EV charging bidding by multi-DQN reinforcement learning in electricity auction market

Yang Zhang,Zhengfeng Zhang,Qingyu Yang,Dou An,Donghe Li,Ce Li
DOI: https://doi.org/10.1016/j.neucom.2019.08.106
IF: 6
2020-01-01
Neurocomputing
Abstract:In this paper, we address the issue of optimal bidding strategy selection for Electric Vehicles (EVs) charging in an auction market. The problem of EV charging has attracted growing attention as EVs become more and more popular. We consider the scenario that EV owners submit their bids for charging to the charging station, and then charging station determines the winning EVs who are admitted to charge and the payments based on an online continuous progressive second price (OCPSP) auction mechanism. In light of this, how to formulate optimal bidding strategy and maximize the economic benefits is crucial for EV owners. To this end, we propose a Multi-Deep-Q-Network (Multi-DQN) reinforcement learning bidding strategy, in which, a value evaluation network and a target network are proposed for each agent to learn the optimal bidding strategy. The extensive experimental results show that our bidding strategy can achieve better economic benefits and help EV owners spend less time on charging compared to the Q-learning based approach and the random approach. (C) 2020 Elsevier B.V. All rights reserved.
What problem does this paper attempt to address?