文献类型：会议论文

英文题名：Order Clustering based Sub-stage Division and Its Application to Batch Process Monitoring

作者：Zhang Cheng[1];Zhao Haitao[1]

机构：[1]East China Univ Sci & Technol, Sch Informat Sci & Engn, Shanghai 200030, Peoples R China

会议论文集：35th Chinese Control Conference (CCC)

会议日期：JUL 27-29, 2016

会议地点：Chengdu, PEOPLES R CHINA

语种：英文

外文关键词：batch process monitoring; order information; clustering algorithm; out of sequence

摘要：Clustering algorithms are often adopted in batch process monitoring, especially in multistage batch process monitoring. In process monitoring, the process data are sampled at a specified time interval and generated as time sequences. In this paper, we propose a novel order clustering based sub-stage division (OCSD) algorithm which utilizes the order information (time sequence information) of batch operation and avoids the problem of out of sequence. It is different from the conventional clustering algorithms, such as k-means and Fuzzy C-means (FCM), which assume the processing data are independent and identically distributed (i.i.d.) and do not consider the time sequence information. We applied the proposed algorithm to a typical batch process, penicillin fermentation process and found that monitoring based OCSD can largely reduce the rate of missing alarm and false alarm compared with conventional clustering algorithms.