华东师范大学学报（自然科学版） ›› 2025, Vol. 2025 ›› Issue (6): 19-28.doi: 10.3969/j.issn.1000-5641.2025.06.003

• • 上一篇    下一篇

基于迁移学习与注意力机制混合神经网络的窃电检测

陈李燊1, 蒲鹏2, 钱江海1,3,*()   

  1. 1. 上海电力大学 数理学院, 上海　200090
    2. 华东师范大学 数据科学与工程学院, 上海　200062
    3. 华东师范大学 软硬件协同设计技术与应用教育部工程研究中心, 上海　200062
  • 收稿日期:2024-01-29 出版日期:2025-11-25 发布日期:2025-11-29
  • 通讯作者: 钱江海 E-mail:qianjianghai@shiep.edu.cn
  • 基金资助:
    华东师范大学软硬件协同设计技术与应用教育部工程研究中心开放研究基金 (OP202102)

Electricity theft detection based on transfer learning and attention hybrid neural network

Lishen CHEN1, Peng PU2, Jianghai QIAN1,3,*()   

  1. 1. College of Mathematics and Physics, Shanghai University of Electric Power, Shanghai　200090, China
    2. School of Data Science and Engineering, East China Normal University, Shanghai　200062, China
    3. Engineering Research Center of Software and Hardware Co-design Technology and Application of the Ministry of Education, East China Normal University, Shanghai　200062, China
  • Received:2024-01-29 Online:2025-11-25 Published:2025-11-29
  • Contact: Jianghai QIAN E-mail:qianjianghai@shiep.edu.cn

RichHTML

3

PDF

2

摘要：

在目前窃电检测的实践中, 采用一维用电数据建模的检测性能往往不足, 而使用二维图像训练模型又面临计算成本过高的问题. 为解决这一困境, 提出了一种基于迁移学习与注意力混合神经网络的窃电检测模型. 模型引入迁移学习策略, 在减少ConvNeXt模型训练开销的同时大幅提升了性能. 同时, 模型混合双向长短期记忆网络模型, 通过提取一维负荷时序数据的全局非线性特征补充纠正了ConvNeXt模型的训练. 此外, 模型分别引入了SimAM和多头自注意力机制以提升混合模型的特征表达能力. 在国家电网公开数据集中, 对所提模型进行了实验验证, 结果表明, 相比其他深度学习分类模型, 所提出模型的$ {A_{{{\mathrm{UC}}} }} $、${M_{{\text{AP@100}}}}$、${M_{{\text{AP@200}}}}$和${F_1}$分数都取得了有效的提升, 相比t-LeNet算法, ${F_1}$提升了9.1%.

关键词: 迁移学习, ConvNeXt, 双向长短期记忆网络, 注意力机制, 混合神经网络

Abstract:

In this study, several issues with current electricity theft detection methods are addressed, notably the reliance on one-dimensional electricity load time series data to develop a singular model. These approaches are often plagued by low detection accuracy, and they require extensive training parameters and a significant number of training samples when computer vision models are directly applied to two-dimensional images of electricity load time series. To overcome these challenges, a novel electricity theft detection method that utilizes a hybrid neural network, combining transfer learning and attention mechanisms, is proposed. The training demands of the ConvNeXt model are reduced via the integration of transfer learning, significantly enhancing its performance. Additionally, a bi-directional long short-term memory (BiLSTM) model is integrated to support the training of the refined ConvNeXt model by extracting global nonlinear features from one-dimensional load time-series data. Furthermore, SimAM and multi-headed self-attention (MHSA) mechanisms are incorporated to improve the feature representation capability of the hybrid model. The experimental verification of the proposed method in the China State Grid public dataset shows that $A_{\mathrm{UC}} $, $M_{{\text{AP@}}100} $, $ M_{{\text{AP@}}200}$, and $F_1 $ metrics of the proposed model can be effectively enhanced when compared to those of other deep learning classification models. For example, $F_1 $ is improved by 9.1% compared to that obtained via t-LeNet algorithm.

Key words: transfer learning, ConvNeXt, bi-directional long short-term memory, attention mechanism, hybrid neural network

中图分类号: