期刊
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
卷 64, 期 5, 页码 4325-4335出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TIE.2017.2677351
关键词
Hidden Markov models (HMMs); multi-modal processes; Viterbi algorithm
资金
- Natural Sciences and Engineering Research Council of Canada
For increased efficiency and profitability, many processes have multiple modes of operation. Switching between different operating modes is performed according to the standard operating procedures. These procedures are set by considering safety and operating limitations of various subsystems and equipment, and thus put restrictions on the switching of the process modes. In this paper, a hidden Markov model based monitoring method is proposed that can not only handle the multimodality of process data but can also capture the mode switching restrictions. A two-step Viterbi algorithm is proposed for effective mode detection in the event of faults, and a reconstruction-based fault isolation algorithm is developed to build the contribution plots. Application examples demonstrate the effectiveness of the proposed monitoring method.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据