• HOME
  • About Journal
    • Historical evolution
    • Journal Honors
  • Editorial Board
    • Members of Committee
    • Director of the Committee
    • President and Editor in chief
  • Submission Guide
    • Instructions for Authors
    • Manuscript Processing Flow
    • Model Text
    • Procedures for Submission
  • Academic Influence
  • Open Access
  • Ethics&Policies
    • Publication Ethics Statement
    • Peer Review Process
    • Academic Misconduct Identification and Treatment
    • Advertising and Marketing
    • Correction and Retraction
    • Conflict of Interest
    • Authorship & Copyright
  • Contact Us
  • Chinese
Site search        
文章摘要
基于生成对抗网络和联邦学习的非侵入式负荷分解方法
Non-intrusive load disaggregation method based on GAN and federated learning
Received:February 26, 2020  Revised:February 26, 2020
DOI:DOI: 10.19753/j.issn1001-1390.2022.06.006
中文关键词: 非侵入式负荷分解  生成对抗网络  联邦学习  云边协同  智能电能表
英文关键词: non-intrusive load disaggregation, GAN, federated learning, cloud-side collaboration, smart meters
基金项目:国家电网公司科技项目
Author NameAffiliationE-mail
Li Xiaolu School of Electrical Engineering,Shanghai University of Electric Power,Yangpu District joabuy@126.com 
Wu Dingjie* School of Electrical Engineering,Shanghai University of Electric Power,Yangpu District okwudingjie@163.com 
Lu Yiming Energy Internet Research Institute jingkoubuyu@163.com 
Hits: 1771
Download times: 373
中文摘要:
      非侵入式负荷分解是对终端用户用电需求的重要感知手段,传统负荷分解方法存在电器识别和功率分解准确度低等问题。为此提出一种基于生成对抗网络的负荷分解模型,生成网络通过构建卷积自编码器对总功率信号去噪,生成指定电器的功率分解序列,而判别网络被用来辨别生成序列的真伪,两者相互对抗,得到更为真实的分解序列。针对集中式模型训练方法的不足,采用深度可分离卷积代替传统卷积来实现模型轻量化,使之能应用于智能电表等终端设备,并提出一种基于联邦学习的网络模型实施方案,以云边协同的方式对模型进行训练,降低了通信传输压力,保护用户隐私和数据安全。基于公开数据集验证了方法的有效性。
英文摘要:
      Non-intrusive load disaggregation is an important means of sensing the power demand of terminal users. Traditional methods have low accuracy of appliance identification and power disaggregation. A load disaggregation model based on GAN (generative adversarial network) is proposed in this paper. The generative network deconstructs the total power signals by constructing a convolutional self-encoder to generate the power sequence of the appliance, and the discrimination network is used to identify the authenticity of the generated sequence. The two oppose each other to get a more realistic sequence. In view of the shortage of centralized model training method, this paper adopts deep separable convolution instead of traditional convolution, so that it can be applied to smart meters and other terminal equipments, and proposes a network model implementation solution based on federated learning. Training the model in a cloud-side collaborative manner reduces communication transmission pressure and protects user privacy and data security. The validity of the proposed method is verified based on public data sets.
View Full Text   View/Add Comment  Download reader
Close
  • Home
  • About Journal
    • Historical evolution
    • Journal Honors
  • Editorial Board
    • Members of Committee
    • Director of the Committee
    • President and Editor in chief
  • Submission Guide
    • Instructions for Authors
    • Manuscript Processing Flow
    • Model Text
    • Procedures for Submission
  • Academic Influence
  • Open Access
  • Ethics&Policies
    • Publication Ethics Statement
    • Peer Review Process
    • Academic Misconduct Identification and Treatment
    • Advertising and Marketing
    • Correction and Retraction
    • Conflict of Interest
    • Authorship & Copyright
  • Contact Us
  • 中文页面
Address: No.2000, Chuangxin Road, Songbei District, Harbin, China    Zip code: 150028
E-mail: dcyb@vip.163.com    Telephone: 0451-86611021
© 2012 Electrical Measurement & Instrumentation
黑ICP备11006624号-1
Support:Beijing Qinyun Technology Development Co., Ltd