基于專家和預(yù)測補償控制策略的連續(xù)碳酸化分解過程控制系統(tǒng).pdf

1、中南大學(xué)碩士學(xué)位論文基于專家和預(yù)測補償控制策略的連續(xù)碳酸化分解過程控制系統(tǒng)姓名：彭勃申請學(xué)位級別：碩士專業(yè)：控制理論與控制工程指導(dǎo)教師：桂衛(wèi)華20060509中南大學(xué)碩士學(xué)位論文摘要ABSTRACTIntheprocessofaluminaproductionbysintering，itisoneofthekeyprocessestoproduceal(OIIhusingthemethodofcontinualcarbonationde

2、composition(CCD)ofsodiumaluminatessolutions，inwhichtheresolutionratiogradientandthelastresolutionratiodirectlyaffecttheoutputandqualityofproductItissuchalongtime—delayandcomplexindustrialprocesswhichexitsmuchuncertaintya

3、ndistoocomplicatedtodescribe、)lrimmathematicalmodels，thatitCannotbecontrolledproperlybytraditionalmethodsInthispaperaimedtocontroloptimalresolutionratiogradientandthelastdecompositionratio，thecontrolsystemforCCDprocessis

4、exploited，whichintegratesexpertcontrolwithpredictivecompensationcontrolstrategyThemechanictheoryandcharacteristicofCCDprocesswereintroducedandthefactorstoquai姆andquantityofAl(Oh％werediscussedBasedontheprincipleknowledgea

5、ndexperts’experienceofCCDprocess，therulebasesofresolutionratiocontrolandsafe哆monitoringweredesignedusingproductionrulemethodTheinferentialoutputsofallrulebaseswereintegratedtoobtaintheinformationofalarmandthecontrolvalue

6、softhevalveswhichusedthemethodofaheadsearchingConsideredthepropertyoflongtimedelayintheproductionapredictivecompensationstrategyinthecontrolofthelastresolutionratioWasproposedonthebasisoftheexpertcontrolmodelforCCDproces

7、sBasedonBPneuralnetworkapredictivemodelofthelastresolutionratioWasdesignedwhichWasperiodicallymodifiedonlineAccordingtotheerrorbetweentargetandprediction，thecontrolvaluewhichwouldcompensatetheoutputofexpertcontrolmodeIWa

8、scalculatedwhichusedanexpertoptimizationalgorithmThus，thesystem’ScapabilityofpredictingandantijammingisimprovedTheinfluenceoflongtimedelayisconqueredeffectivelyandtheprecisioninthecontrolofthelastresolutionratioisalsoimp