This paper proposes a cooperative multi-agent deep reinforcement learning (MADRL) algorithm for energy trading among multiple unmanned aerial vehicles (UAVs) in order to perform big-data processing in a distributed manner. In order to realize UAV-based aerial surveillance or mobile cellular services, seamless and robust wireless charging mechanisms are required for delivering energy sources from charging infrastructure (i.e., charging towers) to UAVs for the consistent operations of the UAVs in the sky. For actively and intelligently managing the charging towers, MADRL-based energy management system (EMS) is proposed and designed for energy trading among the energy storage systems those are equipped with charging towers. If the required energy for charging UAVs is not enough, the purchasing energy from utility company is desired which takes high consts. The main purpose of MADRL-based EMS learning is for minimizing purchasing energy from outside utility company for minimizing operational costs. Our data-intensive performance evaluation verifies that our proposed framework achieves desired performance.
This research was supported by the MSIT (Ministry of Science and ICT), Korea, under the ITRC (Information Technology Research Center) support program (IITP-2020-2017-0-01637) supervised by the IITP (Institute for Information & Communications Technology Planning & Evaluation) and also by National Research Foundation of Korea (2019R1A2C4070663, 2019M3E4A1080391). J. Kim and J.-H. Kim are the corresponding authors of this paper.