電力電纜的廣泛應用對電力網路發展扮演關鍵的角色，電網的運行可靠度常因電纜接頭絕緣故障而顯著惡化，而局部放電技術已被國際認可為檢測設備絕緣狀態的重要指標之一，且有助於維護人員建立設備維護的優先次序及預防性維修計畫。典型研究多以相位分析法搭配機器學習演算法去探討被試物的局部放電行為，企圖評估電纜的瑕疵類型或絕緣狀態，惟因相位分析法仰賴電壓訊號及相位參考之訊息，在艱難的量測場域將面臨阻礙。
因此，本文製作地下電纜接頭瑕疵試驗樣本，採用脈衝序列分析法考慮局部放電連續脈衝之間的動態特性，首先，將被試物之局部放電訊號轉為脈衝序列特徵圖，而非典型相位分析圖譜依賴之相位訊息，本文針對脈衝序列圖譜提出參數設置最佳化之建議，包含輸入圖像之畫素大小、脈衝序列型式、色彩型式以及標記點大小；隨後利用卷積神經網路在圖像識別上的優勢，建立絕緣狀態識別模型，本文提出自我反饋以及擾動觀察之方式，依當回合之絕緣狀態診斷結果，作為下回合訓練集數據之標籤依據，直至診斷結果之變動率低於5%為止；最後搭配轉態密度法作為卷積神經網路模型輸出的後端處理程序，以決定被試物何時進入危險期之依據。結果顯示，大多數樣本之轉態點會收斂於一或二個穩定值位置，且此結果近似於尋找轉折演算法的結果。
本文提出之診斷機制適用於電纜接頭絕緣狀態之評估工作，並以瑕疵識別及破壞路徑識別作為輔助，結果顯示該診斷機制能在被試物發生絕緣破壞前進行危險預測，給予工程人員充裕的時間進行設備維護或汰換之依據。

The wide application of power cables plays an important role in the development of power grids. The reliable operation of a power network is significantly deteriorated by the insulation failure of cable joints. Partial discharge (PD) detection is an effective technique for diagnosing the insulation status and defects of power cable joints, which is expected to provide condition-based monitoring and maintenance for cable accessories. The conventional researches mostly apply phase-resolved analysis (PRA) and machine learning algorithms to investigate the PD behavior of the power equipment in an attempt to evaluate the defect type or insulation status. Whereas PRA relies on the voltage signal and phase position, PD detection may not be feasible, especially in a scenario with a hostile on-site environment.
To address the aforementioned problem, pulse sequence analysis (PSA) is applied to consider the dynamic characteristics between consecutive PD pulses rather than phase positions. Firstly, the experimental PD data from defective cable joints were preprocessed into PSA patterns. The key configurations of the PSA pattern that influence deep learning-based pattern recognition accuracy were analyzed, including the size, type, color, and marker size of the input image. Secondly, the recognition model of insulation status was established through convolutional neural networks (CNNs) which have high accuracy and effectiveness in pattern recognition. This study proposes self-feedback and disturbance analysis to acquire an opportune time of maintenance for cable joints. The diagnosis result of the insulation status in the current round will be used as the label basis for the training set data of the next round until the rate of change of the diagnosis result is less than 5%. Finally, a transition density method is used as the back-end processing program for the output of the CNN model to determine when the cable joint turns into crisis. The results show that the transition point of most samples will converge to one or two stable positions, and this result is similar to the findchangepts algorithm ones.
In this paper, the proposed diagnosis mechanism of insulation status is applicable to the cable joints. In addition, defect recognition and damage recognition are included in the mechanism as auxiliaries. The results show that the proposed diagnosis mechanism can predict the risk before the cable joint occurs insulation breakdown. It is expected that the proposed diagnosis mechanism conforms to economic benefits which may provide condition-based monitoring and maintenance for cable accessories.

[1] A. Eigner and K. Rethmeier, "An overview on the current status of partial dicharge measurements on AC high voltage cable accessories," IEEE Electrical Insulation Magazine, vol. 32, no. 2, pp. 48-55, Mar.-Apr. 2016.
[2] N. Sahoo, M. Salama and R. Bartnikas, "Trends in partial discharge pattern classification: A survey," IEEE Transactions on Dielectrics and Electrical Insulation, vol. 12, no. 2, pp. 248-264, Apr. 2005.
[3] M. Hoof and R. Patsch, "Voltage-difference analysis a tool for partial discharge source identification," Conference Record of the 1996 IEEE International Symposium on Electrical Insulation, vol. 1, pp. 401-406, Jun. 1996.
