A PLANT-FRIENDLY MULTIVARIABLE SYSTEM IDENTIFICATION FRA MEWORK BASED ON IDENTIFICATION TEST MONITORING

Historically, model development for advanced process control applications has been a major consideration, demanding significant time and effort. The increased use of advanced control systems in industry creates a need for efficient methods for multivariable system identification that systematically refine process knowledge, leading to models that achieve desirable control performance. Moreover, time-varying changes and the aging of process equipment frequently demand model maintenance and control system tuning during the life of process operation. A comprehensive identification test monitoring procedure can aid in resolving this significant model development challenge. This dissertation presents a plant-friendly identification framework, aimed at developing dynamic models for multivariable systems. The components of the framework include plant-friendly multisine input design, frequency response estimation, control-relevant parameter estimation, and robust loopshaping. These components are implemented in a plant-friendly manner to facilitate industrial implementation. Deterministic, periodic multisine input signals are developed to perform plant-friendly experimental testing. A series of multisine design guidelines are derived based on a priori knowledge to generate a desirable input power spectral density. The use of constrained optimization enforces requirements on manipulated and controlled variables. A control-relevant parameter estimation procedure is formulated for curvefitting frequency responses generated from data into linear Matrix Fractional Description models with Model Predictive Control (MPC)-relevant weightings. The MPC-relevant weights emphasize closed-loop performance requirements in the curvefit. A set of models defined by the curvefitted model and uncertainty bounds are used in a robust loopshaping procedure, based on Structured Singular Value (μ) analysis. Robust stability and performance bounds are computed and used as criteria for defining model adequacy with respect to the end-use control application, and to decide when to halt or continue experimental testing. The framework provides a viable tool for performing experimental testing and controller design of systems involving strong interaction, ill-conditioning, and gain-directionality considerations. The user can conduct identification experimental testing of multivariable systems displaying these challenges while meeting practical plant-friendliness considerations. A series of case studies involving distillation column control are presented to demonstrate the effectiveness of the integrated framework. iii The Lord is my shepherd; I shall not want. He maketh me to lie down in green pastures: he leadeth me beside the still waters. He restoreth my soul: he leadeth me in the paths of righteousness for his name’s sake. Psalms 23:1-3

[1]  Tegoeh Tjahjowidodo,et al.  Identification of pre-sliding and sliding friction dynamics: Grey box and black-box models , 2007 .

[2]  M. Morari,et al.  LV-Control of a high-purity distillation column , 1987 .

[3]  Jorge Nocedal,et al.  A trust region method based on interior point techniques for nonlinear programming , 2000, Math. Program..

[4]  Daniel E. Rivera,et al.  "Model-on-Demand" Identification for Control: An Experimental Study and Feasibility Analysis for MoD-Based Predictive Control , 2000 .

[5]  Hyunjin Lee,et al.  CR-IDENT: A MATLAB TOOLBOX FOR MULTIVARIABLE CONTROL-RELEVANT SYSTEM IDENTIFICATION , 2006 .

[6]  Fred Y. Hadaegh,et al.  Multivariable Plant Set Estimation using Multisinusoidal Input Designs , 1994 .

[7]  J. Macgregor,et al.  Design of identification experiments for robust control. A geometric approach for bivariate processes , 1993 .

[8]  Jorge Nocedal,et al.  An Interior Point Algorithm for Large-Scale Nonlinear Programming , 1999, SIAM J. Optim..

[9]  John F. MacGregor,et al.  Identification for robust multivariable control: The design of experiments , 1994, Autom..

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

[11]  Tegoeh Tjahjowidodo,et al.  IDENTIFICATION OF PRE-SLIDING AND SLIDING FRICTION , 2005 .

[12]  Lennart Ljung,et al.  Modeling Of Dynamic Systems , 1994 .

[13]  Fredrik Gustafsson,et al.  Just in time models for dynamical systems , 1996, Proceedings of 35th IEEE Conference on Decision and Control.

