基于递归分析的喷雾气固流化床团聚状态识别  

Recursive analysis and agglomerate state recognition of spray gas-solid fluidized bed

作　　者：周云龙[1] 卢志叶 王猛 ZHOU Yunlong;LU Zhiye;WANG Meng(School of Energy and Power Engineering,Northeast Electric Power University,Jilin 132012,Jilin,China;CCCC Gas&Heat Research and Design Institute Limited Company,Shenyang 110000,Liaoning,China)

机构地区：[1]东北电力大学能源与动力工程学院,吉林省吉林市132012 [2]中交煤气热力研究设计院有限公司,辽宁沈阳110000

出　　处：《化工学报》2018年第9期3835-3842,共8页CIESC Journal

基　　金：国家自然科学基金项目(51276033)~~

摘　　要：液体会影响喷雾气固流化中的颗粒循环模式,形成不同程度的颗粒团聚物,从而直接降低反应器的传热、传质效果。实验中向流化床喷入不同黏度的乙二醇溶液,来人工制造团聚现象。研究结果表明,喷入不同黏度的液体后,流化床内会形成不同的团聚结构,按照团聚物大小可分为四类:微团聚、成核团聚、黏结团聚以及糊状团聚。通过对不同工况中的压差波动信号进行递归分析,发现不同团聚状态下的递归图纹理结构有着明显的差异。与此同时,对50种流动条件下的递归特征量分布情况进行分析,以此来识别不同的团聚结构,整体识别率高达96.39%。结果表明:通过对压差信号的递归分析可以快速识别流化床中的团聚状态。Liquid affects particle circulation in spray gas-solid fluidization by forming different extent of particle agglomerates and directly reduce heat and mass transfer of the reactor.Ethylene glycol solutions of various viscosities were sprayed onto fluidized bed to artificially create agglomerates.The results showed that liquids of different viscosities would form different agglomerates in fluidized bed,which can be divided into four groups by agglomerate size,including microaggregation,nucleation agglomerates,coherence agglomerates,and paste agglomerates.Recursive analysis of differential pressure fluctuation signal in flow process indicated some obvious differences in recursive pattern structure under different agglomeration conditions.Further analysis finding on distribution of recursive feature indices of 50 flow conditions was used to identify agglomeration structures which 96.39%overall recognition rate was achieved.Therefore,agglomerated state in fluidized bed was quickly identified by recursive analysis of differential pressure signal.

关 键 词：气固流化床 团聚 液体黏度 递归分析 

分 类 号：O359.2[理学—流体力学]