催化重整流程模拟及优化研究

《催化重整流程模拟及优化研究》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、浙江大学博士学位论文催化重整流程模拟及优化研究姓名：胡永有申请学位级别：博士专业：控制科学与工程指导教师：褚健;苏宏业20040201浙江大学博士学位论文6．在MATLAB平台上开发了基于图形用户界面的专用重整模拟优化软件，利用一套半再生重整工业装置和两套连续重整工业装置数据对建立的过程模型进行了验证，表明了模型的有效性和应用方便性。利用模型对重整过程主要操作变量进行了分析，说明了模型在过程分析和操作指导上的重要意义。最后，在总结全文研究工作的基础上，指出了重整流程模拟优化技术在开发和应用中需要进一步解决的问题。

2、关键词：催化重整，动力学模型，流程模拟，过程优化，动态建模，参数估计，BFGS，sOP，数值算法，MATLABII浙江大学博士学位论文ABSTRACTTheprocesssimulationandoptimizationtechnologyiswidelyappliedtotheprocessindustryandbringslargeprofitsforrelatedplantsandcompanies．EspeciallythespecialsimulationandoptimizationSORWILI'e

3、oncomplexcatalyticreactionprocessismoreimportantandpursuedbythecorporationsandinstitutesinvolved．Inthisdissertation,thereactionkinetics，modelingandoptimizationofnaphthacatalyticreformingprocess，whichisaveryimportantprocessinhydrocarbonprocessingindustryforoct

4、aneimprovementorassourcesofaromaticsandhydrogen，isstudied．Theobjectiveofthisdissertationistodevelopthesimulationandoptimizationsoftwarewithgoodfeaturesofeasyusingandcompletefunctionsandtoapplytocommercialnaphthacatalyticreformingprocesses．Inthisdissertation，t

5、hemaincon廿ibutionsaxeasfollows．1．Thekineticsofnaphthacatalyticreformingreactionsarestudiedandakinetics-basedreactionmodelwithonly17lumpsconsistingofPNA(Paraffins，NaphthenesandAromatics)and17reactionsisdevelopedaccordingtothecharacterandapplyingconditionsofdom

6、esticindustrialreformers．Thelumpingschemeandreactionnetworkofthismodellargelydecreasethecomplexityofmodelanddemandfordetailedcomponentanalysis，SOitiseasytoapplythismodeltothecommercialunits．Thecatalystcokingmodelwimaverageactivationftmctionofcatalystispresent

7、ed．Theheattransferringmodeloffurnaceandgas—liquidbalancingseparationmodelarealsodescribedhere．2．Thecalculationmethodsofphysicalandchemicalpropertiesoflumpsarepresentedandthemodelkineticparametersarebenchmarkedwithplantdata．Theprocedureofparameterestimationisd

8、onebyminimizationofthesuruofthesquaresofthedeviationsbetweenplantandpredictedvaluesofkeyvariables．InordertOdecreasethedifficultyofparametersestimation，anewparameterestimationmethodispropo