遺傳算法優(yōu)化反應(yīng)過程的研究.pdf

1、青島化工學(xué)院碩士學(xué)位論文遺傳算法優(yōu)化反應(yīng)過程的研究姓名：安維中申請學(xué)位級別：碩士專業(yè)：化學(xué)工程指導(dǎo)教師：胡仰棟2000.6.1|t_、～，●些堡塑婆垡些堡生婪塑塑型窒————————塑堡L———一OPTIMIZATIONOFREACTIVEPROCESSUSINGGENETICALGORITHMGenedcAlgodthm(GA)isanewmethodofoptimization，whosemechanismiSbasedonthen

2、aturalselectionprocessinbiologicalevolutionGAfindstheglobalsolutionstepbystepfromitsinitialcandidatesolutionsthroughselectioncrossoverandmutationandSOonNowGAhasbeenusedinmanyfieldsanditsapplicationshavebeenenlargingincre

3、asinglyInthispaperweintroducesystemicallysomefundamentalaspectsoftheGA。includingi蝕theories，methods，implementationandmodificationWewritethecomputerprogramofGAusingBorlandClanguage，andsetuptwomodelsnamedSGAandMGAWbalsoexal

4、flsomemathfunctionsforourmodelsTIliSpaperpresentsanewmethodsbasedonGAincalculationofmulticomponentandmultiphaseequilibriumwhichiScalled懿theMinimizationofGibbsFreeEnergyOurworksarcmainlyasfollow：Firstwepresentanewmethodof

5、parametercodinginGAThedifferenceofourmethodiSthattherangeofparametercodingiSnotinvariablebutchangedwhichiSusedtosolvetherestrictiOIlSinthisproblemWealsopresentvariousparametercodingmethodsinphaseandchemicalreactionequili

6、bfittm，andsimultaneouschemicalandphaseequilibriumSecond，weimprovetheconvergencemethodbymeansofreducingorenlargingtherangeofparametercodingafterthefirstconvergencewhichenhancestherobnsmessofconvergencetoglobaloptimization

7、ThirdwepresentamethodtogetherwiththeKvaluetosolvetheproblemoftracecompositionsinequilibriumcalculationandtheprecisionofGA’Wbsetupageneralcalculatingmodelforthisproblem，andsomeexampleshavebeenOventocomparewithouralgorithm

8、Attheend，thispaperstudiestheproblemofoptimizationanddesigninreactionnetworkusingGAAmethodcombiningGAwithothersimulatingprogramsinreactionnetworkispresented，whichshowsagreatrobustnessthroughthestudiesofsomeexamples惑，，，。、。