基于CEQ(λ)强化学习算法的微电网智能发电控制  

Smart generation control for micro-grids based on correlated equilibrium Q(λ) learning algorithm

作　　者：唐捷 张泽宇[2] 程乐峰 张孝顺 余涛[2] 

机构地区：[1]广东电网有限责任公司韶关供电局,广东韶关512026 [2]华南理工大学电力学院,广州510640

出　　处：《电测与仪表》2017年第1期39-45,共7页Electrical Measurement & Instrumentation

基　　金：国家自然科学基金资助项目(51177051);中国南方电网科技项目资助(02DKHK1200059)

摘　　要：针对微电网孤岛运行模式下新能源发电强随机性导致的系统频率波动,提出了基于多智能体相关均衡强化学习(Correlated Equilibrium Q(λ),CEQ(λ))的微电网智能发电控制方法。在所搭建含有光伏发电、风力发电、小水电、微型燃气轮机和飞轮储能的微电网负荷频率控制(Load frequency Control,LFC)模型基础上,以频率偏差作为状态输入,提出了一种微电网孤岛运行模式下的CEQ(λ)智能发电控制器。仿真结果显示,与PI控制、单智能体R(λ)控制相比,CEQ(λ)控制器具有更好的在线学习能力,能显著增强孤岛微电网的鲁棒性和适应性,有效提高了频率的考核合格率。In view of addressing frequency fluctuation problem of the islanded micro-grid caused by strong randomness of renewable energy, this paper proposes a smart generation control method of micro-grid based on multi-agent correla- ted equilibrium Q （λ）. After establishing the micro-grid load frequency control （LFC） model containing photovohaic generation, wind generation, small hydro unit, micro-gas turbine and flywheel energy storage, which takes frequency deviation as the input, a smart generation controller based on correlated equilibrium Q （λ） has been proposed. Simu- lation analysis, compared with PI controller, R （λ） learning controller, shows that correlated equilibrium controller has better online learning ability, and enhances robustness and adaptability of islanded micro-grid remarkably and im- proves the pass rate of frequency assessment effectively.

关 键 词：多智能体 相关均衡 强化学习 微电网 智能发电控制 

分 类 号：TM76[电气工程—电力系统及自动化]