电工技术学报  2022, Vol. 37 Issue (7): 1789-1799    DOI: 10.19595/j.cnki.1000-6753.tces.210212
电力系统与综合能源 |
基于ResNet-LSTM网络和注意力机制的综合能源系统多元负荷预测
王琛1, 王颖1, 郑涛2,3, 戴则梅2,3, 张凯锋1
1.复杂工程系统测量与控制教育部重点实验室(东南大学) 南京 210096;
2.南瑞集团(国网电力科学研究院)有限公司 南京 211106;
3.国电南瑞科技股份有限公司 南京 211106
Multi-Energy Load Forecasting in Integrated Energy System Based on ResNet-LSTM Network and Attention Mechanism
Wang Chen1, Wang Ying1, Zheng Tao2,3, Dai Zemei2,3, Zhang Kaifeng1
1. Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education Southeast University Nanjing 210096 China;
2. NARI Group (State Grid Electric Power Research Institute) Co. Ltd Nanjing 211106 China;
3. NARI Technology Development Co. Ltd Nanjing 211106 China
全文: PDF (1728 KB)   HTML
输出: BibTeX | EndNote (RIS)      
摘要 综合能源系统中多种负荷之间可能存在复杂的、较强的相互耦合关系。相对于对各类负荷进行单一独立的预测,直接开展多元负荷预测能够进一步挖掘负荷之间的内在联系,提高预测准确度。该文提出一种基于ResNet-LSTM网络和注意力机制的多任务学习模型,用于拟合多能负荷之间的空间耦合关系和时间耦合关系。首先,采用多层ResNet作为多能负荷数据的特征提取单元,挖掘多能之间的空间耦合交互特征;然后,通过双向长短时记忆网络残差结构进一步挖掘多能负荷数据的时序特征;接着,使用注意力机制实现多任务对于共享特征不同程度的关注,体现不同子任务对共享特征的差异化选择,实现多元负荷的联合预测;最后,结合亚利桑那州立大学Campus Metabolism系统的多能负荷数据,与其他预测模型进行对比分析,结果表明所提出的多元负荷预测方法具有更高的预测精度。
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
王琛
王颖
郑涛
戴则梅
张凯锋
关键词 注意力机制残差网络长短时记忆网络多元负荷预测多任务学习    
Abstract:In an integrated energy system, different types of loads, i.e. electrical loads, heat loads, cooling loads, might have complex and strong coupling relationships among them. Compared with forecasting each type of load separately, to forecast multi-energy loads together in a combined multi-task model can further explore the internal connections and therefore improve the accuracy of forecasting. A multi-task learning model based on ResNet-LSTM network and Attention mechanism is proposed to fit the spatial coupling relationship and time coupling relationship between multi-energy loads. Firstly, the multi-layer ResNet is used as the feature extraction unit of the multi-energy load data to mine the spatial coupling interaction characteristics between the multi-energy; secondly, the LSTM residual structure is used to further mine the time series characteristics of the multi-energy loads data; then, the Attention mechanism is used to realize that multiple subtasks have different degrees of attention to shared features, which reflects the differentiated selection of shared features by different subtasks, and realizes joint forecasting of multiple loads. Finally, we applied the proposed method with the data at the Campus Metabolism system of Arizona State University. Compared with other forecasting models, the results show that the proposed method has higher accuracy.
Key wordsAttention mechanism    ResNet    long short-term memory(LSTM)    multi-energy load forecasting    multi-task learning   
收稿日期: 2021-02-19     
PACS: TM715  
基金资助:国家自然科学基金(51907025, 51977033)、国家电网公司总部科技项目(基于大数据的电网趋势预测及操作智能预演技术研究)和东南大学“至善青年学者”支持计划(2242021R41176)资助
通讯作者: 王颖 女,1989年生,博士,讲师,研究方向为电力系统优化、机组组合和经济调度。E-mail:wyseu@seu.edu.cn   
作者简介: 王琛 男,1997年生,硕士研究生,研究方向为综合能源系统多元负荷预测。E-mail:wangchen1073@126.com
引用本文:   
王琛, 王颖, 郑涛, 戴则梅, 张凯锋. 基于ResNet-LSTM网络和注意力机制的综合能源系统多元负荷预测[J]. 电工技术学报, 2022, 37(7): 1789-1799. Wang Chen, Wang Ying, Zheng Tao, Dai Zemei, Zhang Kaifeng. Multi-Energy Load Forecasting in Integrated Energy System Based on ResNet-LSTM Network and Attention Mechanism. Transactions of China Electrotechnical Society, 2022, 37(7): 1789-1799.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.210212          https://dgjsxb.ces-transaction.com/CN/Y2022/V37/I7/1789