Suzhou Electric Appliance Research Institute
期刊號(hào): CN32-1800/TM| ISSN1007-3175

Article retrieval

文章檢索

首頁(yè) >> 文章檢索 >> 文章瀏覽排名

基于ReliefF-mRMR與IAO-SVM的變壓器故障診斷

來(lái)源:電工電氣發(fā)布時(shí)間:2023-02-07 10:07 瀏覽次數(shù):454

基于ReliefF-mRMR與IAO-SVM的變壓器故障診斷

張佳豪1,楊國(guó)華1,2,趙藝青1,張兆坤1,李志遠(yuǎn)1
(1 寧夏大學(xué) 物理與電子電氣工程學(xué)院,寧夏 銀川 750021;
2 寧夏電力能源安全重點(diǎn)實(shí)驗(yàn)室,寧夏 銀川 750004)
 
    摘 要:為進(jìn)一步提高變壓器故障診斷準(zhǔn)確率,提出一種基于 ReliefF-mRMR 與 IAO-SVM 結(jié)合的變壓器故障診斷模型。采用 ReliefF 和最大相關(guān)最小冗余 (mRMR) 算法對(duì)變壓器故障數(shù)據(jù)進(jìn)行特征優(yōu)選;引入混沌反向?qū)W習(xí)和自適應(yīng)混合變異策略改進(jìn)天鷹優(yōu)化算法,并對(duì)最優(yōu)特征集合和支持向量機(jī) (SVM) 參數(shù)聯(lián)合尋優(yōu),構(gòu)建最佳故障診斷模型;利用已有變壓器故障數(shù)據(jù)對(duì)所提模型仿真實(shí)驗(yàn),并與常用故障診斷模型灰狼算法支持向量機(jī) (GWO-SVM)、天鷹優(yōu)化算法支持向量機(jī) (AO-SVM) 相比較, 準(zhǔn)確率分別提高了 10.76% 和 6.15%,高達(dá) 95.38%,結(jié)果表明所提模型能有效提高變壓器故障診斷精度。
    關(guān)鍵詞: 變壓器;故障診斷;特征優(yōu)選;改進(jìn)天鷹優(yōu)化算法;支持向量機(jī)
    中圖分類(lèi)號(hào):TM407     文獻(xiàn)標(biāo)識(shí)碼:A     文章編號(hào):1007-3175(2023)01-0001-07
 
Fault Diagnosis Method of Transformer Based on ReliefF-mRMR and IAO-SVM
 
ZHANG Jia-hao1, YANG Guo-hua1,2, ZHAO Yi-qing1, ZHANG Zhao-kun1, LI Zhi-yuan1
(1 College of Physics and Electrical and Electronic Engineering, Ningxia University, Yinchuan 750021, China;
2 Ningxia Key Laboratory of Power and Energy Security, Yinchuan 750004, China)
 
    Abstract: The paper is aimed at putting forward a transformer fault diagnosis model based on ReliefF-mRMR and IAO-SVM to further increase its accuracy. In order to build an optimal fault diagnosis model, the authors use ReliefF and mRMR for feature optimization of transformer fault data, combine chaotic backward learning and adaptive mixed mutation strategy to improve aquila optimizer, and make joint optimization of optimal feature set and support vector machine (SVM) parameters. By doing simulation experiment of the existing transformer fault data and comparing the optimal fault diagnosis model with gray wolf algorithm support vector machine (GWO-SVM) and aquila optimizer support vector machine (AO-SVM), it is found that the accuracy of the optimal fault diagnosis model rise to 95.38% with the growth rate of 10.76% and 6.15% respectively, verifying its high accuracy of transformer fault diagnosis.
    Key words: transformer; fault diagnosis; feature optimization; improved aquila optimizer; support vector machine
 
