本文主要提出兩種方法應用於以局部放電相位分析圖(PRPD圖譜)為輸入特徵，並使用卷積神經網路辨識局部放電類型。第一種方法以突出度(Prominence)與峰高的概念，分析局部放電波型主峰與子峰之間的差異，以改善傳統波峰檢測方法閥值設定的問題。當閥值設高容易受到較大的局部放電信號影響，而濾除大部分較小的局部放電，閥值設低則容易將子峰當作一次局部放電，兩者都會衍生對局部放電次數計算的問題。第二種方法是為了解決參考相位在局部放電相位分析圖上局部放電圖譜相位移的問題。其結合模算術(module arithmetic)的公式分析局部放電相位分析圖上像素之間的相位差，以產生一種不受相位移影響的新特徵，該特徵由於相位移不變性使得卷積神經網路的準確度穩定並可以成功分辨出不同的局部放電類型。本研究系統使用高頻比流器(HFCT)、暫態對地電壓感測器(TEV)和超高頻感測器(UHF Sensor)三種感測器量測高壓設備，並使用具有內部放電的模鑄式變壓器、尖端對圓盤的尖端放電模型與潑灑鹽水的方式模擬出內部放電、尖端放電與表面放電。實驗結果以平移局部放電相位分析圖圖譜的方式，驗證提出方法的有效性與改善傳統量測分析的方法。

This study proposes two novel methods to be applied to the system that uses Phase Resolution Partial Discharge (PRPD) as input feature of convolutional neural network (CNN) to diagnose the partial discharge type of power equipment. The first method combines the concept of prominence and height of topography to analyze the difference between the main(highest) peak and the sub-peak of partial discharge waveform. For improving the problem that the traditional peak detection hardly determines the threshold. When the threshold value is high, it is often affected by larger partial discharge signals and ignores most smaller partial discharge signals. If the threshold value is low, the sub-peak is treated as once partial discharge. The second method is proposed to solve the phase-shifting problem of PRPD caused by the reference phase. The method uses modular arithmetic to analyze the phase distance between pixels of PRPD to generate a new feature that isn’t affected by phase shifting. In this study, the system uses three types of sensors, high-frequency current transformer (HFCT), transient earth voltage(TEV) sensor, and ultra-high frequency(UHF) sensor to measure power equipment to distinguish internal, surface, and corona partial discharge. The effectiveness of the proposed methods and the fault of the current system will be discussed in the results.

1. Y. Lecun, B. Boser, J. S. Denker, D. Henderson, R. E. Howard, W. Hubbard and L. D. Jackel: Backpropagation Applied to Handwritten Zip Code Recognition, Neural Computation, 1(4):541-551, Winter 1989.
2. Y. Lecun, L. Bottou, Y. Bengio and P. Haffner, "Gradient-based learning applied to document recognition," in Proceedings of the IEEE, vol. 86, no. 11, pp. 2278-2324, Nov. 1998, doi: 10.1109/5.726791.
3. M. D. Zeiler, and R. Fergus. "Visualizing and understanding convolutional networks." European conference on computer vision. Springer, Cham, 2014.
4. K. Simonyan, and A. Zisserman, (2014). Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556.
5. C. Szegedy, W. Liu, Y. Jia, P. Sermanet, S. Reed, D. Anguelov, ... and A. Rabinovich, (2015). Going deeper with convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 1-9).
6. K. He, X. Zhang, S. Ren, and J. Sun,(2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 770-778).
7. F. N. Iandola, S. Han, M. W. Moskewicz, K. Ashraf, W. J. Dally, and K. Keutzer, (2016). SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 0.5 MB model size. arXiv preprint arXiv:1602.07360.
8. Y. Gao, W. Zhu, Cai and Y. Zhang, "Pattern Recognition of Partial Discharge Based on VMD-CWD Spectrum and Optimized CNN With Cross-Layer Feature Fusion," in IEEE Access, vol. 8, pp. 151296-151306, 2020, doi: 10.1109/ACCESS.2020.3017047.
