Computing meaningful models of continuous data

Abstract Parameters controlling continuous chemical processes as well as all instruments found in such processes generate a huge amount of data. Calculating models to predict critical quality attributes (in y ) from process data (in X ) can be fastidious. Partial-Least-Squares (PLS) regression is now widely present in the industry for process analysis and modelling purposes. However, this methodology requires static data while the majority of chemical processes operate continuously. Some dynamic PLS variants have been developed, but they cannot associate specific critical process parameters with meaningful process dynamics. This work presents a novel Dynamic-PLS variant developed to resolve these issues. Using one real dataset and artificial cases, the performances of this novel methodology is compared to standard regression methodologies. It is demonstrated that the proposed methodology calculates the most robust models in validation. With the real dataset, it is also demonstrated that the predictive model calculated from the proposed methodology can easily be interpreted, leading to better process understanding.

[1]  Giuseppe Palermo,et al.  Performance of PLS regression coefficients in selecting variables for each response of a multivariate PLS for omics-type data , 2009, Advances and applications in bioinformatics and chemistry : AABC.

[2]  C. Jun,et al.  Performance of some variable selection methods when multicollinearity is present , 2005 .

[3]  H. Wold.  Path Models with Latent Variables: The NIPALS Approach , 1975 .

[4]  Anil K. Jain,et al.  Artificial Neural Networks: A Tutorial , 1996, Computer.

[5]  P. A. Taylor,et al.  Missing data methods in PCA and PLS: Score calculations with incomplete observations , 1996 .

[6]  Sirish L. Shah,et al.  Constrained nonlinear MPC using hammerstein and wiener models: PLS framework , 1998 .

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

[8]  L. Rabiner,et al.  FIR digital filter design techniques using weighted Chebyshev approximation , 1975, Proceedings of the IEEE.

[9]  Wynne W. Chin.  Bootstrap Cross-Validation Indices for PLS Path Model Assessment , 2010 .

[10]  A. Çinar,et al.  Online batch/fed-batch process performance monitoring, quality prediction, and variable-contribution analysis for diagnosis , 2003 .

[11]  Jorge J. Moré,et al.  The Levenberg-Marquardt algo-rithm: Implementation and theory , 1977 .

[12]  Yi-Zeng Liang,et al.  Monte Carlo cross‐validation for selecting a model and estimating the prediction error in multivariate calibration , 2004 .

[13]  Barry M. Wise,et al.  A comparison of neural networks, non-linear biased regression and a genetic algorithm for dynamic model identification , 1995 .

[14]  S. Joe Qin,et al.  A novel dynamic PCA algorithm for dynamic data modeling and process monitoring , 2017 .

[15]  Jesús Picó,et al.  Multi‐phase analysis framework for handling batch process data , 2008 .

[16]  In-Cheol Park,et al.  FIR filter synthesis algorithms for minimizing the delay and the number of adders , 2001 .

[17]  Jun Liang,et al.  Multi-loop nonlinear internal model controller design under nonlinear dynamic PLS framework using ARX-neural network model , 2012 .

[18]  A. Ferrer,et al.  Dealing with missing data in MSPC: several methods, different interpretations, some examples , 2002 .

[19]  Ali Cinar,et al.  Statistical process monitoring and disturbance diagnosis in multivariable continuous processes , 1996 .

[20]  Leo H. Chiang,et al.  Fault diagnosis in chemical processes using Fisher discriminant analysis, discriminant partial least squares, and principal component analysis , 2000 .

[21]  Damaris Zurell,et al.  Collinearity: a review of methods to deal with it and a simulation study evaluating their performance , 2013 .

[22]  J. Colodette,et al.  The impact of kappa number composition on eucalyptus kraft pulp bleachability , 2007 .

[23]  Bin Hu,et al.  Multi-loop Internal Model Controller Design Based on a Dynamic PLS Framework , 2010 .

[24]  Junghui Chen,et al.  On-line batch process monitoring using dynamic PCA and dynamic PLS models , 2002 .

