3D Real-Time Energy Efficient Path Planning for a Fleet of Fixed-Wing UAVs

Path planning of unmanned aerial vehicle (UAV) targets at finding an optimal and collision free path in a cluttered environment while taking into account the geometric, physical and temporal constraints. The goal is to reach fully autonomous UAVs robust enough to work in every possible situation and under every possible source of uncertainty. The scientific community is increasingly focusing on this topic and a lot of works has been done in the last decades. It is important to try to go ahead in the research to understand what is the state of the art in order to locate which are the steps to be done to move forward. This Thesis work starts showing the state of the art and detecting the challenges still unsolved. After that, a novel node-based algorithm based on the A* Search algorithm has been developed. The algorithm called EEA* has been developed in order to deal with 3D environments, being robust, energy efficient and working in real-time. Along with the EEA*, a local path planner has been developed in order to cope with unknown dynamic threaten in the environment. The UAV have been tested in different mountain-wise 3D environments, with different numbers of unknown dynamic obstacles. Then, the algorithm has been employed in a multi agent environment, dealing with three UAVs working altogether, making them to perform their path safely. Lastly, the energy efficiency of the path planning algorithm has been tested and compared with the A* algorithm.