9. Q. Che, H. Wen, X. Li, Z. Peng and K. P. Chen, "Partial Discharge Recognition Based on Optical Fiber Distributed Acoustic Sensing and a Convolutional Neural Network," in IEEE Access, vol. 7, pp. 101758-101764, 2019, doi: 10.1109/ACCESS.2019.2931040.
10. F. -C. Gu, "Identification of Partial Discharge Defects in Gas-Insulated Switchgears by Using a Deep Learning Method," in IEEE Access, vol. 8, pp. 163894-163902, 2020, doi: 10.1109/ACCESS.2020.3018553.
11. X. Peng et al., "A Convolutional Neural Network-Based Deep Learning Methodology for Recognition of Partial Discharge Patterns from High-Voltage Cables," in IEEE Transactions on Power Delivery, vol. 34, no. 4, pp. 1460-1469, Aug. 2019, doi: 10.1109/TPWRD.2019.2906086.
12. T. -D. Do, V. -N. Tuyet-Doan, Y. -S. Cho, J. -H. Sun and Y. -H. Kim, "Convolutional-Neural-Network-Based Partial Discharge Diagnosis for Power Transformer Using UHF Sensor," in IEEE Access, vol. 8, pp. 207377-207388, 2020, doi: 10.1109/ACCESS.2020.3038386.
13. 林鈺洋，「基於卷積神經網路與資料融合技術之局部放電檢測系統」國立成功大學電機工程學系碩士論文，2020年。
14. S. Kim, J. Jung, Y. Kim, G. Kil and G. Wang, "New diagnosis method of unknown phase-shifted PD signals for gas insulated switchgears," in IEEE Transactions on Dielectrics and Electrical Insulation, vol. 25, no. 1, pp. 102-109, Feb. 2018, doi: 10.1109/TDEI.2018.006739.
15. J. H. Lee, N. Hozumi and T. Okamoto, "Discrimination of phase-shifted partial discharge patterns by neural network using standardization method," Proceedings of 1994 IEEE International Symposium on Electrical Insulation, 1994, pp. 314-317, doi: 10.1109/ELINSL.1994.401504.
16. S. Malla, A Wavelet Tour of Signal Processing, Third Edition: The Sparse Way
17. R. Hussein, K. B. Shaban and A. H. El-Hag, "Wavelet Transform With Histogram-Based Threshold Estimation for Online Partial Discharge Signal Denoising," in IEEE Transactions on Instrumentation and Measurement, vol. 64, no. 12, pp. 3601-3614, Dec. 2015, doi: 10.1109/TIM.2015.2454651.
18. S. Sriram, S. Nitin, K. M. M. Prabhu and M. J. Bastiaans, "Signal denoising techniques for partial discharge measurements," in IEEE Transactions on Dielectrics and Electrical Insulation, vol. 12, no. 6, pp. 1182-1191, Dec. 2005, doi: 10.1109/TDEI.2005.1561798.
19. J. Li, C. Cheng, T. Jiang and S. Grzybowski, "Wavelet de-noising of partial discharge signals based on genetic adaptive threshold estimation," in IEEE Transactions on Dielectrics and Electrical Insulation, vol. 19, no. 2, pp. 543-549, April 2012, doi: 10.1109/TDEI.2012.6180248.
20. J. Brown, "Analytic signals and product theorems for Hilbert transforms," in IEEE Transactions on Circuits and Systems, vol. 21, no. 6, pp. 790-792, November 1974, doi: 10.1109/TCS.1974.1083928.
21. E. Bedrosian, "The Analytic Signal Representation of Modulated Waveforms," in Proceedings of the IRE, vol. 50, no. 10, pp. 2071-2076, Oct. 1962, doi: 10.1109/JRPROC.1962.288236.
22. H. B. Voelcker, "Toward a unified theory of modulation part I: Phase-envelope relationships," in Proceedings of the IEEE, vol. 54, no. 3, pp. 340-353, March 1966, doi: 10.1109/PROC.1966.4695.