基于信息融合的變壓器故障多級(jí)診斷方法
張愛蘭,許志元,楊琦欣,劉春明,朱彥瑋,李曉磊
(國網(wǎng)山東省電力公司濟(jì)南供電公司,山東 濟(jì)南 250012)
摘 要:建立了基于信息融合的變壓器故障多級(jí)診斷模型,該模型融合了在線監(jiān)測、油中溶解氣體、電氣試驗(yàn)等多源數(shù)據(jù)信息。采用自適應(yīng)遺傳算法優(yōu)化的小波神經(jīng)網(wǎng)絡(luò)對(duì)變壓器故障進(jìn)行初級(jí)診斷,通過改進(jìn)D-S證據(jù)理論對(duì)初級(jí)診斷結(jié)果進(jìn)行決策級(jí)融合,實(shí)現(xiàn)對(duì)變壓器故障的深度診斷與定位。通過應(yīng)用實(shí)例證明,該方法可以有效提高變壓器故障診斷的精度和可信度,減小診斷的不確定性。
關(guān)鍵詞:變壓器故障;多級(jí)診斷;改進(jìn)D-S證據(jù)理論;信息融合
中圖分類號(hào):TM411 文獻(xiàn)標(biāo)識(shí)碼:A 文章編號(hào):1007-3175(2019)06-0015-06
Multi-Level Diagnosis Method of Transformer Fault Based on Information Fusion
ZHANG Ai-lan, XU Zhi-yuan, YANG Qi-xin, LIU Chun-ming, ZHU Yan-wei, LI Xiao-lei
(State Grid Shandong Electric Power Company Jinan Power Supply Company, Jinan 250012, China)
Abstract: This paper established a multi-level diagnosis model of transformer faults based on information fusion. This model integrated multi-source data information in transformer faults, such as the on-line monitoring data, dissolved gas in oil and electrical test. The adaptive genetic algorithm was adopted to optimize the wavelet neural network, so as to implement primary diagnosis of transformer faults. The improved D-S evidence theory was used to carry out decision-Level fusion of primary diagnostic results to realize the depth diagnosis and location of transformer faults. The application example shows that this method can improve the accuracy and reliability of transformer fault diagnosis, reducing the diagnostic uncertainty.
Key words: transformer fault; multi-level diagnosis; improved D-S evidence theory; information fusion
參考文獻(xiàn)
[1] 李剛,于長海,劉云鵬,等. 電力變壓器故障預(yù)測與健康管理:挑戰(zhàn)與展望[J]. 電力系統(tǒng)自動(dòng)化,2017,41(23):156-167.
[2] 汪可,李金忠,張書琦,等. 變壓器故障診斷用油中溶解氣體新特征參量[J]. 中國電機(jī)工程學(xué)報(bào),2016,36(23):6570-6578.
[3] 苗長新,申坤,鐘世華,等. 基于優(yōu)化神經(jīng)網(wǎng)絡(luò)和DGA的變壓器故障診斷[J]. 高壓電器,2016,52(11):163-168.
[4] 鄭含博,王偉,李曉綱,等. 基于多分類最小二乘支持向量機(jī)和改進(jìn)粒子群優(yōu)化算法的電力變壓器故障診斷方法[J]. 高電壓技術(shù),2014,40(11):3424-3429.
[5] 吳坤,康建設(shè),池闊. 基于改進(jìn)多分類算法和相關(guān)向量機(jī)的電力變壓器故障診斷方法[J]. 高電壓技術(shù),2016,42(9):3011-3017.
[6] HU Yifei, TAN Jing, DENG Lianbo, et al. Big Data and Advanced Analysis f o r Network Components Fault Diagnosing[C]//International Conference on Electricity Distribution,2018:17-19.
[7] 沈懷榮,楊露,周偉靜,等. 信息融合故障診斷技術(shù)[M]. 北京:科學(xué)出版社,2013.
[8] WANG Xuelei, LI Qingmin, LI Chengrong, et al. Reliability Assessment of the Fault Diagnosis Methodologies for Transformers and a New Diagnostic Scheme Based on Fault Info Integration[J].IEEE Transactions on Dielectrics & Electrical Insulation,2013,20(6):2292-2298.
[9] 梁艷香, 程加堂. 證據(jù)理論在變壓器故障診斷中的應(yīng)用[J]. 科學(xué)技術(shù)與工程,2012,12(5):1158-1160.
[10] 葉漢民,肖尊定. 多傳感器信息融合技術(shù)在變壓器故障診斷中的應(yīng)用[J]. 變壓器,2014,51(9):60-64.
[11] 司馬莉萍,舒乃秋,李自品,等. 基于SVM和D -S證據(jù)理論的電力變壓器內(nèi)部故障部位識(shí)別[J]. 電力自動(dòng)化設(shè)備,2012,32(11):72-76.
[12] 陳偉根,劉娟,曹敏. 基于信息融合的變壓器內(nèi)部故障診斷方法[J]. 高電壓技術(shù),2015,41(11):3797-3803.
[13] KARI T, GAO W, ZHAO D, et al.An integrated method of ANFIS and Dempster-Shafer theory for fault diagnosis of power transformer[J]. IEEE Transactions on Dielectrics and Electrical Insulation,2018,25(1):360-371.
[14] 阮羚,謝齊家,高勝友,等. 人工神經(jīng)網(wǎng)絡(luò)和信息融合技術(shù)在變壓器狀態(tài)評(píng)估中的應(yīng)用[J]. 高電壓技術(shù),2014,40(3):822-828.
[15] 陳偉根,潘翀,王有元,等. 利用小波神經(jīng)網(wǎng)絡(luò)的電力變壓器故障診斷方法[J]. 高電壓技術(shù),2007,33(8):52-55.
[16] 錢國超. 大型電力變壓器基于信息融合故障診斷技術(shù)的研究[D]. 重慶:重慶大學(xué),2008.
[17] 陳偉根,凌云,甘德剛,等. 基于聚類- 小波神經(jīng)網(wǎng)絡(luò)的油紙絕緣氣隙放電發(fā)展階段識(shí)別方法[J].電網(wǎng)技術(shù),2012,36(7):126-132.