資源描述:
《基于小波包與改進(jìn)的pso-pnn變壓器勵(lì)磁涌流識(shí)別算法研究》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在工程資料-天天文庫。
1、基于小波包與改進(jìn)的PSO-PNN變壓器勵(lì)磁涌流識(shí)別算法研究*公茂法1,接怡冰1,李美蓉2,解云興3,宋健3、吳娜1(1.山東科技大學(xué)電氣與自動(dòng)化工程學(xué)院,山東青島266590;2.國(guó)網(wǎng)山東省電力公司棗莊供電公司,山東棗莊277100;3.國(guó)網(wǎng)山東省電力公司東營(yíng)供電公司、東營(yíng)方大電力設(shè)計(jì)規(guī)劃有限公司,山東東營(yíng)257091)摘要:利用小波包對(duì)勵(lì)磁涌流和故障電流信號(hào)進(jìn)行分解并提取小波包能量特征。采用改進(jìn)粒子群(PSO)算法訓(xùn)練概率神經(jīng)網(wǎng)絡(luò)(PNN)尋找全局最優(yōu),對(duì)PNN網(wǎng)絡(luò)的輸入輸出、傳遞函數(shù)以及隱含層節(jié)點(diǎn)數(shù)進(jìn)行確定,建立PNN的網(wǎng)絡(luò)模型,對(duì)網(wǎng)絡(luò)進(jìn)行訓(xùn)練測(cè)試,最后提出保護(hù)判據(jù)。研究發(fā)現(xiàn),該算法不僅訓(xùn)
2、練速度和收斂速度快,而且具有較高的識(shí)別精度。關(guān)鍵詞:勵(lì)磁涌流;小波包能量;粒子群;概率神經(jīng)網(wǎng)絡(luò)中圖分類號(hào):TM761文獻(xiàn)標(biāo)識(shí)碼:A文章編號(hào):1001-1390(2018)08-0000-00ResearchonalgorithmoftransformerinrushcurrentidentificationbasedonwaveletpacketandimprovedPSO-PNNGongMaofa1,JieYibing1,LiMeirong2,XieYunxing3,SongJian3,WuNa1(1.CollegeofElectricalEngineeringandAutomation,S
3、handongUniversityofScienceandTechnology,Qingdao266590,Shandong,China.China.2.ZaozhuangPowerSupplyCompany,StateGridShandongElectricPowerCompany,Zaozhuang277100,Shandong,China.China.3.DongyingFangdaElectricPowerDesign&PlanningCompany,DongyingPowerSupplyCompany,StateofStateGridShandongElectricPowerComp
4、any、,DongyingFangdaElectricPowerDesign&planningcompany,Dongying257091,Shandong,China)Abstract:Thealgorithmapplieswaveletpackettodecomposeexcitationinrushcurrentandfaultcurrentsignalsandthenextracttheenergycharacteristicsofwaveletpacket.Firstly,itusesadoptstheimprovedPSO(ParticleparticleSwarmswarmOpt
5、imizationoptimization)algorithmtotrainPNN(ProbabilisticprobabilisticNeuralneuralNetworknetwork)todeterminetheinputandoutput,thetransferfunctionaswellasthehiddenlayernodesofthePNNnetwork.Then,itestablishesanetworkmodelofPNNtotrainandtestthenetwork.Finally,theprotectioncriterionisproposed.Thestudyfoun
6、dthatnotonlydoesthealgorithmhavefasttrainingspeedandconvergencespeed,butalsohighrecognitionaccuracy.Keywords:inrushcurrent,waveletpacketenergy,PSO,PNN0引言*基金項(xiàng)目:國(guó)家自然科學(xué)基金資助項(xiàng)目(61503224);山東省高等學(xué)??萍加?jì)劃項(xiàng)目(J17KA074)在以特高壓電網(wǎng)為骨干網(wǎng)架的智能電網(wǎng)建設(shè)過程中,電壓等級(jí)逐步提高,用戶用電量越來越高,電力系統(tǒng)規(guī)模逐漸擴(kuò)大[1]。變壓器是電力系統(tǒng)中的核心部分,相當(dāng)于“心臟”,它的運(yùn)行狀態(tài)關(guān)乎著整個(gè)電力系統(tǒng)的
7、安全。一旦出現(xiàn)故障,將會(huì)立即導(dǎo)致整個(gè)區(qū)域內(nèi)的電力網(wǎng)癱瘓。因此,對(duì)變壓器進(jìn)行定期的檢查、維修,能夠及時(shí)彌補(bǔ)設(shè)備缺陷、消除隱患,這對(duì)變壓器來說是極為重要的[2]。在現(xiàn)場(chǎng)實(shí)際生產(chǎn)中,電流縱聯(lián)差動(dòng)保護(hù)的原理簡(jiǎn)單,且具備快速切除各類故障等特點(diǎn),因而逐漸成為了變壓器首選的主保護(hù)之一。將差動(dòng)保護(hù)投入到變壓器中,并不能完全可靠地保護(hù)變壓器,有可能產(chǎn)生誤動(dòng),導(dǎo)致其誤動(dòng)的關(guān)鍵性因素就是變壓器鐵芯飽和引起的勵(lì)磁涌流[3