基于深度分布式强化学习的电力系统分散协调调度方法  

Decentralized Coordinated Dispatching Method for Power System Based on Deep Distributed Reinforcement Learning

作　　者：潮铸 段秦尉 钱峰 黄红伟 薛艳军 CHAO Zhu;DUAN Qinwei;QIAN Feng;HUANG Hongwei;XUE Yanjun(Electric Power Dispatching Control Center of Guangdong Power Grid,Guangzhou Guangdong 510080,China;Beijing Qingda KeYue Co.,Ltd.,Beijing 10084,China)

机构地区：[1]广东电网电力调度控制中心,广东广州510080 [2]北京清大科越股份有限公司,北京100084

出　　处：《电子器件》2022年第4期947-953,共7页Chinese Journal of Electron Devices

摘　　要：针对风、光等可再生能源发电不断并入电力系统,配电网由被动逐渐转变为主动,由此需要主电网、主动配电网和微电网统一调度的问题,基于深度分布式强化学习方法,提出了电力系统分散协调的一体化调度方法。首先,基于传统经济调度模型,提出了主电网、主动配电网和微电网分散协调的调度模型;其次,对传统强化学习方法进行改进,提出了深度分布式强化学习协调模型;第三,将分散协调的深度强化学习方法应用至经济调度中,推导得到分散协调的经济调度方法;最后,以实际电网为例进行验证,表明了所提方法的有效性。In order to solve the problem that wind,light and other renewable energy power generation are continuously integrated into the power system,which gradually changes the distribution network from passive to active,which requires unified dispatching of main grid,active distribution network and microgrid,an integrated dispatching method of decentralized coordination of power system is proposed based on deep distributed reinforcement learning method.Firstly,based on the traditional economic dispatch model,the decentralized coordination dispatching model of main grid,active distribution network and microgrid is proposed.Secondly,the traditional reinforcement learning method is improved and the deep distributed reinforcement learning coordination model is proposed.Thirdly,the decentralized coordination deep reinforcement learning method is applied to economic dispatch,and the decentralized coordinated economic dispatch method is derived.Finally,an actual power grid is taken as an example to verify the effectiveness of the proposed method.

关 键 词：分布式 强化学习 分散协调 调度 电力系统 

分 类 号：TM74[电气工程—电力系统及自动化]