Method and algorithm for task allocation in a heterogeneous group of UAVs in a clustered field of targets

Abstract

The article presents a method for distributing tasks to agents of a heterogeneous UAV group in a cluster field of tasks, when the number of tasks exceeds the number of agents by 5–20 times. The proposed task distribution method based on a three-stage procedure for distributing agents of different specializations among task clusters, taking into account the agent value function. To evaluate the effectiveness, the method compared with the greedy task distribution algorithm, the collective plan improvement algorithm, and the consensus-based linking algorithm with local rescheduling. 2400 experiments were carried out with different group sizes and randomly generated task maps, the results of which revealed the high efficiency of the proposed method. According to the results of the study, a relationship found between the efficiency of the method depending on the concentration of the number of tasks per agent. With an increase in the specific number of tasks per agent, the task execution time improves and the indicator of the path traveled by agents worsens. With a ratio of 5–10 agents per 100 tasks, the method shows the best results in terms of the parameters of the path traveled by agents and task execution time.

Keywords

• Multi-robotic systems
• Decentralized task allocation
• Swarm robotics
• Group control
• Task allocation
• Labor division