Planificador basado en tareas para misiones de vehículos aéreos no tripulados

Abstract

The design of software systems capable of planning and reacting to environmental changes for unmanned aerial vehicles (U.A.V.) is an important challenge nowadays. Although the act of planning seems to be closer to a rational being than robotic agents, the implementation of a system with such capabilities suggests a better performance for the missions executed by U.A.V. in dynamical environments, improving the autonomy of the agents. As a result, the design and development of these systems is the main objective of this work. This document presents the design and development of a task-based mission planner for U.A.V.. In the design of such software, reactive planning techniques and automated planning techniques have been used in order not only to provide the abilities defined before, but also to define the general structure of a mission. In addition, concepts and results from Number Theory and Fuzzy Logic have been used in the construction of a formal model for action feasibility from a physical point of view. Finally, the software has been tested in order to check their correct operation .First from a modular point of view, testing the most important components of each module individually; and second,from a global point of view, checking that the operations among all the components of the system are accomplished correctly.