[14]  R. A. de Callafon,et al.  Multivariable least squares frequency domain identification using polynomial matrix fraction descriptions , 1996 .

[15]  I-Lung Chien,et al.  Nonlinear identification and control of a high-purity distillation column: a case study , 1995 .

[16]  Manfred R. Schroeder,et al.  Synthesis of low-peak-factor signals and binary sequences with low autocorrelation (Corresp.) , 1970, IEEE Trans. Inf. Theory.

[17]  Lennart Ljung,et al.  System identification (2nd ed.): theory for the user , 1999 .

[18]  Jay H. Lee,et al.  Model predictive control: past, present and future , 1999 .

[19]  J. Doyle,et al.  Essentials of Robust Control , 1997 .

[20]  Yucai Zhu,et al.  Multivariable System Identification For Process Control , 2001 .

[21]  Daniel E. Rivera,et al.  Application of minimum crest factor multisinusoidal signals for "plant-friendly" identification of nonlinear process systems , 2000 .

[22]  Asaad Y. Shamseldin.  Review of the Applications of Neural Networks and Fuzzy Systems in Hydrological Modelling , 2004 .

[23]  K. H. Fasol,et al.  Principles of Model Building and Identification , 1979 .

[24]  Ian Postlethwaite,et al.  Multivariable Feedback Control: Analysis and Design , 1996 .

[25]  H. A. Barker,et al.  Nonlinear System Identification Using Pseudorandom Signals with Partially Orthogonal Transforms , 1985 .

[26]  P. V. D. Hof.  System order and structure indices of linear systems in polynomial form , 1992 .

[27]  George Cybenko,et al.  Just-in-Time Learning and Estimation , 1996 .

[28]  Keith R. Godfrey,et al.  Perturbation signals for system identification , 1993 .

[29]  K. H. Fasol,et al.  Principles of model building and identification , 1979, Autom..

[30]  Håkan Hjalmarsson,et al.  From experiment design to closed-loop control , 2005, Autom..

[31]  David S. Bayard,et al.  High-order multivariable transfer function curve fitting: Algorithms, sparse matrix methods and experimental results , 1994, Autom..

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

[33]  Petre Stoica,et al.  Introduction to spectral analysis , 1997 .

[34]  Anders Stenman,et al.  Model on Demand: Algorithms, Analysis and Applications , 1999 .

[35]  Babatunde A. Ogunnaike,et al.  Process Dynamics, Modeling, and Control , 1994 .

[36]  L. Ljung,et al.  On Adaptive Smoothing of Empirical Transfer Function Estimates , 1999 .

[37]  Dale E. Seborg,et al.  Nonlinear Process Control , 1996 .

[38]  Daniel E. Rivera,et al.  An integrated identification and control design methodology for multivariable process system applications , 2000 .

[39]  Michel Gevers,et al.  ARMA models, their Kronecker indices and their McMillan degree , 1986 .

[40]  C. Sanathanan,et al.  Transfer function synthesis as a ratio of two complex polynomials , 1963 .

[41]  D. Riordan,et al.  An integrated identification methodology for control and identification , 1992 .

[42]  Daniel E. Rivera,et al.  SYSTEMATIC TECHNIQUES FOR DETERMINING MODELING REQUIREMENTS FOR SISO AND MIMO FEEDBACK CONTROL PROBLEMS , 1994 .

[43]  Chun Tung Chou,et al.  Nonlinear Identification of High Purity Distillation Columns , 2000 .

[44]  Stanley H. Johnson,et al.  Use of Hammerstein Models in Identification of Nonlinear Systems , 1991 .

[45]  R. A. de Callafon,et al.  Feedback oriented identification for enhanced and robust control: a fractional approach applied to a wafer stage. , 1998 .

[46]  Daniel E. Rivera,et al.  A 'Model-on-Demand' identification methodology for non-linear process systems , 2001 .

[47]  Daniel E. Rivera,et al.  Systematic techniques for determining modelling requirements for SISO and MIMO feedback control , 1995 .

