局部放電(Partial Discharge, PD)是檢測電力設備絕緣狀態的指標之一，如何有效蒐集資料與即時分析為辨識成功與否之關鍵。因此本文提出一個基於局部放電的電力設備診斷平台，特點為增加實驗數據的格式一致化及操作上的便利性。進而建立一套可依循擴展的架構的介面，也有助於後續的維護與創新。
本文選用Flask輕量級網頁框架作為網頁伺服器，其具有基礎的WSGI（Web Server Gateway Interface）功能，負責用戶與後端程式之間的溝通。並使用MongoDB作為本平台之資料庫，能儲存非結構化的資料，相較於傳統SQL資料庫，使用上更為靈活。
本平台使用Python之開源環境進行開發，使用MATLAB Production Server整合MATLAB端的功能並與Python建立連結，利用Flask建立網頁伺服器，讓使用者可於不同設備登入此平台進行操作。透過將量測數據格式與深度學習參數標準化，使資料庫同時儲存不同地點之設備的放電數據，配合深度學習之辨識功能可達到即時辨識的效果。
本文展示一個使用已開發的卷積神經網路（Convolutional Neural Networks, CNN）與本平台整合的案例，使用已設計好的模型架構與資料蒐集流程，簡化複雜的訓練過程，提供易於操作的測試環境。利用設定即時辨識之功能，使用多種訓練模型進行即時的局部放電絕緣瑕疵辨識與絕緣狀態診斷，可將進入危險期之電力設備提早更換，避免發生非預期的電力事故。

Partial Discharge (PD) is one of the indicators to detect the insulation status of power equipment. How to effectively collect data and analyze in real time is the key to the success of the detection. Therefore, this paper proposes a power equipment diagnosis platform based on partial discharge, which is characterized by increasing the consistency of the format of the experimental data and the convenience of operation. Furthermore, the establishment of a set of interfaces that can follow the extended architecture is also helpful for subsequent maintenance and innovation.
This paper chooses the Flask lightweight web framework as the web server, which has the basic WSGI (Web Server Gateway Interface) function and is responsible for the communication between the user and the back-end program. And use MongoDB as the database of this platform, which can store unstructured data. Compared with the traditional SQL database, it is more flexible in use.
This platform uses the open source environment of Python for development, uses MATLAB Production Server to integrate the functions of MATLAB and establish a link with Python, and uses Flask to build a web server, so that users can log in to the platform on different devices to operate. By standardizing the measurement data format and deep learning parameters, the database can store the discharge data of equipment in different locations at the same time, and the effect of real-time analysis can be achieved with online operation.
This paper shows a case of integrating the developed Convolutional Neural Networks (CNN) with this platform, using the designed model structure and data collection process to simplify the complex training process and provide an easy-to-operate test environment. By using the function of setting real-time identification and using various training models for real-time partial discharge insulation defect identification and insulation state diagnosis, When the electric power equipment entering the dangerous period, replace equipment early to avoid unexpected electrical accidents.

