摘要： 针对具有多个目标的多辆电动汽车协调充电问题，包括降低用户的充电成本并保证足够的电池电量，同时避免变压器过载，提出一种基于多智能体深度强化学习的分布式算法。该算法首先将上述问题建模为马尔可夫决策过程，然后采用一种通信神经网络模型进行信息交互，实现了全局信息在多个智能体上的分布式计算。最后，通过仿真验证了该算法在住宅充电区域的有效性和可扩展性。

中图分类号： TP18
文献标识码： A
DOI： 10.19358/j.issn.2096-5133.2022.04.013
引用格式： 张子霖. 基于深度强化学习的电动汽车协调充电算法[J].信息技术与网络安全，2022，41(4)：83-89.

A deep RL-based algorithm for coordinated charging of electric vehicles

Abstract： A distributed algorithm based on multi-agent deep reinforcement learning(DRL) is proposed in this paper to solve the problem of coordinated charging of multiple electric vehicles(EVs) with multiple objectives, including reducing the charging cost, ensuring sufficient battery state of charge(SoC) for users, and avoiding the transformer overload. To this end, the above problem is formulated as a Markov Decision Process(MDP), and then a communication neural network(CommNet) model is adopted for information interaction to realize the distributed computation of global information on multiple EVs. Finally, simulations are presented to verify the effectiveness and scalability of the proposed algorithm in a residential charging area.

Key words : electric vehicle；multi-agent system；deep reinforcement learning；distributed charging scheduling