[48]  Francis J. Doyle,et al.  The identification of nonlinear models for process control using tailored “plant-friendly” input sequences , 2001 .

[49]  H. Anthony Barker.  Design of multi-level pseudo-random signals for system identification , 1993 .

[50]  H. Weyl.  Über die Gleichverteilung von Zahlen mod. Eins , 1916 .

[51]  W Li,et al.  Frequency‐domain closed‐loop identification of multivariable systems for feedback control , 1996 .

[52]  M. Morari,et al.  Control-relevant model reduction problems for SISO H2, H∞, and μ-controller synthesis , 1987 .

[53]  Paul M. J. Van den Hof,et al.  Identification and control - Closed-loop issues , 1995, Autom..

[54]  Yucai Zhu,et al.  Case studies on closed-loop identification for MPC , 2002 .

[55]  A. Forbes.  Modeling and control , 1990, Journal of Clinical Monitoring.

[56]  D. G. Watts,et al.  Spectral analysis and its applications , 1968 .

[57]  P. Lindskog.  Methods, Algorithms and Tools for System Identification Based on Prior Knowledge , 1996 .

[58]  Evanghelos Zafiriou,et al.  Robust process control , 1987 .

[59]  Daniel E. Rivera,et al.  Control-Relevant Input Signal Design for Multivariable System Identification: Application to High-Purity Distillation , 1996 .

[60]  T. J. McAvoy,et al.  Feasibility of Decoupling in Conventionally Controlled Distillation Columns , 1980 .

[61]  K. R. Godfrey,et al.  Pseudorandom signals for the dynamic analysis of multivariable systems , 1966 .

[62]  David S. Bayard.  Statistical plant set estimation using Schroeder-phased multisinusoidal input design , 1993 .

[63]  Yucai Zhu,et al.  Some study on identification of ill-conditioned processes for control , 2001, Proceedings of the 2001 American Control Conference. (Cat. No.01CH37148).

[64]  Peter Lindskog,et al.  A System Identification Software Tool for General MISO ARX-Type of Model Structures , 1996 .

[65]  Hyunjin Lee,et al.  Constrained minimum crest factor multisine signals for "Plant-Friendly" identification of highly interactive systems , 2003 .

[66]  J. Schoukens,et al.  Design of excitation signals for the restoring force surface method , 1995 .

[67]  Anders Stenman,et al.  Model-on-Demand Model Predictive Control for Nonlinear Process Systems , 2000 .

[68]  Hyunjin Lee,et al.  "Plant-Friendly" system identification: a challenge for the process industries , 2003 .

[69]  Wei Li,et al.  Control relevant identification of ill-conditioned systems: Estimation of gain directionalyty , 1996 .

[70]  J. Schoukens,et al.  Crest-factor minimization using nonlinear Chebyshev approximation methods , 1991 .

[71]  Yucai Zhu.  Black-box identification of mimo transfer functions: Asymptotic properties of prediction error models , 1989 .

[72]  S. V. Gaikwad,et al.  Multivariable frequency-response curve fitting with application to control-relevant parameter estimation , 1997, Autom..

[73]  Lorenz T. Biegler,et al.  Failure of global convergence for a class of interior point methods for nonlinear programming , 2000, Math. Program..

[74]  R. Pearson.  Discrete-time Dynamic Models , 1999 .

[75]  Hans D. Mittelmann,et al.  CONSTRAINED MULTISINE INPUTS FOR PLANT-FRIENDLY IDENTIFICATION OF CHEMICAL PROCESSES , 2002 .

[76]  K. Scarbrough,et al.  of Electrical Engineering , 1982 .

[77]  Sigurd Skogestad,et al.  Inconsistencies in Dynamic Models for Ill-Conditioned Plants: Application to Low-Order Models of Distillation Columns , 1994 .

[78]  Carlos E. García,et al.  Fundamental Process Control , 1988 .

[79]  Daniel E. Rivera,et al.  A Control-Relevant Multivariable System Identification Methodology Based on Orthogonal Multifrequency Input Perturbations , 1997 .

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