宝钢分公司电力负荷模拟与预测研究

上海交通大学硕士学位论文宝钢分公司电力负荷模拟与预测研究姓名：王小辉申请学位级别：硕士专业：电气工程指导教师：王西田;李关定20081228I1EMS234IIABSTRACTIIIRESEARCHONTHEPOWERLOADSIMULATIONANDFORECASTINGOFBAOSTEELBRANCHABSTRACTWiththedeeplyreformingoftheelectricitymarketandthecontinuousimprovementoftheironandsteelenterpriseenergymanagement,loadforecastingoftheironandsteelenterpriseshasbecomeahottopicresearch.Becauseofthespecialloadcharacteristics,themainloadfluctuationsoftheironandsteelenterprisearerelatedwiththeproductionprocess.Therefore,theappropriateloadmodelsandtheirsimulationalgorithmstogetherwiththeproductionplaninformationshouldbetakenintoaccounttoimprovetheloadforecastingaccuracy.InordertoimproveBaosteelenergymanagement,aco-operativeresearchproject“BaosteelBranchpowergridloadmanagementandforecastingsystem”wassponsored.Inthecontextoftheproject,thesteelloadsimulationandthemethodofloadforecastingwerefurtherstudiedinthisthesis.Themainworkislistedasfollows:1ThenecessarydataandinformationwerecollatedandanalyzedbytheuseoftheBaosteelBranchEMSDataSystems.Therelationsbetweenthesteelproductionprocess,maintenancescheduling,powersupplylinesandtheenterpriseloadweresummarized.2TwotypesofloadsimulationmethodswereproposedonaccountofthepracticalloadcharacteristicsoftheBaosteelBranch.Theoneissimulatingdirectlybasedonthehistoryloaddata.Theotherisresolvingthetotalload,simulatingseparately,andthenaddingtheloadmodelstogettheloadcurve.ABSTRACTIV3Usingtheconventionalloadforecastingalgorithms,comparingofvariousalgorithmserrorandthenchoosingthebestalgorithmtocompletethedesignofBaosteelBranchloadforecastingsystem.4Themethodofcorrectionthesteelloadforecastingresultwasproposedbyuseofmaintenanceschedulingtable.Theloadcorrectionparametersofmajormaintenanceschedulesweredeterminedbythedetectionmethod.Whensomemaintenancescheduleswereoccurred,theparameterswereintroducedtoimprovetheaccuracyoftheloadforecasting.Thisthesisisclosewiththepracticalproductionprocess.ItisadapttonotonlyBaosteelBranch,butalsotheotherbranchesoftheBaosteelGroup.Theproposedcorrectionmethodoftheloadforecastingandtheloadsimulationmodelcanbeveryusefulattheloadstudyoftheironandsteelenterprises.KEYWORDSironandsteelenterprisesloadforecastingsimulationmodelmaintenancescheduling1111.1[1][2]1.21.2.11998111719781600900MW---1300MW1.2.2[3]EMS(AGC)12288[4][5]1.31.3.1[1][4]1-1[6](1)[7][8][9]Delphi131-1Table1-1Mainmethodscomparisonofpowernetworkloadforecast14ANNPAREW-MVR(2)[10]s(3)[11][11](4)[12]15Box.Jenkins[13](5)[14,15]:[16][17](DFT),,T7dT≤7dGM(1,1)[18],(6)[19][20][21]Boosting.Boosting(7)[22,23][24]161.3.2[25]48[26][27][28][29]2881.411995EMSS5PLC20S1720081~200835[30]22883[31][23][32]3[33,34,35]4282[36][37][38]2.112342-12~30KW100MW2-2292-1200719Fig.2-1PowerdailyconsumptionfromJanuarytoSeptember05010015020025030075080085090095010002-2Fig.2-2ThetotalloadwaveofBaosteel52102.22.2.12-3205020301580155014202-3Fig.2-3ProductionProcesschartofBaosteel2-42.2.22-51-12112-4Fig.2-4Thechartofironandsteelproductionprocess2-5Fig.2-5ThenetworkofBaosteelmainsubstationsandgeneratorsG0G3G1G22123220KV110KV10KV33KVS72.32-62-6Fig.2-6Thechartofironandsteelproductioninformationanalysis2.3.114141-11-21-3…2-12-22-2…N-1N-2N-3…12……N123……nn+1n+2…J123……M213281112345678910112-7200611190%18%22%1%1%11%2%13%2%2%1%27%2-7Fig.2-7Percentagechartofvariousproductionplants2-8742141%1%1%2%3%2%2%1%2%3%6%4%2%3%1%1%5%1%1%6%5%2%5%3%5%0%2%0%0%1%0%0%12%14%0%2%1%ERWCDQ2-8Fig.2-8Percentagechartofeachunit2.3.22.3.12-81215A.2-92-10„670V105KA99MVA80MW15040~45223190/„10„2-9Fig.2-9ProductionprocessofElectricfurnaceABLFVD33kV2162-10Fig.2-10ProductionprocesslogisticmapofElectricfurnaceB110KV133033KV10132-112172-11Fig.2-11PowersupplysystemofElectricfurnaceC2-122-132-14200391940-452-121Fig.2-12Electricsmeltingprocesscurve12182-132Fig.2-13Electricsmeltingprocesscurve22-143Fig.2-14Electricsmeltingprocesscurve3215802050A2-1515801.51~2.7mm,700~1430mm26.5t84%16%3mm77%1580288219R13R23~535~60mm1.5~12.7*700~1430mmS/PR1/R22-15Fig.2-15Hot-rolledprocessflowchart3165-175SPR1R22000A20B2-162-16Fig.2-16PowersupplysystemofHot-rolledplant110KV1416141714181580R1R2220314201550A2-172-17Fig.2-17Cold-rolledprocessflowchartCDCMCAPLCPLETLSTLSHL14203-513.772.284032.281550142010---1314201-21060400-500A155011-/155014202211420142015501081A155014201550155041MWB2-18110KV1419142014212-18Fig.2-18PowersupplysystemofCold-rolledplant2.3.31-22222.432333.1[39][40]2883.23-13-1Fig.3-1ThedirectSimulationofthetotalload3-23-1324050100150200250300820840860880900920940960980100010205/3-2Fig.3-2Dailyloadcurvebysplicingrandom3-1Table3-1Historicaldatatableforsplicingdailyloadcurve2008102401200811201162008103116202008111320241.2.3253.3.32008.1~2008.31288/95%2008.1.15-2008.2.143-33-43-20100020003000400050006000700080009000300400500600700800900100011001200MWr-b-;g-*2;y-*3;k-*53-32008.1.15-2008.2.14Fig.3-3Thecomparativeloadmapbyremovingeachunit2008.1.15-2008.2.14326051015202530354045500.20.30.40.50.60.70.80.91y-;b-;g-*2;r-*3;k-*53-42008.1.15-2008.2.14Fig.3-4Thecomparativechartofcorrelationcoefficient3-22008.1.15-2008.2.14Table3-2Statisticsbyremoving