This paper presents a multi-agent mission planning and task allocation framework designed to coordinate autonomous aerial vehicles engaged in a competition scenario. The development was a part of an inter-university UAV Swarm competition that was supported by BAE Systems. The proposed centralised system was developed with the main objectives of robustness and scalability. The system consists of a general mission planning module which decomposes the overall mission into identified sub-stages to achieve the overall mission goal. In order to enable autonomous defence actions a dynamic task allocation approach is proposed. The dynamic task allocation is using received information of detected enemies and utilises the information for a further combinatorial optimisation problem. In this work, we discuss the structure of the framework and present results obtained in a high-fidelity simulation environment. Moreover, a comparative study of the performance of three different optimization algorithms for the given combinatorial problem, namely Kuhn-Munkres, Jonker-Volgenant and Gale-Shapley, implemented in the system is included. The results demonstrate that the best allocation result performances, in terms of minimal costs, are obtained with utilising, both Kuhn-Munkres or Jonker-Volgenant methods, while the Gale-Shapley algorithms have benefits in terms of time efficiency for cases in which minimal costs are not the highest priority.
A Mission Planning and Task Allocation Framework For Multi-UAV Swarm Coordination
2019-11-01
2212601 byte
Conference paper
Electronic Resource
English
DOAJ | 2022
|DTTA - Distributed, Time-division Multiple Access based Task Allocation Framework for Swarm Robots
BASE | 2017
|Multiple UAV Task Allocation Using Particle Swarm Optimization
British Library Conference Proceedings | 2008
|