基于数据更新长短期记忆网络的多能源微网集群优化调度方法  

OPTIMAL DISPATCH METHOD OF MULTI-ENERGY MICROGRID CLUSTER BASED ON DATA UPDATING LONG SHORT-TERM MEMORY NETWORK

作　　者：邱实 张琨 程嵩晴 陈哲 Qiu Shi;Zhang Kun;Cheng Songqing;Chen Zhe(Shenyang University of Technology,School of Electrical Engineering,Shenyang 110870,China;Aalborg University,the Faculty of Engineering and Science,Aalborg 9220,Denmark)

机构地区：[1]沈阳工业大学电气工程学院,沈阳110870 [2]奥尔堡大学工程与科学学院,奥尔堡9920

出　　处：《太阳能学报》2025年第4期193-199,共7页Acta Energiae Solaris Sinica

基　　金：国家重点研发计划(2017YFB0902100)。

摘　　要：为实现多能源微网集群海量存储数据的准确分析和利用,提出一种利用数据更新长短期记忆网络的多能源微网集群优化调度方法。首先,研究多能源微网间的能量交互机理,建立多能源微网集群能量交互模型。然后,基于多能源微网间的能量传输,综合考虑负荷需求和新能源出力历史数据对多能源微网集群调度决策的影响,建立基于长短期记忆网络的多能源微网集群优化调度模型。最后,对提出的多能源微网集群优化调度方法进行算例验证,并与传统微网调度方法进行对比,结果表明提出的调度方法可有效提升多能源微网集群调度的经济性和精准性。Considering the uncertainty brought by high proportion of new energy supply and change in multi-source loads to multienergy microgrid,this paper proposed a method of using long short-term memory networks to solve the optimization scheduling problem of multi energy microgrid clusters based on historical operating data of multi-energy microgrids.Firstly,a multi-energy microgrid cluster model was established in this paper.Then,considering the impact of historical data on load demand and new energy output,a data updating long short-term memory network multi-energy microgrid cluster optimization scheduling model was established.Finally,the simulation result shows that the proposed scheduling method can effectively improve the economy and accuracy of multi-energy microgrid cluster scheduling compared to traditional physical scheduling methods.

关 键 词：多能源微网 新能源电源 长短期记忆网络 微网集群 神经网络 电网调度 

分 类 号：TM73[电气工程—电力系统及自动化]