Decentralized fault diagnosis using multiblock kernel independent component analysis

• Published in: Chemical engineering research & design Vol. 90; no. 5; pp. 667 - 676
• Main Authors: Zhang, Yingwei, Ma, Chi
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.05.2012; Elsevier
• Subjects: Applied sciences; Chemical engineering; Exact sciences and technology; Fault detection and diagnosis; Kernel independent component analysis; Multiblock kernel methods; Process monitoring; Safety; Process monitoring; Multiblock kernel methods; Kernel independent component analysis; Fault detection and diagnosis; Annealing; Surveillance; Independent component analysis; Failure detection; Algorithm; Fault diagnostic
• ISSN: 0263-8762
• DOI: 10.1016/j.cherd.2011.09.011

► A multiblock kernel independent component analysis (MBKICA) algorithm is proposed. ► A new fault diagnosis approach based on MBKICA is proposed to monitor large-scale processes. ► The nonlinearity and non-Gaussianity in the block process variables are extracted. In this paper, a multiblock kernel independent component analysis (MBKICA) algorithm is proposed. Then a new fault diagnosis approach based on MBKICA is proposed to monitor large-scale processes. MBKICA has superior fault diagnosis ability since variables are grouped and the non-Gaussianity is considered compared to standard kernel methods. The proposed method is applied to fault detection and diagnosis in the continuous annealing process. The proposed decentralized nonlinear approach effectively captures the nonlinear relationship and non-Gaussianity in the block process variables, and shows superior fault diagnosis ability compared to other methods.