引用本文: | 薛健侗,马宏忠,杨洪苏,等.基于格拉姆角场与迁移学习-AlexNet的变压器绕组松动故障诊断方法[J].电力系统保护与控制,2023,51(24):154-163.[点击复制] |
XUE Jiantong,MA Hongzhong,YANG Hongsu,et al.A fault diagnosis method for transformer winding looseness based on Gramianangular field and transfer learning-AlexNet[J].Power System Protection and Control,2023,51(24):154-163[点击复制] |
|
摘要: |
绕组松动故障是变压器最主要的机械故障之一,尚缺乏有效的智能化诊断方法。为此提出基于格拉姆角场与迁移学习-AlexNet的变压器绕组松动故障诊断方法。变压器稳态运行时的振动信号存在周期性的特点,导致其构建足量具有时间相关性的图像集十分困难,提出了一种样本构建方法用于生成变压器振动信号的格拉姆角场图像集。将生成的图像集送入AlexNet进行迁移学习,获得微调后的神经网络模型。实验结果表明:利用该样本构建方法生成的图像集作为训练集和验证集,建立的卷积神经网络模型训练准确率与验证准确率均达到99%以上;利用变压器周期性振动信号生成的图像集作为测试集,测试准确率达到99%以上,实现了变压器绕组松动故障的准确诊断,并为周期性信号运用具有时间相关性的图像变换方法构建足量样本集提供了一种新思路。 |
关键词: 变压器 绕组松动 振动信号 格拉姆角场 AlexNet 迁移学习 样本构建 故障诊断 |
DOI:10.19783/j.cnki.pspc.230659 |
投稿时间:2023-06-01修订日期:2023-07-27 |
基金项目:国家自然科学基金项目资助(51577050);国网江苏省电力有限公司重点科技项目资助(J2021053) |
|
A fault diagnosis method for transformer winding looseness based on Gramianangular field and transfer learning-AlexNet |
XUE Jiantong,MA Hongzhong,YANG Hongsu,NI Yiming,WAN Keli,ZE Hengpeng |
(College of Energy and Electrical Engineering, Hohai University, Nanjing 211100, China) |
Abstract: |
The winding looseness fault is one of the main mechanical faults in transformers, and there is still a lack of effective intelligent diagnosis methods. Therefore, a diagnosis method for such a fault based on Gramian angular field and transfer learning-AlexNet is proposed. The periodic characteristics of the vibration signals during steady-state operation of transformers make it difficult to construct a sufficient set of images with time correlation. Therefore, a sample construction method is proposed to generate the Gramian angular field image set of transformer vibration signals. The generated image set is sent to AlexNet for transfer learning to obtain the fine-tuned neural network model. The image set generated by the sample construction method is used as the training and validation set, and the experimental results are that training and validation accuracy of the convolutional neural network model established are both above 99%. The image set generated by the periodic vibration signal of the transformer is used as the test set, with a testing accuracy of over 99%, achieving accurate diagnosis of transformer winding looseness faults. It also provides a way for constructing a sufficient sample set using time-related image transformation methods for periodic signals. |
Key words: transformer winding looseness vibration signals Gramian angular field AlexNet transfer learning sample construction fault diagnosis |