Browsing by Author "Skaltsis, George Marios"

Now showing 1 - 3 of 3
  • Thumbnail Image
    ItemOpen Access
    A review of task allocation methods for UAVs
    (Springer, 2023-11-21) Skaltsis, George Marios; Shin, Hyosang; Tsourdos, Antonios
    Unmanned aerial vehicles, can offer solutions to a lot of problems, making it crucial to research more and improve the task allocation methods used. In this survey, the main approaches used for task allocation in applications involving UAVs are presented as well as the most common applications of UAVs that require the application of task allocation methods. They are followed by the categories of the task allocation algorithms used, with the main focus being on more recent works. Our analysis of these methods focuses primarily on their complexity, optimality, and scalability. Additionally, the communication schemes commonly utilized are presented, as well as the impact of uncertainty on task allocation of UAVs. Finally, these methods are compared based on the aforementioned criteria, suggesting the most promising approaches.
  • Thumbnail Image
    ItemOpen Access
    Self play with parameter sharing in n-player mixed competitive-cooperative games
    (AIAA, 2021-12-29) Skaltsis, George Marios; Shin, Hyosang; Tsourdos, Antonios
    We introduce some parameter sharing multi-agent reinforcement learning schemes, combined with self-play for n-players mixed competitive-cooperative games. Except for the pure self-play scheme, an another one using the best policy, outperform the pure parameter sharing baseline, leading to better exploration of the state space and protecting from the performance deterioration observed with the baseline.
  • Thumbnail Image
    ItemOpen Access
    A survey of task allocation techniques in MAS
    (IEEE, 2021-07-19) Skaltsis, George Marios; Shin, Hyosang; Tsourdos, Antonios
    Multi-agent systems and especially unmanned vehicles, are a crucial part of the solution to a lot of real world problems, making essential the improvement of task allocation techniques. In this review, we present the main techniques used for task allocation algorithms, categorising them based on the techniques used, focusing mainly on recent works. We also analyse these methods, focusing mainly on their complexity, optimality and scalability. We also refer to common communication schemes used in task allocation methods, as well as to the role of uncertainty in task allocation. Finally, we compare them based on the above criteria, trying to find gaps in the literature and to propose the most promising ones.