浙江电力

2020, v.39;No.288(04) 29-35

[打印本页] [关闭]
本期目录(Current Issue) | 过刊浏览(Archive) | 高级检索(Advanced Search)

基于EEMD-LSTM的区域能源短期负荷预测
Short-term Load Forecasting Based on EEMD-LSTM for Regional Energy

马梦冬,彭道刚,王丹豪
MA Mengdong,PENG Daogang,WANG Danhao

摘要(Abstract):

短期负荷预测既是电网规划的重要组成部分,也是系统可靠、高效运行的前提和基础。采用EEMD(集合经验模态分解)方法将区域能源互联网历史负荷数据分解成若干分量,再对各个分量分别建立模型,运用LSTM(长短期记忆神经网络)设置对应的隐藏层数,对各个分量分别进行预测,最后叠加预测值得出最终预测结果。将EEMD-LSTM算法与LSTM算法、 EMD-LSTM算法以及Elman算法进行比较,结果表明EEMD-LSTM算法在区域能源互联网负荷预测中能够实现较好的预测精度,具有很好的应用前景和推广价值。
Short-term load forecasting is not only an important part of power grid planning but a prerequisite and basis for the reliable and efficient operation of the system. The ensemble empirical mode decomposition method(EEMD) is used to decompose historical load data of the regional energy Internet into several components, and then models for the components built separately. The long-short-term memory(LSTM) is used to set the corresponding hidden layer number to predict each component separately and conclude the final prediction result by superimposing the prediction values. Compared with LSTM algorithm, EMD-LSTM algorithm and Elman algorithm, it is verified that EEMD-LSTM can achieve better prediction accuracy in regional energy Internet load forecasting and has a good application prospect and is worthy of promotion.

关键词(KeyWords): 短期负荷预测;区域能源;集合经验模态分解;长短期记忆神经网络
short-term load forecasting;regional energy;EEMD;LSTM

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金项目(71871160);; 上海市“科技创新行动计划”高新技术领域项目(19511101600)

作者(Author): 马梦冬,彭道刚,王丹豪
MA Mengdong,PENG Daogang,WANG Danhao

DOI: 10.19585/j.zjdl.202004005

参考文献(References):

扩展功能
本文信息
服务与反馈
本文关键词相关文章
本文作者相关文章
中国知网
分享