Model Based Control: Case Studies in Process Engineering

Filling a gap in the literature for a practical approach to the topic, this book is unique in including a whole section of case studies presenting a wide range of applications from polymerization reactors and bioreactors, to distillation column and complex fluid catalytic cracking units. A section of general tuning guidelines of MPC is also present.These thus aid readers in facilitating the implementation of MPC in process engineering and automation. At the same time many theoretical, computational and implementation aspects of model-based control are explained, with a look at both linear and nonlinear model predictive control. Each chapter presents details related to the modeling of the process as well as the implementation of different model-based control approaches, and there is also a discussion of both the dynamic behaviour and the economics of industrial processes and plants. The book is unique in the broad coverage of different model based control strategies and in the variety of applications presented. A special merit of the book is in the included library of dynamic models of several industrially relevant processes, which can be used by both the industrial and academic community to study and implement advanced control strategies.

Table of Contents:
Preface. 1 Introduction. 1.1 Introductory Concepts of Process Control. 1.2 Advanced Process Control Techniques. 1.2.1 Key Problems in Advanced Control of Chemical Processes. 1.2.1.1 Nonlinear Dynamic Behavior. 1.2.1.2 Multivariable Interactions between Manipulated and Controlled Variables. 1.2.1.3 Uncertain and Time-Varying Parameters. 1.2.1.4 Deadtime on Inputs and Measurements. 1.2.1.5 Constraints on Manipulated and State Variables. 1.2.1.6 High-Order and Distributed Processes. 1.2.1.7 Unmeasured State Variables and Unmeasured and Frequent Disturbances. 1.2.2 Classification of the Advanced Process Control Techniques. 2 Model Predictive Control. 2.1 Internal Model Control. 2.2 Linear Model Predictive Control. 2.3 Nonlinear Model Predictive Control. 2.3.1 Introduction. 2.3.2 Industrial Model-Based Control: Current Status and Challenges. 2.3.2.1 Challenges in Industrial NMPC. 2.3.3 First Principle (Analytical) Model-Based NMPC. 2.3.4 NMPC with Guaranteed Stability. 2.3.5 Artificial Neural Network (ANN)-Based Nonlinear Model Predictive Control. 2.3.5.1 Introduction. 2.3.5.2 Basics of ANNs. 2.3.5.3 Algorithms for ANN Training. 2.3.5.4 Direct ANN Model-Based NMPC (DANMPC). 2.3.5.5 Stable DANMPC Control Law. 2.3.5.6 Inverse ANN Model-Based NMPC. 2.3.5.7 ANN Model-Based NMPC with Feedback Linearization. 2.3.5.8 ANN Model-Based NMPC with On-Line Linearization. 2.3.6 NMPC Software for Simulation and Practical Implementation. 2.3.6.1 Computational Issues. 2.3.6.2 NMPC Software for Simulation. 2.3.6.3 NMPC Software for Practical Implementation. 2.4 MPC General Tuning Guidelines. 2.4.1 Model Horizon (n). 2.4.2 Prediction Horizon (p). 2.4.3 Control Horizon (m). 2.4.4 Sampling Time (T). 2.4.5 Weight Matrices ( l y and l u). 2.4.6 Feedback Filter. 2.4.7 Dynamic Sensitivity Used for MPC Tuning. 3 Case Studies. 3.1 Productivity Optimization and Nonlinear Model Predictive Control (NMPC) of a PVC Batch Reactor. 3.1.1 Introduction. 3.1.2 Dynamic Model of the PVC Batch Reactor. 3.1.2.1 The Complex Analytical Model of the PVC Reactor. 3.1.2.2 Morphological Model. 3.1.2.3 The Simplified Dynamic Analytical Model of the PVC Reactor. 3.1.3 Productivity Optimization of the PVC Batch Reactor. 3.1.3.1 The Basic Elements of GAs. 3.1.3.2 Optimization of the PVC Reactor Productivity through the Initial Concentration of Initiators. 3.1.3.3 Optimization of PVC Reactor Productivity by obtaining an Optimal Temperature Policy. 3.1.4 NMPC of the PVC Batch Reactor. 3.1.4.1 Multiple On-Line Linearization-Based NMPC of the PVC Batch Reactor. 3.1.4.2 Sequential NMPC of the PVC Batch Reactor. 3.1.5 Conclusions. 3.1.6 Nomenclature. 3.2 Modeling, Simulation, and Control of a Yeast Fermentation Bioreactor. 3.2.1 First Principle Model of the Continuous Fermentation Bioreactor. 3.2.2 Linear Model Identification and LMPC of the Bioreactor. 3.2.3 Artificial Neural Network (ANN)-Based Dynamic Model and Control of the Bioreactor. 3.2.3.1 Identification of the ANN Model of the Bioreactor. 3.2.3.2 Using Optimal Brain Surgeon to Determine Optimal Topology of the ANN-Based Dynamic Model. 3.2.3.3 ANN Model-Based Nonlinear Predictive Control (ANMPC) of the Bioreactor. 3.2.4 Conclusions. 3.2.5 Nomenclature. 3.3 Dynamic Modeling and Control of a High-Purity Distillation Column. 3.3.1 Introduction. 3.3.2 Dynamic Modeling of the Binary Distillation Column. 3.3.2.1 Model A: 164th Order DAE Model. 3.3.2.2 Model B: 84th Order DAE Model. 3.3.2.3 Model C: 42nd Order ODE Model. 3.3.2.4 Model D: 5th Order ODE Model. 3.3.2.5 Model E: 5th Order DAE Model. 3.3.2.6 Comparison of the Models. 3.3.3 A Computational Efficient NMPC Approach for Real-Time Control of the Distillation Column. 3.3.3.1 NMPC with Guaranteed Stability of the Distillation Column. 3.3.3.2 Direct Multiple Shooting Approach for Efficient Optimization in Real-Time NMPC. 3.3.3.3 Computational Complexity and Controller Performance. 3.3.4 Using Genetic Algorithm in Robust Optimization for NMPC of the Distillation Column. 3.3.4.1 Motivation. 3.3.4.2 GA-Based Robust Optimization for NMPC Schemes. 3.3.5 LMPC of the High-Purity Distillation Column. 3.3.6 A Comparison Between First Principles and Neural Network Model-Based NMPC of the Distillation Column. 3.3.7 Conclusions. 3.3.8 Nomenclature. 3.4 Practical Implementation of NMPC for a Laboratory Azeotropic Distillation Column. 3.4.1 Experimental Equipment. 3.4.2 Description of the Developed Software Interface. 3.4.3 First Principles Model-Based Control of the Azeotropic Distillation Column. 3.4.3.1 Experimental Validation of the First Principles Model. 3.4.3.2 First Principle Model-Based NMPC of the System. 3.4.4 ANN Model-Based Control of the Azeotropic Distillation Column. 3.4.5 Conclusions. 3.5 Model Predictive Control of the Fluid Catalytic Cracking Unit. 3.5.1 Introduction. 3.5.2 Dynamic Model of the UOP FCCU. 3.5.2.1 Reactor Model. 3.5.2.2 Regenerator Model. 3.5.2.3 Model of the Catalyst Circulation Lines. 3.5.3 Model Predictive Control Results. 3.5.3.1 Control Scheme Selection. 3.5.3.2 Different MPC Control Schemes Results. 3.5.3.3 MPC Using a Model Scheduling Approach. 3.5.3.4 Constrained MPC. 3.5.4 Conclusions. 3.5.5 Nomenclature. 3.6 Model Predictive Control of the Drying Process of Electric Insulators. 3.6.1 Introduction. 3.6.2 Model Description. 3.6.3 Model Predictive Control Results. 3.6.4 Neural Networks-Based MPC. 3.6.4.1 Neural Networks Design and Training. 3.6.4.2 ANN-Based MPC Results. 3.6.5 Conclusions. 3.6.6 Nomenclature. 3.7 The MPC of Brine Electrolysis Processes. 3.7.1 The Importance of Chlorine and Caustic Soda. 3.7.2 Industrially Applied Methods for Brine Electrolysis. 3.7.3 Mathematical Model of the Mercury Cell. 3.7.3.1 Model Structure. 3.7.3.2 The Main Equations of the Mathematical Model. 3.7.4 Mathematical Model of Ion-Exchange Membrane Cell. 3.7.4.1 Model Structure. 3.7.4.2 The Main Equations of the Mathematical Model. 3.7.5 Simulation of Brine Electrolysis. 3.7.5.1 Simulation of the Mercury Cell Process. 3.7.5.2 Simulation of the Ion-Exchange Membrane Cell Process. 3.7.6 Model Predictive Control of Brine Electrolysis. 3.7.6.1 MPC of Mercury Cell. 3.7.6.2 MPC of IEM Cell. 3.7.7 Conclusions. 3.7.8 Nomenclature. Index.

