This thesis deals with the mission management of a team of autonomous vehicles. The two main issues of the mission management are considered: the path-planning among obstacles, and the task-assignment.
In the context of path-planning a novel procedure based on a dynamic version of the Constrained Delaunay Triangulation (CDT) is proposed. This procedure takes advantage of the geometrical properties of the triangles in order to obtain obstacle-free paths with a low computational load.
In the context of task-assignment, two novel approaches are proposed. The former is based on a dynamic task clustering in order to obtain a less myopic view of the scenario. The latter is based on a dynamic task ranking, that is, a set of dynamic weights that the vehicles have for each task. The main advantages of the proposed procedures are the low computational cost and a full decentralization, together with a certain degree of optimality.