SCI和EI收录∣中国化工学会会刊

›› 2016, Vol. 24 ›› Issue (10): 1413-1422.DOI: 10.1016/j.cjche.2016.06.011

• Process Systems Engineering and Process Safety • Previous Articles     Next Articles

Multivariate time delay analysis based local KPCA fault prognosis approach for nonlinear processes

Yuan Xu, Ying Liu, Qunxiong Zhu   

  1. College of Information Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China
  • Received:2015-10-19 Revised:2016-02-17 Online:2016-11-19 Published:2016-10-28
  • Supported by:
    Supported by the National Natural Science Foundation of China (61573051, 61472021), the Natural Science Foundation of Beijing (4142039), Open Fund of the State Key Laboratory of Software Development Environment (SKLSDE-2015KF-01), and Fundamental Research Funds for the Central Universities (PT1613-05).

Multivariate time delay analysis based local KPCA fault prognosis approach for nonlinear processes

Yuan Xu, Ying Liu, Qunxiong Zhu   

  1. College of Information Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China
  • 通讯作者: Qunxiong Zhu,E-mail address:zhuqx@mail.buct.edu.cn.
  • 基金资助:
    Supported by the National Natural Science Foundation of China (61573051, 61472021), the Natural Science Foundation of Beijing (4142039), Open Fund of the State Key Laboratory of Software Development Environment (SKLSDE-2015KF-01), and Fundamental Research Funds for the Central Universities (PT1613-05).

Abstract: Currently, some fault prognosis technology occasionally has relatively unsatisfied performance especially for incipient faults in nonlinear processes duo to their large time delay and complex internal connection. To overcome this deficiency,multivariate time delay analysis is incorporated into the high sensitive local kernel principal component analysis. In this approach, mutual information estimation and Bayesian information criterion (BIC) are separately used to acquire the correlation degree and time delay of the process variables. Moreover, in order to achieve prediction, time series prediction by back propagation (BP) network is applied whose input is multivariate correlated time series other than the original time series. Then the multivariate time delayed series and future values obtained by time series prediction are combined to construct the input of local kernel principal component analysis (LKPCA)model for incipient fault prognosis. The newmethod has been exemplified in a simple nonlinear process and the complicated Tennessee Eastman (TE) benchmark process. The results indicate that the new method has superiority in the fault prognosis sensitivity over other traditional fault prognosis methods.

Key words: Fault prognosis, Time delay estimation, Local kernel principal component analysis

摘要: Currently, some fault prognosis technology occasionally has relatively unsatisfied performance especially for incipient faults in nonlinear processes duo to their large time delay and complex internal connection. To overcome this deficiency,multivariate time delay analysis is incorporated into the high sensitive local kernel principal component analysis. In this approach, mutual information estimation and Bayesian information criterion (BIC) are separately used to acquire the correlation degree and time delay of the process variables. Moreover, in order to achieve prediction, time series prediction by back propagation (BP) network is applied whose input is multivariate correlated time series other than the original time series. Then the multivariate time delayed series and future values obtained by time series prediction are combined to construct the input of local kernel principal component analysis (LKPCA)model for incipient fault prognosis. The newmethod has been exemplified in a simple nonlinear process and the complicated Tennessee Eastman (TE) benchmark process. The results indicate that the new method has superiority in the fault prognosis sensitivity over other traditional fault prognosis methods.

关键词: Fault prognosis, Time delay estimation, Local kernel principal component analysis