows to include the vehicle's model in the planning so the solutions obtained are always feasible. With time enough it also can obtain trajectories very close to the optimal. Cellular Automata was used to define an heuristic function. This approach was tested on different manoeuvre scenarios. For all of them the algorithm was able to obtain good solutions.