标题：Hidden semi-Markov model based monitoring algorithm for multimode processes
作者：Lou, Zhijiang ;Wang, Youqing
作者机构：[Lou, Zhijiang ;Wang, Youqing ] College of Information Science and Technology, Beijing University of Chemical Technology, Beijing; 100029, China;[Wang 更多
会议名称：6th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2017
会议日期：May 26, 2017 - May 27, 2017
来源：Proceedings of 2017 IEEE 6th Data Driven Control and Learning Systems Conference, DDCLS 2017
摘要：Several studies have adopted hidden Markov model (HMM) to monitor multimode processes. The drawback of HMM is that its inherent duration probability density is exponential and hence it is inappropriate for the modeling of multimode processes. To address this problem, hidden semi-Markov model (HSMM), which introduces the mode duration probability into HMM, is combined with principal component analysis (PCA) in this paper, named as HSMM-PCA. With the restriction of mode duration probability, HSMM-PCA can successfully identify the operation mode affiliation and build the precise PCA model for each mode. As a result, HSMM-PCA is more sensitive to abnormal conditions and has better fault detection ability for multimode processes.
© 2017 IEEE.