人工神經(jīng)網(wǎng)絡(luò)在電廠汽溫系統(tǒng)中的研究與應(yīng)用.pdf

1、Y79“12盔j二二二毳j簍芝東南碩士掌大學(xué)位論文人工神經(jīng)網(wǎng)絡(luò)在電廠汽溫系統(tǒng)中的研究與應(yīng)用研咒7r姓“：龔室劇甘帥姓私：工絲咀副熬攫巾請世目螻L疆土二一掌枷翱f嶄稱越藍(lán)I止氈土出遺監(jiān)盟，、∥，“論女挺2L唧—墅匣鹽二＆i斟引孵——型型i一^一～’串忸授P啦n—盤直灶做授fⅫ——靜卅盛。l會k譫二—一wⅫ^生：疊二墜一l‘11H、東南大學(xué)碩士學(xué)位論文AbstractIfthesuper—heatedsteamtemperatureistO

2、Ohighortoolow，itwillbringondisadvantageousfactorsattheprocessofthermodynamicengineeringinpowerplantThesuperheatedsteamtemperatureoftheoutputofthesuperheatedimplementmustbecontrolledtoanexpectedrangeSomedisturbancefactors

3、invaportemperatureaIeanalyzedonthebasisofproducetechniquesinthepowerplantandthevaportemperaturemathmodelforsimulationisbuiltbyuseofmechanismanalyzingTheneuronplusinfluenceoncontrolperformanceisanalyzedonthefoundationofth

4、ebasicneuroncontrolalgorithmwhichbasedonneuroncontroltheoryAselfadjustingplussingleneuralcontrollerisdesignedbyusingoftheerrorbetweenactualoutputandexpectedoutputanditsradioofitschangingThecontrolprojectaboutthesuperheat

5、edsteamsystemisdevisedTheseriescontrolsystemisappliedtotheproject，andselfadjustingplussingleneuralcontrollerisadoptedasmaincontrollerandconventionalPIcontrollerisadoptedasininorcontrollerThenewcontrollerissuperiortotheco

6、nventionalPIDcontrollerwhichprovedbyemulationalexperimentTheattenuationfunctionneuralcontroUerisinvestigatedinordertoimprovetheslowresponsevelocityofselfadjustingplussingleneuralcontrollerSomeparameterssuchasneuronplus，l

7、earningvelocityandinitialweightofcontrolleraleanalyzedinordertocoachpeoplehowtoimprovecontrollerperformanceandinitializeparametersTheneuralnetworkscompensatorisdesignedtodealwiththeadverseeffectsproducedbythelargetimedel

8、aybasedonimprovedStrlithpredictorschemebyHang，anditisusedforidentifythedelaytimeDigitalemulationalresultsshowthatthenewmethodCallcontrolthetimedelayplantswithoutknowingtheirmathematicalmodelsThenewmethodcanalsoautomatica