Robust Nonlinear PLS Based on Neural Networks and Application to Composition Estimator for High-Purity Distillation Columns

The accurate and reliable on-line estimation of product quality is an essential task for successful process operation and control. This paper proposes a new estimation method that extends the conventional linear PLS (Partial Least Squares) regression method to a nonlinear framework in a more robust manner. To handle the nonlinearities, nonlinear PLS based on linear PLS and neural network has been employed. To improve the robustness of the nonlinear PLS, the autoassociative neural network has been integrated with nonlinear PLS. The integration allows us to handle the nonlinear correlation as well as nonlinear functional relationship with fewer components in a more robust manner. The application results have shown that the proposed Robust Nonlinear PLS (RNPLS) performs better than previous linear and nonlinear regression methods such as PLS, NNPLS, even for the nonlinearities due to operating condition changes, limited observations, and measurement noise.

[1]  M. Morari,et al.  Understanding the Dynamic Behavior of Distillation Columns , 1988 .

[2]  S. Skogestad,et al.  Composition estimator in a pilot-plant distillation column using multiple temperatures , 1991 .

[3]  Sophocles J. Orfanidis,et al.  GramSchmidt Neural Nets , 1990, Neural Computation.

[4]  S. Skogestad,et al.  Estimation of distillation compositions from multiple temperature measurements using partial-least-squares regression , 1991 .

[5]  T. McAvoy,et al.  Integration of multilayer perceptron networks and linear dynamic models : a Hammerstein modeling approach , 1993 .

[6]  Thomas J. McAvoy,et al.  Nonlinear PLS Modeling Using Neural Networks , 1992 .

[7]  Lennart Ljung,et al.  System Identification: Theory for the User , 1987 .

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

[9]  Babu Joseph,et al.  Inferential control of processes: Part I. Steady state analysis and design , 1978 .

[10]  L. E. Scales,et al.  Introduction to Non-Linear Optimization , 1985 .

[11]  Julian Morris,et al.  Real-Time Optimization and Model-Based Control of Polymer Reactors , 1995 .

[12]  T. McAvoy,et al.  Nonlinear principal component analysis—Based on principal curves and neural networks , 1996 .

[13]  John B. Moore,et al.  Singular Value Decomposition , 1994 .

[14]  Manfred Morari,et al.  PLS/neural networks , 1992 .

[15]  S. Skogestad,et al.  Output estimation using multiple secondary measurements: High‐purity distillation , 1993 .

[16]  M. Kramer.  Nonlinear principal component analysis using autoassociative neural networks , 1991 .

[17]  Mark A. Kramer,et al.  Autoassociative neural networks , 1992 .

[18]  Jan de Leeuw,et al.  Nonlinear Principal Component Analysis , 1982 .

[19]  Massimo Morbidelli,et al.  Development of a composition estimator for binary distillation columns. Application to a pilot plant , 1995 .

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

[21]  Ivan Dvořák,et al.  Singular-value decomposition in attractor reconstruction: pitfalls and precautions , 1992 .

[22]  Petre Stoica,et al.  Decentralized Control , 2018, The Control Systems Handbook.

[23]  Kurt Hornik,et al.  Universal approximation of an unknown mapping and its derivatives using multilayer feedforward networks , 1990, Neural Networks.

[24]  James L. McClelland,et al.  Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations , 1986 .

[25]  S. Wold.  Cross-Validatory Estimation of the Number of Components in Factor and Principal Components Models , 1978 .

[26]  J. Macgregor,et al.  Development of inferential process models using PLS , 1994 .

[27]  Composition Estimators for a Distillation Column: Theory and Experiments , 1995 .