热工大迟延过程的内模控制算法及应用研究

《热工大迟延过程的内模控制算法及应用研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、华北电力大学（保定）博士学位论文热工大迟延过程的内模控制算法及应用研究姓名：金秀章申请学位级别：博士专业：热能工程指导教师：刘吉臻20051201华北电力大学博士学位论文型的摄动，基于自适应多内模切换控制策略设计了主蒸汽温度自动控制系统，并根据主汽温系统的特点设计了燃烧前馈、机组负荷前馈以及磨煤机启停前馈，燃烧前馈采用辐射能作为中间信号，不仅考虑了燃料量变化，而且考虑了风量的变化。仿真结果表明了所设计控制系统优良的控制性能。最后是全文的总结与展望。关键词：大迟延，热工过程，内模控制，自适应，多模型控制，神经网络，过热汽温ABSTRA

2、CTThePIDcontrolstrategyisusuallyadoptedinthecontrolsystemsinpowerplantatpresent．However，theprocessofthepowerplantisverycomplex，whichinvolvesthetransmissionofthesubstanceinthetubes，combustionofthecoalinthefumaceandheatexchange．Furthermore，thepoweroftheunitischangedwitht

3、hechangeofneededloadfrequently．So，thecombustionprocessisakindofnonlinearsystemwithstrongcoupling，moreconditionsandsignificantdelay．Theneedofthecontrolsystemofcombustionprocesscannotbesatisfiedbythenormalcontrolstrategies．Fortheadvantagesofsimpletodesign，excellentrobust

4、ness，andbetterpredictionability，theinternalmodelcontrol(IMC)showsthestrongvigortosolvethecontrolproblemofthethermalprocesswithlargetimedelay．However，thedynamiccharacteristicsofcontrolledprocessesinpowerplantaredifferentwitheachother．BasedonthestatusofIMC，wehavestudieds

5、ystematicallyinthisdissertationtheIMCmethodanditsrelatedtheoriesfortheprocesseswithuncertainparametersandlargetimedelay．Ourstudyaddressesthefollowingtopics：Firstly，theIMCstrategyforthestablelinearprocessisintroduced．Wesummarizethestructure，themethodtodesignandpropertie

6、sofIMC．Todesignthecontrolsystemabilitiestotrackthesetpointandrejectexternaldisturbancesrespectively，thetwo—freedomIMCisadopted．TheresearchandsimulationresultsshowedthatsomeexternaldisturbancesoftheunstablecontrolledprocessescannotberejectedbyclassicalIMC．Sotheinternalf

7、eedbackcontrollerisusedtostabilizetheunstableprocess，andthentheIMCsystemcanbedesigned．However，thiscontrolsystemcannotrejecttheoffsetintroducedbysomeexternaldisturbances，aPIDcompensationcontrollerisdesignedinthecontrolsystem．AkindofIMCcontrollerisdesignedforthestablepro

8、cesswithlargetimedelay，andthepermittedvariationrangesofcontrolledprocessparametersaregiven．TheadaptiveIMCcontrolleris