Statistical process monitoring of a base metal flotation plant using a Gaussian mixture model and dissimilarity scale based singular spectrum analysis

Singular spectrum analysis (SSA) is a data-adaptive multimodal method based on singular value decomposition or equivalently, principal component analysis, which is a promising tool for process monitoring and fault diagnosis in chemical process systems. Statistical monitoring of a base metal flotation plant using dissimilarity scale based singular spectrum analysis (DSSA) is considered in this paper. Monitoring of multiscale signals using principal component analysis obtained from the multilevel decomposition of process data using conventional SSA and DSSA assume the data to have a Gaussian distribution. This assumption limits the performance of SSA-based approaches when applied to the monitoring of complex nonlinear processes. To address this issue, a Gaussian mixture model was used to estimate the probability density function for the Hotelling’s T2 and the Qstatistics of the model. Application of the proposed study demonstrated that, in comparison with conventional SSA-based monitoring, the proposed process monitoring scheme is more reliable and efficient in detecting faults in a smaller number of modes.

[1]  R. Vautard,et al.  Singular-spectrum analysis: a toolkit for short, noisy chaotic signals , 1992 .

[2]  J. Macgregor,et al.  Experiences with industrial applications of projection methods for multivariate statistical process control , 1996 .

[3]  Chris Aldrich,et al.  Classification of process dynamics with monte carlo singular spectrum analysis , 2004 .

[4]  Chris Aldrich,et al.  Fault detection in the Tennessee Eastman benchmark process with nonlinear singular spectrum analysis , 2017 .

[5]  Jie Zhang,et al.  Utilizing of Dissimilarity Scale-based PCA in Multivariate Statistical Process Monitoring Application , 2014 .

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

[7]  S. J. Wierda.  Multivariate statistical process control—recent results and directions for future research , 1994 .

[8]  Erkki Oja,et al.  Modified Hebbian learning for curve and surface fitting , 1992, Neural Networks.

[9]  G. P. King,et al.  Extracting qualitative dynamics from experimental data , 1986 .

[10]  G. G. Stokes.  "J." , 1890, The New Yale Book of Quotations.

[11]  Tao Chen,et al.  On-line multivariate statistical monitoring of batch processes using Gaussian mixture model , 2010, Comput. Chem. Eng..

[12]  Chris Aldrich,et al.  Detecting faults in process systems with singular spectrum analysis , 2016 .

[13]  John F. MacGregor.  STATISTICAL PROCESS CONTROL OF MULTIVARIATE PROCESSES , 1994 .

[14]  W. Melssen,et al.  Multivariate statistical process control using mixture modelling , 2005 .

[15]  Anatoly A. Zhigljavsky,et al.  Analysis of Time Series Structure - SSA and Related Techniques , 2001, Monographs on statistics and applied probability.

[16]  Chris Aldrich,et al.  Multiscale Process Monitoring with Singular Spectrum Analysis. , 2007 .

[17]  W. Marsden.  I and J , 2012 .

[18]  D. W. Moolman,et al.  The identification of perturbations in a base metal flotation plant using computer vision of the froth surface , 1997 .

[19]  Chris Aldrich,et al.  Process system identification strategies based on the use of singular spectrum analysis , 2003 .

[20]  Jin Hyun Park,et al.  Process monitoring using a Gaussian mixture model via principal component analysis and discriminant analysis , 2004, Comput. Chem. Eng..