基于深度强化学习的变电站倒闸操作序列生成方法  

Generation Method of Switching Operation Sequence in Substation Based on Deep Reinforcement Learning

作　　者：陈铁[1,2] 曹颖 蔡东阁[1,2] 何思敏 CHEN Tie;CAO Ying;CAI Dongge;HE Simin(College of Electrical Engineering and New Energy,China Three Gorges University,Hubei Yichang 443002,China;Hubei Provincial Key Laboratory for Operation and Control of Cascaded Hydropower Station,China Three Gorges University,Hubei Yichang 443002,China)

机构地区：[1]三峡大学电气与新能源学院,湖北宜昌443002 [2]三峡大学梯级水电站运行与控制湖北省重点实验室,湖北宜昌443002

出　　处：《电工材料》2023年第1期78-84,共7页Electrical Engineering Materials

摘　　要：针对当前智能电网操作票系统通用性差、智能化程度不足等问题,提出一种基于深度强化学习的变电站倒闸操作序列生成方法。首先,采用知识图谱技术建立变电站知识图谱模型,利用知识图谱路径搜索确定操作空间,结合变电站运行规则对任务设备运行状态进行推理更新;然后,构建求解倒闸操作序列的强化学习模型;最后,应用DDQN深度强化学习算法求解倒闸操作序列。测试结果表明,该方法适用于不同的操作任务,能够根据操作任务自动生成符合倒闸操作逻辑的操作序列,无需建立复杂的规则库,通用性较强。In view of the problems such as poor universality and insufficient intelligence of current smart grid operation ticket system, a substation switching operation sequence generation method based on deep reinforcement learning is presented. Firstly, the knowledge map model of substation is established by using knowledge map technology, the operation space is determined by knowledge map path search, and the operation status of task equipment is inferred and updated with substation operation rules. Then, an intensive learning model is built to solve the sequence of the switching operation. Finally, the DDQN deep reinforcement learning algorithm is applied to solve the sequence of the switching operation. The test results show that this method is suitable for different operation tasks, can automatically generate operation sequence according to the operation tasks, does not need to build a complex rule base, and has strong versatility.

关 键 词：深度强化学习 倒闸操作序列 操作票 知识图谱 

分 类 号：TM734[电气工程—电力系统及自动化]