Optimisation of autonomous ship manoeuvres applying Ant Colony Optimisation metaheuristic

摘要

This paper introduces the use of a swarm algorithm, derived from Ant Colony Optimisation, to solve path planning problems for autonomous vehicles. The purpose is to obtain optimal trajectories for manoeuvres of Autonomous Surface Vessels. The algorithm works with a model of the vehicle, and the solutions achieved are always feasible. With enough time, it can also obtain trajectories very close to the optimal. Provided the appropriate modifications the algorithm can be applied to solve other combinatorial optimisations problems with a non restricted number of feasible solutions. The methodology is tested through simulations on open sea manoeuvres and scenarios with the presence of obstacles.