基於變壓器油中氣體分析之 故障可信度研究及設備管理系統開發

Design of Transformer Fault Analysis and Facility Management System Based on Confidence Values

顏志丞(Chih-Cheng Yen)

指導教授 ： 姚立德

國立臺北科技大學/電資學院/電機工程系所/碩士(2010年)

摘要

為確保台電變電所設備資料之統一與準確性，落實計畫性維護點檢作業，並提升變電所設備管理效能，本文將既有『二次變電所設備管理系統』分散於各區處之Access資料庫，移轉至大型單一關聯式資料庫，並擴充開發既有之『二次變電所設備管理系統』，建立以網路為基礎之設備管理系統，供各區處得以透過企業網路(Intranet)查詢或維護所屬之資料庫，並結合無線射頻辨識技術運用於巡檢管理系統當中。本計畫將改善『二次變電所設備管理系統』，自動計算各類設備維護點檢週期及汰換時程，進行自動派工。且統計各種主要設備發生事故之種類、時間及地點等資料，提供日後事故分析、設備改善及預防性維護工作之參考。此系統亦可因應各項功能性操作，供系統人員印製各類報表。另一方面，本系統也將變壓器各項試驗維護歷史資料進行統計分析，開發油浸式電力變壓器之狀態基準維護(CBM)系統，使用本文提出之遺傳型橢圓分類法則(EECA)學習分類之判定空間，依此學習後之判定空間進而計算油中氣體成分對應之故障分類可信度，決定電力變壓器之量化健康狀態，進而提昇設備可靠度。

關鍵字

無線射頻辨識系統 ； 變電所 ； 模型辦認 ； 油中氣體分析

並列摘要

In order to unify Taiwan Power Compayn’s substation facility data, to improve data accuracy, to put substation facility inspection into effect and to improve the facility management efficiency, the currently installed substation facility management system (SSFMS) using Access database for individual disctrict office will be improved in this project. All of the distributed database at discrit offices will be concentrated and transferred into a large correlated database. Functions and capacity of SSFMS will be extended and expanded. SSFMS will be transformed to a Web based system allowing system operators at district offices search and maintain the database through Intranet and applications of RFID to the inspections on Substations. The SSFMS will also be modified to automatically determine scheduling for replacement or maintenance and to automatically distatch maintenance crew. Statistics of major substation facility malfucntion types, occurance time and spots will also be made by SSFMS for future analyses, improvements and preventive maintenance. The modified SSFMS will also provide system operators with various kinds of statistics reports, statements and tables.
Secondly, the maintaining historical data of transformer for statistical analysis ,we design a conditional base maintenance (CBM) system of the oil-immersed power transformers , using this proposed genetic-based classification rules oval (EECA) to learn the classification decision region,and then calculate the corresponding reliability of fault classification to determine the health status.

並列關鍵字

RFID ； Substation ； Patten Recognition ； DGA

參考文獻

[1]M. Kezunovic, B. Perunicic and J. Mrkic, “An Accurate Fault Location Algorithm Using Synchronized Sampling,” Electric Power Systems Research Journal, Vol. 29, No. 3, pp. 161-169, May 1994.

[3]M. Kezunovic and I. Rikalo, "Automating the Analysis of Faults and Power Quality," IEEE Computer Applications in Power, Vol. 12, No. 1, pp. 46-50, Jan. 1999.

[4]M. Kezunovic, Z. Ren, G. Latisko, D.R. Sevcik, J. Lucey, W. Cook, and E. Koch, "Automated monitoring and analysis of circuit breaker operation,” IEEE Transactions on Power Delivery, vol. 20, no. 3, pp. 1910-1918, Jul. 2005.

[5]Yen Ting Chao, Sheng Ta Lee, Hong-Chan Chang and Tsai-Hsiang Chen,“An Improvement Project for Distribution Transformer Load Management in Taiwan,” IEEE Trans. Power Syst., vol. 18, no. 2, pp. 875-881, May 2003

[6]Tsai Hsiang Chen and Jeng Tyan Cherng, “Design of a TLM Application Program Based on an AM/FNI/GIS System,” IEEE Trans. Power Syst., vol. 13, no. 3, pp. 904-909, Aug. 1998.