參考文獻(xiàn)
[1] 盛戈皞,錢(qián)勇,羅林根,等. 面向新型電力系統(tǒng)的電力設(shè)備運(yùn)行維護(hù)關(guān)鍵技術(shù)及其應(yīng)用展望[J] . 高電壓技術(shù),2021,47(9):3072-3084.
[2] 江秀臣,許永鵬,李曜丞,等. 新型電力系統(tǒng)背景下的輸變電數(shù)字化轉(zhuǎn)型[J] . 高電壓技術(shù),2022,48(1):1-10.
[3] 梁永亮,郭漢琮,薛永端. 基于特征氣體關(guān)聯(lián)特征的變壓器故障診斷方法[J] . 高電壓技術(shù),2019,45(2):386-392.
[4] 張丞鳴,謝菊芳,胡東,等. 基于 QPSO-SVM 與 DGA 五邊形解釋工具的變壓器故障診斷方法[J] . 高壓電器,2021,57(12):117-124.
[5] 徐龍舞,張英,張倩,等. 基于正交實(shí)驗(yàn)法改進(jìn)的蝠鲼算法優(yōu)化 BP 在變壓器故障診斷上的研究[J] .南方電網(wǎng)技術(shù),2022,16(7):46-54.
[6] WU Y H, SUN X B, ZHANG Y, et al.A Power Transformer Fault Diagnosis Method-Based Hybrid Improved Seagull Optimization Algorithm and Support Vector Machine[J].IEEE Access,2022,10:17268-17286.
[7] 劉仲民,翟玉曉,張?chǎng)?,? 基于 DBN-IFCM 的變壓器故障診斷方法[J] . 高電壓技術(shù),2020,46(12):4258-4265.
[8] YANG X H, CHEN W K, LI A Y, et al.BAPNN-based methods for power transformer fault diagnosis[J].Advanced Engineering Informatics,2019,39:178-185.
[9] 謝國(guó)民,倪樂(lè)水,曹媛. 基于 VSRP 與 β-GWO-SVM 的變壓器故障辨識(shí)方法[J] . 高電壓技術(shù),2021,47(10):3635-3641.
[10] 蘇磊,陳璐,徐鵬,等. 基于深度信念網(wǎng)絡(luò)的變壓器運(yùn)行狀態(tài)分析[J] . 高壓電器,2021,57(2):56-62.
[11] 廖偉涵,郭創(chuàng)新,金宇,等. 基于四階段預(yù)處理與 GBDT 的油浸式變壓器故障診斷方法[J]. 電網(wǎng)技術(shù),2019,43(6):2195-2203.
[12] HUANG Z K,YANG C H,ZHOU X J, et al.A Hybrid Feature Selection Method Based on Binary State Transition Algorithm and ReliefF[J].IEEE Journal of Biomedical and Health Informatics,2019,23(5) :1888-1898.
[13] PENG H C, LONG F H, DING C.Feature selection based on mutual information criteria of max-dependency , max-relevance , and min-redundancy[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2005,27(8) :1226-1238.
[14] LAITH A, DALIA Y, MOHAMED A E, et al.Aquila Optimizer : A Novel Meta-Heuristic Optimization Algorithm[J] . Computers & Industrial Engineering,2021,157:107250.
[15] 王雨虹,王志中,付華,等. 多策略改進(jìn)麻雀算法與 BiLSTM 的變壓器故障診斷研究[J]. 儀器儀表學(xué)報(bào),2022,43(3):87-97.
[16] TANG A D , TANG S Q , HAN T , et al . A Modified Slime Mould Algorithm for Global Optimization[J].Computational Intelligence and Neuroscience,2021,2021:2298215.
[17] 單亞峰,段金鳳,付華,等. 基于 SSA-AdaBoost-SVM 的變壓器故障診斷[J] . 控制工程,2022,29(2):280-286.
[18] 尹金良. 基于相關(guān)向量機(jī)的油浸式電力變壓器故障診斷方法研究[D]. 北京:華北電力大學(xué),2013.
[19] 李春茂,周妺末,劉亞婕,等. 基于鄰域粗糙集與多核支持向量機(jī)的變壓器多級(jí)故障診斷[J] . 高電壓技術(shù),2018,44(11):3474-3482.