综合智慧能源 ›› 2022, Vol. 44 ›› Issue (9): 33-39.doi: 10.3969/j.issn.2097-0706.2022.09.005

• 综合能源系统 • 上一篇    下一篇

基于EEMD-BiLSTM的可调节负荷预测方法

李彬1(), 胡纯瑾1,*(), 王婧2()   

  1. 1.华北电力大学 电气与电子工程学院,北京 102206
    2.国网综合能源服务集团有限公司,北京 100052
  • 收稿日期:2022-02-25 修回日期:2022-05-02 出版日期:2022-09-25 发布日期:2022-09-26
  • 通讯作者: 胡纯瑾
  • 作者简介:李彬(1983),男,副教授,工学博士,从事电力通信、电气信息技术等方面的研究, direfish@163.com;
    王婧(1984),女,高级工程师,工学硕士,从事综合能源系统、虚拟电厂方面的研究, pingjingbei@126.com
  • 基金资助:
    国网综合能源服务集团有限公司科技项目(52789921N00D)

Prediction method for adjustable load based on EEMD-BiLSTM

LI Bin1(), HU Chunjin1,*(), WANG Jing2()   

  1. 1. School of Electrical and Electronic Engineering,North China Electric Power University,Beijing 102206,China
    2. State Grid Integrated Energy Service Group Company Limited,Beijing 100052,China
  • Received:2022-02-25 Revised:2022-05-02 Online:2022-09-25 Published:2022-09-26
  • Contact: HU Chunjin

摘要:

“双碳”目标下,可调节负荷成为电网新兴调节资源。为解决经验模态分解(EMD)的模态混叠现象,同时获取负荷序列良好的时间感知能力,提出了一种集合经验模态分解(EEMD)和双向长短期记忆(BiLSTM)组合的可调节负荷预测方法EEMD-BiLSTM。首先分析了EEMD和BiLSTM的原理,通过将预处理的可调节负荷序列通过EEMD算法进行分解,然后将分解后的分量数据和原始数据分别进行预测建模及重构。试验结果表明EEMD-BiLSTM能够有效表达可调节负荷的时序关系,预测精度高。

关键词: “双碳”目标, 集成经验模态分解, 双向长短记忆网络, 可调节负荷, 负荷预测, 需求响应

Abstract:

To achieve the goals of carbon peaking and carbon neutrality,adjustable load has become emerging regulation resources of the power grid. In order to solve the mode aliasing phenomenon in EMD and obtain good time perception of load series,an adjustable load prediction method that combines EEMD with BiLSTM is proposed. The working principles of EEMD and BiLSTM were expounded. The preprocessed adjustable load series was decomposed by EEMD algorithm, and then the decomposed component data and original data were used for load prediction modeling and reconstruction, respectively. Experimental results show that EEMD-BiLSTM can effectively express the relation between time sequence and adjustable load with a high accuracy.

Key words: dual carbon target, ensemble empirical mode decomposition(EEMD), bidirectional long short term memory(BiLSTM), adjustable load, load forecasting, demand response

中图分类号: