周树孝,彭黔荣,周勇,石炎福,余华瑞

• 1:四川大学化学工程学院
• 2:四川大学化学工程学院
• 3:四川大学化学工程学院
• 4:四川大学化学工程学院

摘要(Abstract)：

由于流化系统的复杂性和非线性的特性,到目前为止,仍没有一个能很好地判断这两类流态化的可靠方法,而人工神经网络能够进行复杂的逻辑操作和实现对非线性系统流型的识别。由此,在文献数据(含15种颗粒和11种流体)的基础上,利用三层BP人工神经网络,提出了一种识别散式流化和聚式流化的新方法。以经过归一化处理的3个无因次准数pff()/rrr-、mfRe和mfFr(判别因子)作为神经网络的三个输入结点,以散式流化、过渡状态和聚式流化对应于神经网络的三个输出结点,由训练样本集得到隐层结点数等最佳网络参数,然后对待测样本进行了流型识别。研究结果表明,神经网络用于散式流化和聚式流化的模式识别,结果与实际较一致。新方法优于传统的识别方法,具有较好的应用前景。

关键词(KeyWords)： 散式流态化;聚式流态化;BP神经网络;流型识别

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金;; 海外青年基金资助项目(29928005)。

作者(Author): 周树孝,彭黔荣,周勇,石炎福,余华瑞

参考文献(References)：

• [1]JIN Yong(金涌),ZHU Jing-xu(祝京旭),WANG Zhan-wen(汪展文),YU Zhi-qing(俞芷青).FluidizationEngineeringPrinciples(流态化工程原理)[M].Beijing(北京):TsinghuaUniversityPress(清华大学出版社),2001.
• [2]WilhelnR H,KwaukM.Fluidization of solid particles[J].ChemicalEngineeringProgress,1948,44(3):201-218.
• [3]HarrisonD,DavidsonJ F.On the nature of aggregative and particulate fluidization[J].Transction ofInstitute ofChemicalEngineering.1961,39(3):202-211.
• [4]GeldartD.Types of gas fluidization[J].PowderTechnology,1973,7(5):285-292.
• [5]DejongA H.Studies on the transition between particulate and aggregative fluidization in gas-solids fluidization beds[J].PowderTechnology,1974,9(2):91-102.
• [6]BroadhurstT E,BeckerH A.Onset of fluidization and slugging in beds of uniform particles[J].AIChE Journal,1975,21(2):238-247.
• [7]LIU De-jin,KaukM,LI Hong-zhong.Aggregative and particulate fluidization the two extremes of a continuous spectrum[J].ChemicalEngineeringScience,1996,51(17):4045-4063.
• [8]HarrisonD,DavidsonJ F.On the nature of aggregative and particulate fluidization[J].Transction ofInstitute ofChemicalEngineering.1961,39(3):202-211.
• [9]RemeroJ B,JohnsonL N.Factors affecting fluidized bed qualities[J].ChemicalEngineeringProgress,SymposiumSeries,1962,58:28-45.
• [10]VerloopJ,HeertjesP M.Shock waves as a criterion for the transition from homogeneous to heterogeneous fluidization[J].ChemicalEngineeringScience,1970,25(5):825-832.
• [11]FoscoloP U,GibilaroL G.Fully prediction criterion for the transition between particulate and aggregative fluidization[J].ChemicalEngineeringScience,1984,39(12):1667-1675.
• [12]DoichevK,TodorovS,DimitrovV.Transition between particulate and aggregative fluidization at different state-of-flow of the solids[J].ChemicalEngineeringScience,1975,30(4):419-424.
• [13]ZHOU Zhang-yu(周章玉),SHI Yan-fu(石炎福),YU Hua-rui(余华瑞).Discrimination between particulate and aggregative fluidization by pressure fluctuation(由压力波动判断气-固流化床中的流化类型)[J].ChemicalReactionEngineering andTechnology(化学反应工程与工艺),1999,15(3):262-267.
• [14]LetzelH M,SchoutenJ C,KrishnaR, van denBleekC M.Characterization of regimes and regime transitions in bubble columns by chaos analysis of pressure signals[J].ChemicalEngineeringScience,1997,52(24):4447-4459.
• [15]ZHAO Gui-bing(赵贵兵),GU Li-li(顾丽莉),SHI Yan-fu(石炎福)YU Hua-rui(余华瑞).Distinguishing between particulate and aggregative fluidization using chaos theory.(散式流态化到聚式流态化的混沌识别)[J].J ChemInd andEng(化工学报),2000,51(4):468-473.
• [16]YAN Ping-fan(阎平凡),ZHANG Chang-shui(张长水).ArtificalNeuralNetwork andGeneticAlgorithms(人工神经网络与模拟进化计算)[M].Beijing(北京):TsinghuaUniversityPress(清华大学出版社),2000.
• [17]LI Xiao-xiang(李晓祥),SHI Yan-fu(石炎福),HUANG Wei-xing(黄卫星),QI Xiao-bo(漆小波),YU Hua-rui(余华瑞),ZHU Jing-xu(祝京旭).Nonlinear model for generation of the solids holdup time series in theCFB riser based on combination of chaos with artificial neural network(基于混沌和人工神经网络方法建立计算气固循环流化床局部颗粒浓度时间序列的非线性模型)[J].J ChemEng ofChineseUniv(高校化学工程学报),2003,17(5):580-584.