Automatic Generation Control in a Distributed Power Grid Based on Multi-step Reinforcement Learning

作     者：Wenmeng Zhao Tuo Zeng Zhihong Liu Lihui Xie Lei Xi Hui Ma 

作者机构：the Electric Power Research InstituteSouthern Power GridGuangzhou 510663China the Heyuan Power Supply BureauGuang-dong Power Grid Co.Ltd.Heyuan 517000China the College of Electrical Engineering&New EnergyChina Three Gorges UniversityYichang 443002China the College of Electrical Engineering&New Energy and Hubei Provincial Key Laboratory for Operation and Control of Cascaded Hydropower StationChina Three Gorges UniversityYichang 443002China IEEE 

出 版 物：《Protection and Control of Modern Power Systems》 (现代电力系统保护与控制（英文）)

年 卷 期：2024年第9卷第4期

页      面：39-50页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 080802[工学-电力系统及其自动化] 0808[工学-电气工程] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported by the National Natural Sci-ence Foundation of China(No.52277108) Guangdong Provincial Department of Science and Technology(No.2022A0505020015) 

主　　题：Automatic generation control Dyna framework distributed power grid multi-agent mod-el-based reinforcement learning 

摘      要：The increasing use of renewable energy in the power system results in strong stochastic disturbances and degrades the control performance of the distributed power *** this paper,a novel multi-agent collaborative reinforcement learning algorithm is proposed with automatic optimization,namely,Dyna-DQL,to quickly achieve an optimal coordination solution for the multi-area distributed power *** proposed Dyna framework is combined with double Q-learning to collect and store the environmental *** can iteratively update the agents through buffer replay and real-time *** the environmental data can be fully used to enhance the learning speed of the *** mitigates the negative impact of heavy stochastic disturbances caused by the integration of renewable energy on the control *** are conducted on two different models to validate the effectiveness of the proposed *** results demonstrate that the proposed Dyna-DQL algorithm exhibits superior stability and robustness compared to other reinforcement learning algorithms.