基于图像处理的循环流化床团聚物体积分数及其容积份额  

Cluster Density and Fraction in a Circulating Fluidized Bed Based on Image Processing

作　　者：殷上轶[1,2] 钟文琪[1] 卢平[2] 宋涛 陈宇航 Yin Shangyi;Zhong Wenqi;Lu Ping;Song Tao;Chen Yuhang(School of Energy and Environment Engineering,Southeast University,Nanjing 210096,China;School of Energy and Mechanical Engineering,Nanjing Normal University,Nanjing 210042,China)

机构地区：[1]东南大学能源与环境学院,南京210096 [2]南京师范大学能源与机械工程学院,南京210042

出　　处：《燃烧科学与技术》2018年第6期506-512,共7页Journal of Combustion Science and Technology

基　　金：国家自然科学基金资助项目(51706109;51390490);江苏省高校自然科学研究资助项目(16KJB470012);江苏省博士后科研资助计划(1601093B)

摘　　要：为了获得循环流化床内颗粒团聚行为特性,利用高速摄像技术,以玻璃珠为床料,在提升管截面为100 mm×25 mm、高为3.2 m的矩形循环流化床试验台上获得不同操作条件下提升管内气固两相运动的连续图像.采用基于K-means算法的聚类方法对图像进行多阈值分割,实现对提升管内团聚物及其内部结构的自动识别,并进一步获得时均颗粒体积分数、团聚物体积分数、团聚核心体积分数、团聚物容积份额和团聚核心容积份额等特征参数及其与操作条件的变化关系.结果表明,团聚物内部存在体积分数梯度;当颗粒体积分数较大时,团聚物内部会形成一个致密的团聚核心,此时团聚物由团聚核心及其周围的团聚云两部分组成;当颗粒体积分数较小时,团聚物内部核心消失,团聚物内部体积分数梯度变小.To understand the behavior of particle clusters in circulating fluidized bed(CFB),experiments were conducted with glass beads to acquire the image sequences of gas-solid flow on a100mm×25mm square crosssection and3.2m height CFB riser by adopting high-speed image technology.An efficient image multilevel thresholding approach was applied to perform image segmentation via clustering analysis based on the K-means algorithm and automatically identify clusters and the internal structure of the riser.Cluster characteristics,such as mean solids holdup,cluster concentration,cluster core concentration,cluster and cluster core fractions,and the relationship with operating conditions,were subsequently obtained.The results show that a concentration gradient exists inside the clusters.When solids holdup is high,a dense core is formed inside the cluster,indicating that the cluster comprises a cluster core region and its surrounding cluster cloud.In contrast,the cluster core disappears at low solids holdup,regardless of whether the concentration gradient of the cluster decreases.

关 键 词：循环流化床 颗粒团聚 高速摄像 多阈值 图像分割 

分 类 号：TQ021[化学工程] TQ022.3