A Kernel Time Structure Independent Component Analysis Method for Nonlinear Process Monitoring

• Published in: Chinese journal of chemical engineering 2014-11, Vol.22 (11-12), p.1243-1253
• Main Authors: Cai, Lianfang, Tian, Xuemin, Zhang, Ni
• Format: Article
• Language: English
• Subjects: Fault identification; Faults; Feature extraction; Independent component analysis; Kernel trick; Kernels; Monitoring; Nonlinearity; Process monitoring; Sensitivity analysis; Statistics; Time structure
• ISSN: 1004-9541; 2210-321X
• DOI: 10.1016/j.cjche.2014.09.021

Kernel independent component analysis (KICA) is a newly emerging nonlinear process monitoring method, which can extract mutually independent latent variables called independent components (ICs) from process variables. However, when more than one IC have Gaussian distribution, it cannot extract the IC feature effectively and thus its monitoring performance will be degraded drastically. To solve such a problem, a kernel time structure independent component analysis (KTSICA) method is proposed for monitoring nonlinear process in this paper. The original process data are mapped into a feature space nonlinearly and then the whitened data are calculated in the feature space by the kernel trick. Subsequently, a time structure independent component analysis algorithm, which has no requirement for the distribution of ICs, is proposed to extract the IC feature. Finally, two monitoring statistics are built to detect process faults. When some fault is detected, a nonlinear fault identification method is developed to identify fault variables based on sensitivity analysis. The proposed monitoring method is applied in the Tennessee Eastman benchmark process. Applications demonstrate the superiority of KTSICA over KICA. Fig. 1 The monitoring result of KTSICA for fault 4 in TE process. Fig. 2 The fault identification result of KTSICA for fault 4 in TE process. A KTSICA-based nonlinear process monitoring method is proposed. Firstly, a TSICA method without the requirement for IC distributions is developed. Then, to meet the monitoring demand of nonlinear process, a KTSICA method is proposed by extending TSICA with kernel technique. With the extracted ICs by KTSICA, two monitoring statistics are constructed to detect process faults and a new nonlinear contribution plot method based on sensitivity analysis is established to identify fault variables. Simulation results for the nonlinear TE process demonstrate that the fault detection performance of KTSICA outperforms that of KICA and the built contribution plot method can give an effective indication for fault variables. [Display omitted]