Partial Discharge Pattern Recognition Method Based on Transfer Learning and DenseNet Model

• Published in: IEEE transactions on dielectrics and electrical insulation Vol. 30; no. 3; pp. 1240 - 1246
• Main Authors: Yuwei, Fu, Liejuan, Liang, Weihua, Huang, Guobin, Huang, Peijun, Huang, Zhiyu, Zhang, Chi, Chen, Chuang, Wang
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.06.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Artificial neural networks; Data models; Deep learning; Defects; DenseNet; Discharge; Feature extraction; Finite difference time domain method; Gas insulation; gas-insulated switchgear (GIS); Image compression; Image enhancement; Image processing; Insulation; Machine learning; Model accuracy; partial discharge (PD); Partial discharges; Pattern recognition; Structured data; Switchgear; Time-domain analysis; Transfer learning; Waveforms
• ISSN: 1070-9878; 1558-4135
• DOI: 10.1109/TDEI.2023.3239032

With the development of intelligent sensing technology, a large amount of partial discharge (PD) time-domain waveform images are generated in the on-site detection of gas-insulated switchgear (GIS) PD. Traditional pattern recognition methods are mostly aimed at structured data and cannot directly identify defect types of such data. At the same time, the deep learning method for GIS PD pattern recognition is generally faced with the problem of small samples. In order to solve the above problems, this article proposes a PD pattern recognition method based on transfer learning and DenseNet model. First, the time-domain waveform images are processed by image enhancement, normalization, image compression, and other image processing techniques. The finite-difference time-domain (FDTD) method was used to simulate GIS PD, and the time-domain waveform image database of four PD defects is established. Using convolutional neural network (CNN) and transfer learning, the recognition accuracy of the model is increased to 95%, with better robustness. The recognition performance of different CNN structures is studied. The results show that DenseNet model has higher accuracy than other structures and shorter training time. This study can be used to diagnose the insulation status of GIS equipment in-site.