[4] N. H. Aziz, V. M. Catterson, S. M. Rowland and S. Bahadoorsingh, "Analysis of partial discharge features as prognostic indicators of electrical treeing," IEEE Transactions on Dielectrics and Electrical Insulation, vol. 24, no. 1, pp. 129-136, Feb. 2017.
[5] Q. Khan, S. S. Refaat, H. Abu-Rub, and H. A. Toliyat, "Partial discharge detection and diagnosis in gas insulated switchgear: State of the art," IEEE Electrical Insulation Magazine, vol. 35, no. 4, pp. 16–33, Jul./Aug. 2019.
[6] R. Umamaheswari and R. Sarathi, "Identification of partial discharges in gas-insulated switchgear by ultra-high-frequency technique and classification by adopting multi-class support vector machines," Electric Power Components and Systems, vol. 39, no. 14, pp. 1577–1595, 2011.
[7] Y. Wang, D. Chang, S. Qin, Y. Fan, H. Mu and G. Zhang, "Separating multi-source partial discharge signals using linear prediction analysis and isolation forest algorithm," IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 6, pp. 2734-2742, Jun. 2020
[8] K. B. Lee, S. Cheon and C. O. Kim, "A convolutional neural network for fault classification and diagnosis in semiconductor manufacturing processes," IEEE Transactions on Semiconductor Manufacturing, vol. 30, no. 2, pp. 135-142, May 2017.
[9] 賴佳駿，基於卷積神經網路於交連聚乙烯電纜接頭之狀態診斷及瑕疵識別研究，碩士論文，臺北，國立臺灣科技大學電機工程系，民國110年。
[10] 劉鑑，地下電纜接頭絕緣狀態轉變之研究，碩士論文，臺北，國立臺灣科技大學電機工程系，民國108年。
[11] 陳毅哲，運用支持向量機理論評估電力電纜絕緣劣化狀態，碩士論文，臺北，國立臺灣科技大學電機工程系，民國108年。
[12] 謝沅澔，電力電纜接頭絕緣狀態智能化診斷規則之研究，碩士論文，臺北，國立臺灣科技大學電機工程系，民國107年。
[13] 雷宗憲，暴力演算法應用於地下電纜接頭絕緣狀態評估規則之研究，碩士論文，臺北，國立臺灣科技大學電機工程系，民國106年。
[14] 戴嘉宏，支持向量機應用於電力電纜接頭絕緣狀態評估之研究，碩士論文，臺北，國立臺灣科技大學電機工程系，民國105年。
[15] 張加明，電力電纜接頭瑕疵演進之絕緣狀態警示研究，碩士論文，臺北，國立臺灣科技大學電機工程系，民國105年。
[16] 張瑞村，數位局部放電量測應用於高壓電纜終端接頭絕緣狀態之評估，碩士論文，臺北，國立臺灣科技大學電機工程系，民國94年。
[17] F. Duan, "Induction motor parameters estimation and faults diagnosis using optimization algorithms," PhD. dissertation, University of Adelaide, Australia, School of Electrical and Electronic Engineering, 2014.
[18] C. K. Chang, C. S. Lai and R. N. Wu, "Decision tree rules for insulation condition assessment of pre-molded power cable joints with artificial defects," IEEE Transactions on Dielectrics and Electrical Insulation, Vol. 26, No. 5, pp. 1636-1644, 2019.
[19] S. T. Li and J. Y. Li, "Condition monitoring and diagnosis of power equipment: Review and prospective," High Voltage, Vol. 2, No. 2, 2017.
[20] "High-voltage test techniques - partial discharge measurements," IEC 60270, 3rd ed., 2000.
[21] 張建國，地下電纜接頭局部放電線上監控系統之研製，碩士論文，臺北，國立臺灣科技大學電機工程系，民國95年。
[22] 台灣積體電路製造股份有限公司，局部放電偵測方法、特高頻天線、局部放電偵測系統以及訊號處理單元，中華民國發明專利第I598601號，民國105年。
[23] 恒揚電機技術顧問股份有限公司，局部放電檢測設備，中華民國新型專利第M568372號，民國107年。
[24] H. Prasetia, U. Khayam, Suwarno, A. Itose, M. Kozako and M. Hikita, "PD pattern of various defects measured by TEV sensor," 2017 International Conference on High Voltage Engineering and Power Systems (ICHVEPS), Sanur, 2017.
