| 基于改进深度强化学习的智能微电网群控制优化方法 |
| |
| 引用本文： | 毛亚哲,何柏娜,王德顺,姜仁卓,周宇洋,张靖茹,贺兴民,董彦辰.基于改进深度强化学习的智能微电网群控制优化方法[J].陕西电力,2021,0(3):19-25,58. |
| |
| 作者姓名： | 毛亚哲 何柏娜 王德顺 姜仁卓 周宇洋 张靖茹 贺兴民 董彦辰 |
| |
| 作者单位： | （1. 山东理工大学 电气与电子工程学院，山东 淄博 255000；2. 中国电力科学研究院（南京），江苏 南京 210003） |
| |
| 摘 要： | 针对微电网群控制的经济效益、负荷波动以及碳排放问题，提出一种基于改进深度强化学习的智能微电网群运行优化方法。首先，计及分布式电源、电动汽车及负荷特性，提出微电网的系统模型。然后，针对微电网群的运行特点，提出4个系统优化目标和5个约束条件，并且引入分时电价机制调控负荷运行。最后，利用改进深度强化学习算法对微电网群进行优化，合理调控多种能源协同出力，调整负荷状态，实现电网经济运行。仿真结果表明了所提方法的有效性，与其他方法相比，其收益较高且碳排放量较小，可实现系统的经济环保运行。
|
| 关 键 词： | 智能微电网群 改进深度强化学习 电动汽车 能量控制优化 经济环保 |
Optimization Method for Smart Multi-microgrid Control Based on Improved Deep Reinforcement Learning |
| |
| MAO Yazhe,HE Baina,WANG Deshun,JIANG Renzhuo,ZHOU Yuyang,ZHANG Jingru,HE Xingmin,DONG Yanchen.Optimization Method for Smart Multi-microgrid Control Based on Improved Deep Reinforcement Learning[J].Shanxi Electric Power,2021,0(3):19-25,58. |
| |
| Authors: | MAO Yazhe HE Baina WANG Deshun JIANG Renzhuo ZHOU Yuyang ZHANG Jingru HE Xingmin DONG Yanchen |
| |
| Affiliation: | （1. College of Electric and Electronic Engineering， Shandong University of Technology， Zibo 255000，China；2. China Electric Power Research Institute （Nanjing）， Nanjing 210003， China） |
| |
| Abstract: | Aiming at the economic benefits， load fluctuations and carbon emissions of multi-microgrid control，the paper proposes a optimization method for smart multi-microgrid operation based on improved deep reinforcement learning. Under the premise of the comprehensive analysis of distributed generation，electric vehicles and load characteristics， the system model of the microgrid is proposed. According to the operating characteristics of the multi-microgrid system， four system optimization objectives and five constraints are proposed. Furthermore， time-of-use power price mechanism is introduced to regulate the load operation. Based on the above analysis， the improved deep reinforcement learning algorithm is used to optimize the multi-microgrid system， ensuring the economic operation of the power grid by rationally regulating the coordinated output of multiple energy sources and the load conditions. The simulation results show that， compared with other methods， the proposed method has higher profits and lower carbon emissions， which can realize the economic and environmental operation of the system. |
| |
| Keywords: | smart multi-microgrid improved deep reinforcement learning electric vehicle energy control optimization economic and environmental operation |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《陕西电力》浏览原始摘要信息 |
|
点击此处可从《陕西电力》下载全文 |
