安娜,潘立登,陈标华,牛学坤,李成岳

• 1:可控化学反应科学与技术基础教育部重点实验室
• 2:北京化工大学
• 3:信息科学与技术学院
• 4:可控化学反应科学与技术基础教育部重点实验室

摘要(Abstract)：

清除工业废气中低浓度挥发性有机物(VOCS) 的流向变换催化燃烧反应器的床层温度依时变化,为了实现实时预测和控制,用动态RBF(Radial Basis Function)神经网络建立了反应器床层瞬态温度分布的预测模型。着重讨论了动态RBF神经网络的基本结构,依据RBF网络线性输出的特点,给出了预测模型参数的在线修正方法。仿真结果与中试装置现场数据的对照表明,所建立的模型简单、精度高,能满足控制要求。

关键词(KeyWords)： 挥发性有机物;流向变换;催化燃烧;动态系统;RBF神经网络;在线参数校正

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金资助项目(20076002)。

作者(Author): 安娜,潘立登,陈标华,牛学坤,李成岳

参考文献(References)：

• [1]AN Na(安娜),AppliedBasicResearch of theReversed-flowCatalyticCombustionProcess and itsControl(流向变换催化燃烧及其控制技术的应用基础研究)[D].Beijing(北京):University ofChemicalTechnology(北京化工大学).2003.
• [2]CittadiniM,VanniM,BarresiA A.Transient behaviour and start-up of periodical flow reversal reactors for catalyticdecontamination of waste gases[J].ChemEng andProcessing,2002,41(5):437-443.
• [3]MarcoCittadini,MarcoVanni,AntonelloA Barresi,GiancarloBaldi.Reverse-flow catalytic burners: response to periodical variations in the feed[J].ChemicalEngineeringScience,2001,56:1443-1449.
• [4]BudmanH,KzyonsekM,SilvestonP.Control of a nonadiabatic packed bed reactor under periodic flow reversal[J].TheCanadianJournal ofChemicalEngineering,1996,74(10):751-759.
• [5]MA Xiao-min(马晓敏),ZHOU Mang-lai(周忙来).Nonlinear dynamic system identification based on the neural networks(基于神经网络的非线性动态系统辨识)[A].ControlConferenceDissertation ofChina(中国控制会议论文集)[C],Hunan(湖南):HunanPress(湖南出版社),1995:708-712.
• [6]NarendraK S, et al.Identification and control for dynamic systems using neural networks[J].IEEE,TransNN,1990,1(3):4-27.
• [7]WANG Yong-ji(王永骥),TU Jian(涂健).NeuralNetworkControl(神经元网络控制)[M].Beijing(北京):Mechanism andIndustryPress(机械工业出版社),1998.
• [8]ZHANG Guo-ping(张国萍),CAO Liu-lin(曹柳林).A solution to the measurement lag in prediction of viscosity using dynamic neural network(用动态神经网络解决粘度预测的方法)[J].Joural ofBeijingUniversity ofChemicalTechnology(北京化工大学学报),1999,26(3):54-56.
• [9]JieZhang.Developing robust neural network models by using both dynamic and static process operating data[J].IndEngChemRes,2001,40:234-241.
• [10]XU Ying(徐英),YANG Er-fu(杨尔辅),XU Yong-mao(徐用懋).An adaptive approach to on-line identification of dynamic system based onRBF neural networks(一类基于RBF神经网络的动态系统在线自适应辨识方法)[J].Information andControl(信息与控制),2001,30(6):508-512.
• [11]WANG Hua-zhong(王华忠),HUA Xiang-ming(华向明).Inferential control of a fixed-bed reactor usingRBF network(固定床反应器基于RBF网络的推断控制)[J].J ChemInd andEng(化工学报),1995,46(5):631-634.
• [12]WU Hui-xiong,ZHANG Shu-zeng,LI Cheng-yue.Study of unsteady-state catalytic oxidation of sulfur dioxide by periodic flow reversal[J].TheCanadianJournal ofChemicalEngineering,1996,74(10):766-771.
• [13]WU Peng(吴鹏),LI Shao-fen(李绍芬),LIAO Hui(廖晖).The criterion for runaway of the n-th order reaction in fixed bed reactor(固定床反应器进行n级反应的飞温判据)[J].J ChemEng ofChineseUniv(高校化学工程学报),1995,8(2):163-167.
• [14]PAN Li-deng(潘立登),MA Jun-ying(马俊英).SystemModeling andSoftwareSensorsTechnique(系统模型化与软测量技术)[Z].BeijingUniversity ofChemicalTechnologyPrintedTeachingMaterials(北京化工大学讲义),2000.