[1] L. Niemeyer, "A generalized approach to partial discharge modeling", IEEE Transactions on Dielectrics and Electrical Insulation, Aug. 1995
[2] 張軒豪，應用局部放電脈衝序列分析與卷積神經網路於電纜接頭 之絕緣狀態評估，碩士論文，臺北，國立臺灣科技大學電機工程 系，民國111年。
[3] 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。
[4] 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。
[5] 吳佑軒，環氧樹脂材料電樹生成與局部放電之關聯性研究，碩士論文，臺北，國立臺灣科技大學電機工程系，民國111年。
[6] L. Renforth, R. Giussani, M. T. Mendiola and L. Dodd , "Online partial pischarge insulation condition monitoring of complete high-voltage networks", IEEE Transactions on Industry Applications, Jan.-Feb. 2019
[7] L. Renforth, P. S. Hamer, D. Clark, S. Goodfellow and R.Tower, "Continuous, remote on-line partial discharge (OLPD) monitoring of HV EX/ATEX motors in the oil and gas industry", Industry Applications Society 60th Annual Petroleum and Chemical Industry Conference, Nov. 2013
[8] G. C. Montanari and R. Ghosh, "Wind farm electrical asset components: insulation diagnostics through partial discharge measurements", 2021 IEEE Green Technologies Conference (GreenTech), Jun. 2021
[9] A. Caprara, G. Ciotti and L. Paschini, "Analysis of the results of on-line partial discharge monitoring and the impact of the maintenance actions on a 30MVA synchronous generator", 2020 IEEE Electrical Insulation Conference (EIC), Aug. 2020
[10] A. Khalyasmaa, A. Stepanova, S. Eroshenko, V. Bolgov and T. D. Chung, "The application of partial discharge monitoring system for instrument transformers: special issues", 2020 21st International Symposium on Electrical Apparatus & Technologies (SIELA), Aug. 2020
[11] Y. Liao, B. Feng, X. Gu, T. Sun, Y. Xu and Z. Zhang, "Application of the online partial discharge monitoring for the EHV XLPE cable system", 2016 International Conference on Condition Monitoring and Diagnosis (CMD), Dec. 2016
[12] J. Wang, Z. Wang, D. Zhou, Y. He, J. Zhang and X. Yao, "The research and application on partial discharge intensive care system in GIS based on detection technology combined ultrasonic and electrical", 2020 IEEE International Conference on High Voltage Engineering and Application (ICHVE), Dec. 2020
[13] B. Ward and T. Whyte, "Online partial discharge monitoring and failure analysis case study of a 275 kV oil filled cable termination", 2022 IEEE Electrical Insulation Conference (EIC), Jul. 2022
[14] Y. Wang, J. Gong, Z. Zeng and Z. Wang, "Research on on-line monitoring fault diagnosis system of GIS UHF partial discharge", 2019 IEEE Innovative Smart Grid Technologies - Asia (ISGT Asia), Oct. 2019
[15] Y. Hao, Y. Chen, Y. Chen, T. Huang, J. Xiao and L. Yang, "A partial discharge detection system for XLPE cable terminals and acoustic wave sensing characteristics based on fiber optic mach-zehnder interferometer", IEEE Sensors Journal, Oct. 2021
[16] B. Niu, "Research and application of intermittent partial discharge characteristics and easy-warning system for electric equipment", 2022 The 4th International Conference on Clean Energy and Electrical Systems (CEES 2022), Apr. 2022
[17] G. C. Montanari, R. Hebner, P. Seri and R. Ghosh, "Self-assessment of health conditions of electrical assets and grid components: a contribution to smart grids", IEEE Transactions on Smart Grid, Mar. 2021
[18] G. C. Montanari and R. Ghosh, " Wind farm electrical asset components: insulation diagnostics through partial discharge measurements", 2021 IEEE Green Technologies Conference (GreenTech), Jun. 2021
[19] 詹婉渝，基於深度學習影像辨識之稻米期別辨識系統，碩士論文，臺北，國立臺灣大學工程科學及海洋工程學研究所，民國110年。
[20] 周昱安，非監督式網頁資料擷取、轉置、載入與輸出系統，碩士論文，臺北，國立中央大學軟體工程研究所，民國106年。
[21] 劉文誠，可用於自動訓練深度學習網路的網頁服務，碩士論文，桃園，國立中央大學資訊工程學系，民國107年。
[22] 王喬韋，通過整合資料虛擬化和資料編目邁向 Data Fabric，碩士論文，臺南，國立成功大學人工智慧科技碩士學位學程，民國111年。
[23] 張力仁，基於聯邦式資料庫技術的 FHIR 中介系統設計與實作，碩士論文，臺北，國立臺灣科技大學資訊管理系，民國108年。
[24] 佐藤昌基、平田哲也、寺田學，實戰Python Flask開發｜基礎知識x物件偵測x機器學習應用，民國111年。