[25]  S. Wold,et al.  The Collinearity Problem in Linear Regression. The Partial Least Squares (PLS) Approach to Generalized Inverses , 1984 .

[26]  Armando B. Corripio,et al.  Dynamic neural networks partial least squares (DNNPLS) identification of multivariable processes , 2003, Comput. Chem. Eng..

[27]  William J. Mcbride,et al.  PERT and the beta distribution , 1967 .

[28]  Ramesh A. Gopinath,et al.  Least squared error FIR filter design with transition bands , 1992, IEEE Trans. Signal Process..

[29]  Elaine B. Martin,et al.  Dynamic Non-Linear PLS Modelling , 2000 .

[30]  S. Joe Qin,et al.  Dynamic-Inner Partial Least Squares for Dynamic Data Modeling , 2015 .

[31]  Rafael Herrerías Pleguezuelo,et al.  Revisiting the PERT mean and variance , 2011, Eur. J. Oper. Res..

[32]  Pedro Hernandez-Abad,et al.  Quality by design case study: an integrated multivariate approach to drug product and process development. , 2009, International journal of pharmaceutics.

[33]  S. Wold,et al.  PLS-regression: a basic tool of chemometrics , 2001 .

[34]  T. Gustafsson,et al.  Model predictive control of the chip level in a continuous pulp digester, a case study , 2005 .

[35]  A. Höskuldsson.  PLS regression methods , 1988 .

[36]  George-Othon Glentis,et al.  Efficient least squares adaptive algorithms for FIR transversal filtering , 1999, IEEE Signal Process. Mag..

[37]  Sirish L. Shah,et al.  Modeling and control of multivariable processes: Dynamic PLS approach , 1997 .

[38]  W. Harmon Ray,et al.  Dynamic PLS modelling for process control , 1993 .

[39]  Julian Morris,et al.  Non-Linear Model Based Predictive Control Through Dynamic Non-Linear Partial Least Squares , 2002 .

[40]  P. A. Taylor,et al.  The impact of missing measurements on PCA and PLS prediction and monitoring applications , 2006 .

[41]  John F. MacGregor,et al.  Modeling of dynamic systems using latent variable and subspace methods , 2000 .

[42]  A. J. Morris,et al.  Non-linear dynamic projection to latent structures modelling , 2000 .

[43]  Dae Sung Lee,et al.  Application of a Moving-Window-Adaptive Neural Network to the Modeling of a Full-Scale Anaerobic Filter Process , 2005 .

[44]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[45]  Gavin C. Cawley,et al.  Preventing Over-Fitting during Model Selection via Bayesian Regularisation of the Hyper-Parameters , 2007, J. Mach. Learn. Res..

[46]  Dae Sung Lee,et al.  Nonlinear dynamic partial least squares modeling of a full-scale biological wastewater treatment plant , 2006 .

[47]  Manel Alcalà,et al.  Real-time determination of critical quality attributes using near-infrared spectroscopy: a contribution for Process Analytical Technology (PAT). , 2012, Talanta.

[48]  A. J. Morris,et al.  Non-linear projection to latent structures revisited (the neural network PLS algorithm) , 1999 .

[49]  Francis B. Lavoie,et al.  Generalization of Powered–Partial-Least-Squares , 2018, Chemometrics and Intelligent Laboratory Systems.

[50]  Furong Gao,et al.  A survey on multistage/multiphase statistical modeling methods for batch processes , 2009, Annu. Rev. Control..

[51]  Thomas J. McAvoy,et al.  Nonlinear FIR Modeling via a Neural Net PLS Approach , 1996 .

[52]  T. I. Kirkpatrick,et al.  PERT as an aid to logic design , 1966 .

[53]  W. Harmon Ray,et al.  Chemometric methods for process monitoring and high‐performance controller design , 1992 .

[54]  George Stephanopoulos,et al.  Inverse modeling using multi-block PLS to determine the environmental conditions that provide optimal cellular function , 2004, Bioinform..

[55]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[56]  Richard D. Braatz,et al.  Data-driven Methods for Fault Detection and Diagnosis in Chemical Processes , 2000 .