[25] L. Lundgaard, "Partial discharge. XIII. Acoustic partial discharge detection-fundamental considerations," IEEE Electrical Insulation Magazine, Vol. 8, pp. 25-31, 1992.
[26] 台灣電力公司，配電技術手冊(四)地下配電線路設計，民國85年。
[27] 台灣電力公司，高壓單芯交連PE電纜(A043)，材料規範，民國102年。
[28] 李卓翰，地下電纜接頭局部放電之差動量測研究，碩士論文，臺北，國立臺灣科技大學電機工程系，民國109年。
[29] J. Wang, P. E. C. Stone, D. Coats, Y. Shin and R. A. Dougal, "Health monitoring of power cable via joint time-frequency domain reflectometry," IEEE Transactions on Instrumentation and Measurement, vol. 60, no. 3, pp. 1047-1053, Mar. 2011.
[30] R. J. V. Brunt, E. W. Cernyar and P. V. Glahn, "Importance of unraveling memory propagation effects in interpreting data on partial discharge statistics," IEEE Transactions on Electrical Insulation, vol. 28, pp. 905-916, 1993.
[31] D. Paul, "Failure analysis of dry-type power transformer," IEEE Transactions on Industry Applications, vol. 37, no. 3, pp. 689-695, May-Jun. 2001.
[32] 林育勳，數位局部放電量測應用於模鑄式比流器，碩士論文，臺北，國立臺灣科技大學電機工程系，民國93年。
[33] 潘彥竹，局部放電量測系統之研製，碩士論文，臺北，國立臺灣科技大學電機工程系，民國93年。
[34] R. Patsch and F. Berton, "Pulse sequence analysis—a diagnostic tool based on the physics behind partial discharges," Journal of Physics D: Applied Physics, vol. 35, no. 1, pp. 25-32, 2002.
[35] Y. Lecun, L. Bottou, Y. Bengio and P. Haffner, "Gradient-based learning applied to document recognition," Proceedings of the IEEE, vol. 86, no. 11, pp. 2278-2324, Nov. 1998.
[36] A. Krizhevsky, I. Sutskever and G. E. Hinton, "Imagenet classification with deep convolutional neural networks," Advances in neural information processing systems, 2012.
[37] X. Peng et al., "A convolutional neural network-based deep learning methodology for recognition of partial discharge patterns from high-voltage cables," IEEE Transactions on Power Delivery, vol. 34, no. 4, pp. 1460-1469, Aug. 2019.
[38] M. A. Ranzato, "Deep learning lecture 6," Microsoft AI Research.
[39] D. Liu et al., "Learning temporal dynamics for video super-resolution: A deep learning approach," IEEE Transactions on Image Processing, vol. 27, no. 7, pp. 3432-3435, Jul. 2018.
[40] X. Glorot, A. Bordes and Y. Bengio, "Deep sparse rectifier neural networks," Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, PMLR 15:315-323, 2011.
[41] N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov, "Dropout: A simple way to prevent neural networks from overfitting," Journal of Machine Learning Research, 2014.
[42] P. Mishra, P. Pai, M. Patel and R. Sonkusare, "Extraction of information from handwriting using optical character recognition and neural networks," 2020 4th International Conference on Electronics, Communication and Aerospace Technology (ICECA), pp. 1328-1333, 2020.
[43] 郭至恩，深度學習：從入門到實戰(使用MATLAB)，全華圖書，民國109年。
[44] C. K. Chang, R. N. Wu, C. C. Lai, H. H. Chang and B. K. Boyanapalli, "Partial discharge pattern recognition for underground cable joints using convolutional neural network," 2020 International Conference on Pervasive Artificial Intelligence (ICPAI), pp. 234-239, 2020.
[45] P. Donoso, R. Schurch, J. Ardila and L. Orellana, "Analysis of partial discharges in electrical tree growth under very low frequency (VLF) excitation through pulse sequence and nonlinear time series analysis," IEEE Access, vol. 8, pp. 163673-163684, 2020.
[46] J. Densley, T. Kalicki and Z. Nodolny, "Characteristics of PD pulses in electrical trees and interfaces in extruded cables," IEEE Transactions on Dielectrics and Electrical Insulation, vol. 8, no. 1, pp. 48–57, Feb. 2001.
[47] 張建國，高電壓地下電纜接頭絕緣狀態之監測與診斷系統之研究，博士論文，臺北，國立臺灣科技大學電機工程系，民國101年。