| 引用本文: | 赵莹莹,何怡刚,邢致恺,杜博伦.基于信息融合与深度残差收缩网络的DAB变换器开路故障诊断方法[J].电力自动化设备,2023,43(2): |
| ZHAO Yingying,HE Yigang,XING Zhikai,DU Bolun.Open-circuit fault diagnosis method of DAB converter based on information fusion and DRSN[J].Electric Power Automation Equipment,2023,43(2): |
|
| 摘要: |
| 针对双有源桥(DAB)变换器开路故障诊断存在的多信号诊断和诊断阈值设置问题，提出了一种基于信息融合和深度残差收缩网络(DRSN)的DAB变换器开路故障诊断方法。首先，将DAB变换器的开路故障诊断信号减少至3个，减少了信号传感器的数量；其次，采用递归图法和脉冲耦合神经网络将3个诊断信号的时间序列转化为图像进行信息融合，生成的融合图像可以反映不同故障状态下的故障特征且便于深度学习网络进行分类；最后，将融合图像输入构建的DRSN进行故障诊断，可以避免设置诊断阈值。使用RT-LAB搭建DAB变换器半实物系统进行实验。实验结果表明选择的3个诊断信号能够有效区分DAB变换器各IGBT开路故障状态。对比分析表明所提出的方法具有较高的故障诊断精度，平均诊断精度可达98.44 %。 |
| 关键词: 双有源桥变换器 故障诊断 递归图 脉冲耦合神经网络 深度残差收缩网络 |
| DOI：10.16081/j.epae.202207010 |
| 分类号:TM46 |
| 基金项目:国家重点研发计划项目(2020YFB0905905,2016YFF0102200)；国家自然科学基金重点资助项目(51637004)；国家自然科学基金资助项目(51977153,51977161,51577046)；中央高校基本科研业务费专项资金资助项目(2042021kf0233)；装备预先研究重点项目(41402040301)；湖北省重点研发计划项目(2021BEA162)；武汉市局科技计划项目(20201G01) |
|
| Open-circuit fault diagnosis method of DAB converter based on information fusion and DRSN |
|
ZHAO Yingying, HE Yigang, XING Zhikai, DU Bolun
|
|
School of Electrical Engineering and Automation, Wuhan University, Wuhan 430072, China
|
| Abstract: |
| Aiming at the problem of multi-signal diagnosis and diagnostic thresholds setting in open-fault diagnosis of dual active bridge(DAB) converter, an open-circuit fault diagnosis method of DAB converter based on information fusion and deep residual shrinkage network(DRSN) is proposed. Firstly, the diagnostic signals of DAB converter are reduced to three, which reduces the number of signal sensor. Secondly, the time series of the three diagnostic signals are transformed into images for information fusion by using the recurrence plot and the pulse-coupled neural network. The fused images contain the fault feature of different fault states and they are convenient to input into deep learning network for classification. Finally, the fused images are input into DRSN for fault diagnosis, which avoids setting the diagnostic thresholds. The hardware-in-the-loop system of DAB converter is built in RT-LAB for the experiments. The experimental results show that selected three diagnostic signals can effectively distinguish all insulated gate bipolar transistor(IGBT) open-circuit faults of DAB converter. The comparison analysis shows that the proposed method has high fault diagnosis accuracy, and the average diagnosis accuracy reaches 98.44 %. |
| Key words: dual active bridge converter fault diagnosis recurrence plot pulse-coupled neural network deep residual shrinkage network |
