Risk‐based process system monitoring using self‐organizing map integrated with loss functions

Free to read The conventional dynamic risk assessment technique does not consider the effect of nonlinear interactions among process variables in its operational risk estimation. Thus, this type of technique fails to provide a realistic estimation of the operational risk of complex industrial processes. To address this issue, a multivariate risk-based process monitoring technique is proposed. This technique takes advantage of the powerful nonlinear dimensionality reduction and visualization power of the self-organizing map to identify the origin and propagation path of the fault. Through integration with the inverted normal loss function, a robust estimation of the hazard potential and operational risk of process operations can be achieved. The proposed technique is tested with two fault conditions in the benchmark Tennessee Eastman chemical process. The results show promising performance.

[1]  Vikram Garaniya,et al.  Self-Organizing Map Based Fault Diagnosis Technique for Non-Gaussian Processes , 2014 .

[2]  Faisal Khan,et al.  Dynamic Risk Assessment and Fault Detection Using Principal Component Analysis , 2013 .

[3]  Erkki Oja,et al.  Independent component analysis: algorithms and applications , 2000, Neural Networks.

[4]  Geoffrey E. Hinton,et al.  A View of the Em Algorithm that Justifies Incremental, Sparse, and other Variants , 1998, Learning in Graphical Models.

[5]  Stelios Psarakis,et al.  Multivariate statistical process control charts: an overview , 2007, Qual. Reliab. Eng. Int..

[6]  E. F. Vogel,et al.  A plant-wide industrial process control problem , 1993 .

[7]  Faisal Khan,et al.  Loss functions and their applications in process safety assessment , 2014 .

[8]  Richard D. Braatz,et al.  Fault Detection and Diagnosis in Industrial Systems , 2001 .

[9]  Thomas E. Marlin,et al.  Multivariate statistical monitoring of process operating performance , 1991 .

[10]  James O. Ramsay,et al.  Functional Data Analysis , 2005 .

[11]  Fred A. Spiring,et al.  The reflected normal loss function , 1993 .

[12]  B. Bakshi.  Multiscale PCA with application to multivariate statistical process monitoring , 1998 .

[13]  Manabu Kano,et al.  A new multivariate statistical process monitoring method using principal component analysis , 2001 .

[14]  John F. MacGregor,et al.  Process monitoring and diagnosis by multiblock PLS methods , 1994 .

[15]  N. Lawrence Ricker,et al.  Decentralized control of the Tennessee Eastman Challenge Process , 1996 .

[16]  Faisal Khan,et al.  Risk-based operational performance analysis using loss functions , 2014 .

[17]  Faisal Khan,et al.  A risk-based approach to design warning system for processing facilities , 2011 .

[18]  S. Joe Qin,et al.  Statistical process monitoring: basics and beyond , 2003 .

[19]  H. Abdi,et al.  Principal component analysis , 2010 .

[20]  Lijuan Cao,et al.  A comparison of PCA, KPCA and ICA for dimensionality reduction in support vector machine , 2003, Neurocomputing.

[21]  Faisal Khan,et al.  Development of Economic Consequence Methodology for Process Risk Analysis , 2015, Risk analysis : an official publication of the Society for Risk Analysis.

[22]  S. Joe Qin,et al.  Subspace approach to multidimensional fault identification and reconstruction , 1998 .

[23]  Aapo Hyvärinen,et al.  A Fast Fixed-Point Algorithm for Independent Component Analysis , 1997, Neural Computation.

[24]  Manabu Kano,et al.  Comparison of multivariate statistical process monitoring methods with applications to the Eastman challenge problem , 2002 .

[25]  Barry Lennox,et al.  Monitoring a complex refining process using multivariate statistics , 2008 .

[26]  Ping Zhang,et al.  A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process , 2012 .