欢迎来到天天文库
浏览记录
ID:31981895
大小:10.97 MB
页数:105页
时间:2019-01-30
《基于条件建模的聚合反应过程动态优化-研究》由会员上传分享,免费在线阅读,更多相关内容在教育资源-天天文库。
1、浙江大学硕士学位论文温操作切换的回溯模拟及产品设计动态优化,验证了方法的可行性。关键词:自由基聚合;条件建模;动态优化;灵敏度分析;互补约束AbstractPolymerisachemicalcompoundormixtureofcompoundsconsistingofrepeatingstructuralunitscreatedthroughaprocessofpolymerization.Withintherangeofwholeconversion,Typicalfreeradicalpolymerizationshavemanycomplicatedphenom
2、enasuchasgeleffect,glasseffectandcageeffectseverelyinfluencesomekeyphysicalquantities(suchasviscosity,diffusioncoefficientsandreactionkineticcoefficients),whichresultsinconditionalmodelingproblemsinpolymerizationprocess.Conditionalmodelbasedondifferential-algebraicequations(DAE)includesno
3、n-continuousgradients,causinghugedifficultyofsolvingdynamicoptimizationproblem.Thus,itmakesagreatpartofpolymerizationprocessthathowtosolvethisdynamicoptimizationproblem.Inthispaper,conditionalmodelingoffreeradicalpolymerizationsisf'wstanalyzedthoroughly,bothsequentialandsimultaneousdynami
4、coptimizationstrategyarepresentedtosolvetheproblem.Sequentialoptimizationalgorithmincludessensitivityanalysisandmodelsmoothingstrategy,whilesimultaneousapproachisdiscretizedbasedonDAEmodelandMPCCalgorithm.Themainresearchcontentsandcontributionsofthispaperareasfollows:1.Mechanismoffreeradi
5、calpolymerizationisfirststudiedindetail,andconditionalmodelingproblemcausedbythethreeeffectsisanalyzedconsequently.Withagoodunderstandingofthecauseandeffect,asummaryofthedifficultiesontraditionalwaystosolvetheproblemisproposed.2.Sequentialdynamicstrategybasedonsensitivityanalysisandmodels
6、moothingmethodisestablished.Sensitivityanalysisalgorithmcanimprovetheaccuracyofgradients,andmodelsmoothingcontributestoproblemsolving.:Basedonwhich,adynamicproblemofproductsdesigninfreeradicalpolymerizationprocessisdevelopedandsolvedwell.Abetterresultisobtainedthanthatwithtraditionalway.B
7、esides,theefficiencyofoptimizationsolvingisimproved.3.TheoriginalconditionalDAEmodelisreformulatedtoalarge-scalenonlinearprogramming(NLP)probleminsimultaneousapproachafterdiscretization.Withcomplementarityconstraintsstrategyandmodelreformulation,MPCCalgorithmisfocus
此文档下载收益归作者所有