About the Author :
Professor Paul Serban Agachi graduated 1970 in Control Engineering at the Politehnica University of Bucharest and obtained his Ph.D. in Chemical Engineering from the University for Petroleum & Gas in Ploiesti, Romania. His professional experience ranges from design engineer and system analyst in process control design to researcher in fuel cells, process modeling, optimization and control, and also professor of process control at the Department of Chemical Engineering of Babes-Bolyai University, Cluj-Napoca. He was visiting associate at California Institute of Technology, invited professor at E?s Lorand University, UNESCO Higher Education consultant, member of the Academy of Technical Sciences of Romania, chair of CAPE Forum 2005, and co-chair of ESCAPE 17. He has published 7 books and 85 scientific papers. Zoltan K. Nagy received his M.S. and Ph.D. degrees in chemical engineering from Babes-Bolyai University of Cluj-Napoca in 1995 and 2001, respectively, where is currently working. Between 1999 and 2005 he was research associate and visiting lecturer in different international research teams, e.g., at ETH Zurich the University of Heidelberg, the University of Stuttgart, and the University of Illinois at Urbana-Champaign. He worked on industrial implementation of model-based control strategies with companies such as BASF and ABB, and has published over 60 papers in the field. Cristea Vasile Mircea graduated the Faculty of Electrotechnics, Romania, with specialization on process control and computer science and holds a Ph.D. degree in process control. After 8 years spent in industry he is at present Associate Professor at Babes-Bolyai University, Cluj-Napoca; his interests lie in systems theory, chemical process control, advanced process control, data acquisition and control, linear and nonlinear model based predictive control, and fuzzy control. He was director of CNCSIS Projects and has published 3 books as well as over 55 scientific papers. Arpad Imre-Lucaci received his M.S. and Ph.D. degrees in chemical engineering from Babes-Bolyai University of Cluj-Napoca in 1985 and 1999, respectively. Since 1988 he has worked in the Chemical Engineering Department of BBU Cluj-Napoca, Romania, and spent research stays at University of Stuttgart (1994) and ETH Zurich (in 2002 and 2003). His main research fields are mathematical modeling, simulation and optimization in process industries, on which he has published over 20 scientific papers.