Adaptive Evolutionary Search Algorithm with Obstacle Avoidance for Multiple UAVs

Authors:	Carruthers Barry, University of Glasgow, United Kingdom McGookin Euan W., University of Glasgow, United Kingdom Murray-Smith David J., University of Glasgow, United Kingdom
Topic:	7.5 Intelligent Autonomous Vehicles
Session:	Multiple Vehicles I
Keywords:	Autonomous vehicles, obstacle avoidance, search algorithms

Abstract

Unmanned air vehicles (UAVs) can be used for surveillance and searching, usually with a specified goal i.e. a given area to search or, person or object to find. In these cases, multiple UAVs can be more effective for searching than a single vehicle due to the greater area of coverage that they offer. This paper outlines a search methodology, based on evolutionary algorithms (EAs), which guides a swarm of UAVs throughout a search mission. The autonomously navigating vehicles also avoid unforeseen obstacles via reactive obstacle sensing.