半月刊

ISSN 1000-1026

CN 32-1180/TP

+高级检索 English
基于强化学习的产消用户端对端电能交易决策
作者:
作者单位:

1.智能电网教育部重点实验室(天津大学),天津市 300072;2.天津市智慧能源与信息技术重点实验室(天津大学),天津市 300072;3.国网江苏省电力有限公司镇江供电公司,江苏省镇江市 212000;4.国网天津市电力公司,天津市 300010

摘要:

强化学习是一种促进智能体在与环境交互过程中通过学习策略达成回报最大化的人工智能方法。在不进行优化计算和不充分了解市场机制的情况下,该方法非常适合处理小规模用户电能交易行为。文中首先建立了包含交易主体、交易电价和交易物理约束的产消用户端对端电能交易模型。其次,将电能交易问题等效为一个马尔可夫决策过程并对各学习要素进行建模。然后,基于Q-Learning强化学习算法,对马尔可夫决策过程中储能动作和交易策略选择问题进行均匀离散处理,并进行分析求解。最后,采用含多类型产消用户和消费用户的电能交易案例进行分析,验证强化学习方法在解决小规模产消用户端对端电能交易问题上的合理性和可行性。

关键词:

基金项目:

国家重点研发计划资助项目(2018YFB0905000);国家自然科学基金资助项目(51977141);国家电网公司总部科技项目(SGTJDK00DWJS1800232)。

通信作者:

作者简介:

王丹(1981—),男,通信作者,博士,副教授,主要研究方向:综合能源电力系统分析、智能用电技术、分布式能源系统与储能。E-mail:wangdantjuee@tju.edu.cn
刘博(1996—),男,硕士研究生,主要研究方向:用户侧端对端电能交易。E-mail:boliustu@126.com
贾宏杰(1973—),男,博士,教授,主要研究方向:大电网稳定性分析、电网规划、新能源集成、综合能源系统分析。E-mail:hjjia@tju.edu.cn


Peer-to-Peer Energy Transaction Decision of Prosumers Based on Reinforcement Learning
Author:
Affiliation:

1.Key Laboratory of the Ministry of Education on Smart Power Grid (Tianjin University), Tianjin 300072, China;2.Key Laboratory of Smart Energy & Information Technology of Tianjin Municipality (Tianjin University), Tianjin 300072, China;3.Zhenjiang Power Supply Company of State Grid Jiangsu Electric Power Co., Ltd., Zhenjiang 212000, China;4.State Grid Tianjin Electric Power Company, Tianjin 300010, China

Abstract:

Reinforcement learning is an artificial intelligence method to maximize the payback of intelligent agents through learning strategies in the process of interaction with the environment. Without the optimal calculation and the full knowledge of the market mechanism, this method is very suitable for prosumers dealing with the small-scale energy transaction behavior of users. Firstly, a peer-to-peer energy transaction model including transaction subjects, price and physical constraints is established in this paper. Secondly, the energy transaction problem is equivalent to a Markov decision process and each learning element is modeled. Then, based on the Q-learning reinforcement learning algorithm, the problem of energy storage action and transaction strategy selection in Markov decision process is discretized uniformly, and then analyzed and solved. Finally, the case of energy transaction including multi-type prosumers and consumers is used to verify the rationality and feasibility of the reinforcement learning method in solving the peer-to-peer power transaction problem of small-scale prosumers.

Keywords:

Foundation:
This work is supported by National Key R&D Program of China (No. 2018YFB0905000), National Natural Science Foundation of China (No. 51977141), and State Grid Corporation of China (No. SGTJDK00DWJS1800232).
引用本文
[1]王丹,刘博,贾宏杰,等.基于强化学习的产消用户端对端电能交易决策[J].电力系统自动化,2021,45(3):139-147. DOI:10.7500/AEPS20200515005.
WANG Dan, LIU Bo, JIA Hongjie, et al. Peer-to-Peer Energy Transaction Decision of Prosumers Based on Reinforcement Learning[J]. Automation of Electric Power Systems, 2021, 45(3):139-147. DOI:10.7500/AEPS20200515005.
复制
分享
历史
  • 收稿日期:2020-05-15
  • 最后修改日期:2020-09-10
  • 录用日期:
  • 在线发布日期: 2021-02-03
  • 出版日期:
相关附件