一種基于綜合推理的變壓器故障診斷方法
胡善芝,周冬旭,唐仁權
(國網南京供電公司,江蘇 南京 210019)
摘 要:針對現有變壓器油中溶解氣體分析方法的不足,提出一種基于綜合優化的故障診斷方法。將本體論與案例推理方法相結合,通過定義變壓器故障診斷領域知識與案例知識的使用、聚集關系,建立了基于本體論的變壓器故障案例庫,并進行案例推理。若無匹配源案例,則轉入基于判錯損失最小的可拓推理機制,通過建立變壓器各類故障的物元集模型,在考慮故障先驗概率的基礎上,引入錯判損失最小函數,提高了可拓推理的準確度。通過對收集的200余例變壓器故障實際DGA數據計算,并將診斷結果與IEC三比值法、可拓推理方法相比較,驗證了所提方法的有效性。
關鍵詞:變壓器;故障診斷;案例推理;可拓推理;錯判損失
中圖分類號:TM401+.1 文獻標識碼:A 文章編號:1007-3175(2019)03-0023-06
A Kind of Fault Diagnosis Method of Power Transformer Based on Integrated Reasoning
HU Shan-zhi, ZHOU Dong-xu, TANG Ren-quan
(State Grid Nanjing Power Supply Company, Nanijng 210019, China)
Abstract: Aiming at the shortage of existing gas dissolved analysis method in transformer oil, this paper proposed a fault diagnosis method based on comprehensive optimization. Combining the ontology with the case reasoning method, this paper established the case base of transformer faults based on the ontology and carried out case-based reasoning, with the definition of usage and aggregation relationship of knowledge of transformer faults diagnosis domain and cases. The extension reasoning mechanism based on the minimum of erroneous judgement started if there was no matching source case. The matter-element model of transformer faults was established, in consideration of the prior probability of all kinds of faults, which improved the accuracy of the extension reasoning. The effectiveness of this method is proved by 200 faults diagnosis examples of transformers, by comparing the results of the diagnosis with the IEC three-ratio method and the extension
reasoning mechanism.
Key words: power transformer; fault diagnosis; case-based reasoning; extension reasoning mechanism; erroneous judgement loss
參考文獻
[1] 鄭燁,黃海,趙海峰,等. 變壓器故障診斷專家系統的改良設計[J]. 電工電氣,2014(1):47-50.
[2] 束洪春.孫向飛,司大軍. 電力變壓器故障診斷專家系統知識庫建立和維護的粗糙集方法[J]. 中國電機工程學報,2002,22(2):3l-35.
[3] LEEH J, AHN B S, PARKY M. A Fault Diagnosis Expert System for Distribution Substations[J]. IEEE Transactions on Power Delivery,2000,15(1):92-97.
[4] 張珂斐,郭江,聶德鑫,等. 基于化學反應優化神經網絡與融合DGA算法的油浸式變壓器模型研究[J]. 高電壓技術,2016,42(4):1275-1281.
[5] 莫娟,王雪,董明,等. 基于粗糙集理論的電力變壓器故障診斷療法[J]. 中國電機工程學報,2004,24(7):162-167.
[6] 王建元,紀延超. 模糊Petri網絡知識表示方法及其在變壓器故障診斷中的應用[J]. 中國電機工程學報,2003,23(1):121-125.
[7] 朱永利,吳立增,李雪玉. 貝葉斯分類器與粗糙集相結合的變壓器綜合故障診斷[J]. 中國電機工程學報,2005,25(10):159-165.
[8] 陳小青,劉覺民,黃英偉,等.采用改進人工魚群優化粗糙集算法的變壓器故障診斷[J]. 高電壓技術,2012,38(6):1403-1409.
[9] 楊兵,丁輝,羅為民,等. 基于知識庫的變壓器故障診斷專家系統[J]. 中國電機工程學報,2002,22(10):121-124.
[10] GUARINO N.Formal Ontology and Information System[M].Trento:IOS Press,1998.
[11] 邰麗君, 趙韓, 董玉德, 等. 基于本體的機械產品知識倉庫研究與應用[J]. 工程圖學學報,2010(2):1-8.
[12] 黃文濤,趙學增,王偉杰,等. 基于物元模型的電力變壓器故障的可拓診斷方法[J]. 電力系統自動化,2004,28 (13):45-49.
[13] 原電力工業部. 變壓器油中溶解氣體分析和判斷導則:GB/T 7252—2001[S]. 北京:中國電力出版社,2001:15-16.