基于特征分類算法的GIS故障診斷方法研究
張湛1,楊光2,黃志2,張峰2,張士文2
(1 中國(guó)電力工程顧問(wèn)集團(tuán)中南電力設(shè)計(jì)院, 湖北 武漢 430071; 2 上海交通大學(xué) 電子信息與電氣工程學(xué)院, 上海 200240)
摘 要:針對(duì)高壓斷路器操動(dòng)機(jī)構(gòu)故障監(jiān)測(cè)問(wèn)題,提出了一種基于核主成分分析和支持向量機(jī)的氣體絕緣開(kāi)關(guān)故障檢測(cè)方法,利用核主成分分析對(duì)分( 合) 閘線圈電流波形的特征值進(jìn)行降維,然后將降維后的特征值輸入多類分類SVM 進(jìn)行故障診斷和分類。通過(guò)實(shí)際樣本的實(shí)驗(yàn),驗(yàn)證了算法的準(zhǔn)確性和可靠性,并通過(guò)參數(shù)討論,測(cè)算了最優(yōu)的分類參數(shù)。
關(guān)鍵詞:故障檢測(cè),特征分類;氣體絕緣金屬封閉開(kāi)關(guān);核主成分分析;支持向量機(jī)
中圖分類號(hào):TM561 文獻(xiàn)標(biāo)識(shí)碼:A 文章編號(hào):1007-3175(2016)11-0016-05
Gas Insulated Switch Fault Diagnosis Method Research Based on
Characteristic Classification Algorithm
ZHANG Zhan1, YANG Guang2, HUANG Zhi2, ZHANG Feng2, ZHANG Shi-wen2
(1 Central Southern China Electric Power Design Institute of China Power Engineering Consulting Group, Wuhan 430071,China;
2 School of Electrical Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China)
Abstract: In allusion to the fault monitoring problem of high voltage circuit breaker operating mechanism, this paper raised a kind of fault detection method for gas insulated switch (GIS) based on kernel principal component analysis (KPCA) and support vector machine (SVM). The KPCA algorithm was used to reduce dimension of eigenvalue of coil current waveform, which was input multi-classified SVM. The practical sample experiment verifies the correctness and reliability of the algorithm, and the discussion is proposed to calculate the optimal parameter.
Key words: failure detection; characteristic classification; gas insulated switch (GIS); kernel principal component analysis (KPCA); support
vector machine (SVM)
參考文獻(xiàn)
[1] 劉亞芳. 國(guó)內(nèi)外高壓S F6 斷路器運(yùn)行狀況及維修策略綜述[J]. 電力設(shè)備,2002,3(1):26-29.
[2] 李娟,焦邵華. 基于D S P 的高壓斷路器狀態(tài)在線監(jiān)測(cè)裝置[J]. 電力自動(dòng)化設(shè)備,2004,24(8):44-47.
[3] 張弛. 高壓斷路器在線監(jiān)測(cè)與故障診斷系統(tǒng)研究[D]. 北京:北京交通大學(xué),2007.
[4] 曹飛. 斷路器在線監(jiān)測(cè)數(shù)據(jù)分析的研究與應(yīng)用[D].杭州:浙江大學(xué),2008.
[5] 汪濤. 基于小波分析的斷路器動(dòng)特性在線監(jiān)測(cè)系統(tǒng)的研究與設(shè)計(jì)[D]. 西安:西安電子科技大學(xué),2010.
[6] 郭武,戴禮榮,王仁華. 采用主成分分析的特征映射[J]. 自動(dòng)化學(xué)報(bào),2008,34(8):876-879.
[7] SCH C, LAPTEV I, CAPUTO B.Recognizing Human Actions: A Local SVM Approach[C]// Proceedings of the Pattern Recognition, 17th International Conference on (ICPR'04)IEEE Computer Society,2004(3):32-36.
[8] RAKOTOMAMONJY A.Variable selection using SVM based criteria[J].Journal of Machine Learning Research,2003,3(7/8):1357-1370.
[9] CHERKASSKY V, MA Y.Practical selection of SVM parameters and noise estimation for SVM regression[J].Neural Networks the Official
Journal of the International Neural Network Society,2004,17(1):113-126.
[10] RAZI-KAZEMI A A, Vakilian M, Niayesh K, et al. Circuit-Breaker Automated Failure Tracking Based on Coil Current Signature[J].IEEE Transactions on Power Delivery,2014,29(1):283-290.
[11] NI J, ZHANG C, YANG S X.An adaptive approach based on KPCA and SVM for real-time fault diagnosis of HVCBs[J].IEEE Transactions on
Power Delivery,2011,26(3):1960-1971.