| 引用本文: | 卢昱宏,范培潇,杨军,李蕊.基于近端策略优化算法的含电动汽车孤岛微电网智能频率控制策略[J].电力自动化设备,2025,45(10):135-143. |
| LU Yuhong,FAN Peixiao,YANG Jun,LI Rui.Intelligent frequency control strategy for islanded microgrid with electric vehicles based on proximal policy optimization algorithm[J].Electric Power Automation Equipment,2025,45(10):135-143. |
|
| 本文已被：浏览 1808次 下载 421次 |
码上扫一扫！ |
|
|
| 基于近端策略优化算法的含电动汽车孤岛微电网智能频率控制策略 |
|
卢昱宏1,2,3, 范培潇1,2,4, 杨军1,2, 李蕊1,2
|
|
1.交直流智能配电网湖北省工程中心，湖北 武汉 430072;2.武汉大学 电气与自动化学院，湖北 武汉 430072;3.清华大学深圳国际研究生院 电力系统国家重点实验室深圳研究室，广东 深圳 518055;4.香港理工大学 电力与电子工程学系，香港 999077
|
|
| 摘要: |
| 随着电动汽车数量的快速增长，其作为有限挂网储能设备参与电网调控的潜力备受关注，但用户行为的随机性与时空移动性给车网互动带来了挑战。为此，提出一种基于近端策略优化算法的含电动汽车孤岛微电网智能频率控制策略。构建包含广义聚合电动汽车充电站的微电网负荷频率控制框架；基于闵可夫斯基求和方式，提出两阶段电动汽车充电站的可调控裕度计算方法；通过重要性采样比率剪切、多步经验回放、小批量优化等方式提升算法训练效率，并通过设计状态与动作空间、奖励函数以及选取合适超参数完成频率控制器的构建。仿真结果表明，所设计的控制器在训练时间和控制效果方面显著优于传统频率控制方法，为微电网的稳定运行提供了有力的技术支持。 |
| 关键词: 电动汽车 孤岛微电网 频率控制 近端策略优化算法 闵可夫斯基求和 |
| DOI：10.16081/j.epae.202503021 |
| 分类号: |
| 基金项目:国家自然科学基金资助项目(51977154) |
|
| Intelligent frequency control strategy for islanded microgrid with electric vehicles based on proximal policy optimization algorithm |
|
LU Yuhong1,2,3, FAN Peixiao1,2,4, YANG Jun1,2, LI Rui1,2
|
|
1.Hubei Engineering and Technology Research Center for AC/DC Intelligent Distribution Network, Wuhan 430072, China;2.School of Electrical and Automation, Wuhan University, Wuhan 430072, China;3.National Key Laboratory of Power Systems in Shenzhen, Tsinghua Shenzhen International Graduate School, Shenzhen 518055, China;4.Department of Electrical and Electronic Engineering, The Hong Kong Polytechnic University, Hong Kong 999077, China
|
| Abstract: |
| With the rapid growth in the number of electric vehicles(EVs),their potential as the limited grid-connected energy storage devices to participate in power grid regulation has attracted significant attention. However, the randomness and spatio-temporal mobility of user behavior bring challenges to the interaction between EVs and networks. Therefore, an intelligent frequency control strategy for an islanded microgrid with EVs based on a proximal policy optimization algorithm is proposed. The microgrid load frequency control framework including generalized aggregated EV charging stations is constructed. Based on the Minkowski summation method, a two-stage calculation method for adjustable margin of EV charging stations is proposed. The training efficiency of the algorithm is improved by means of importance sampling ratio clipping, multi-step experience replay, small-batch optimization, etc.,and the frequency controller is constructed by designing the state space, action space and reward function, and selecting suitable hyperparameters. Simulative results show that the designed controller is significantly superior to the traditional frequency control methods in terms of training time and control effect, which provides strong technical support for the stable operation of microgrid. |
| Key words: electric vehicles islanded microgrid frequency control proximal policy optimization algorithm Minkowski summation |
