化工过程动态响应特征的提取及分类方法  

FEATURE EXTRACTION AND CLASSIFICATION FOR DYNAMIC RESPONSE IN CHEMICAL PROCESSES

作　　者：余红路[1] 沈静珠[1] 胡山鹰[1] 

机构地区：[1]清华大学化学工程系,北京100084

出　　处：《化工学报》1998年第1期73-79,共7页CIESC Journal

摘　　要：以加氢裂化反应器为例进行了动态响应分类方法的探讨，研究了针对化工动态过程的特点提取动态响应特征的方法，以及RBF神经网络在动态响应分类中的应用和与其他几种分类器的比较．The classification of dynamic response is important in dynamic fault diagnosis and on-line operation optimization.In this work,different methods of classification for dynamic response are discussed based on a hydrocracking reactor model.It is used to extract features for dynamic response according to the characteristics of chemical processes.Radial Basis Function Network(RBFN) is applied to make classification for dynamic response and is compared with other classification methods.

关 键 词：动态响应 RBF神经网络 故障诊断 加氢裂化反应器 

分 类 号：TQ063[化学工程] TE624.432[石油与天然气工程—油气加工工程]