CHE 494/561 Advanced Process Control (Spring 2013)
Meets on the ASU Tempe campus in PSH 552
1:30 -3:10 pm M W F, Session A (1/7/13 - 2/27/13)
ChE 494: 26953
ChE 561: 26952
Professor Daniel E. Rivera
School for the Engineering of Matter, Transport, and Energy (SEMTE)
Arizona State University, Tempe, AZ 85287-6106
Office: Engineering Research Center (ERC) Room 275
Undergraduate controls course or equivalent from any engineering discipline. Knowledge of basic linear algebra and complex number arithmetic is desirable. Prior experience with MATLAB/SIMULINK is helpful but not required.
The course will make substantial use of class and presentation notes developed by the instructor. The following texts are recommended:
B.A. Ogunnaike and W.H. Ray, 1994, Process Dynamics, Modeling, and Control, Oxford University Press, ISBN 0-19-509119-1.
D.E. Seborg, T.F. Edgar, & D.A. Mellichamp, 2004, Process Dynamics and Control, 2nd edition, Wiley and Sons, ISBN 0-471-00077-9.
Morari, M. and E. Zafiriou, 1987, Robust Process Control, Prentice-Hall, ISBN 0-13-782153-0.
Relevant chapters to the course from the first edition of Seborg, Edgar, and Mellichamp, as well as all sections of the Morari and Zafiriou text are available for download free of charge from the CACHE Virtual Process Control book (http://www.cse.sc.edu/~gatzke/cache/)
The course introduces students to advanced process control concepts with emphasis on methods and techniques with significant impact on industrial practice. Homework assignments are highly design-oriented and rely on MATLAB w/SIMULINK to provide students with a "hands-on" experience on the course material.
The course presents a wide variety of concepts and techniques for designing process control systems that extend beyond the standard classical feedback/PID tuning methods that form part of a typical undergraduate course. Advanced analog single loop control, digital (computer) control, and multivariable control analysis and design procedures are presented. These include: 1) a detailed presentation of the Internal Model Control design procedure for both continuous-time (analog) and discrete-time (digital) control systems, 2) how to make single-loop designs work in a multivariable system through the judicious design of decentralized and decoupled control systems, and 3) the design of control systems for constrained multivariable processes via Model Predictive Control. The application of these ideas to important problem areas that lie beyond process control (specifically, enterprise systems / supply chain management and adaptive interventions in behavioral health) will be presented. A brief overview of system identification techniques in support of advanced control system design will also be discussed, as time permits.
Homework: 6 - 8 problem sets (total) .
Examinations: One in-class midterm exam.
Computer Facilities: Students must have access to MATLAB with SIMULINK and the Control System, Signal Processing, and Model Predictive Control toolboxes.
Laboratory and Projects: There is no laboratory associated with this course. Students will be required to perform a final design assignment on a topic of industrial or research interest.
Course Outline by Topical Areas:
I. Modeling and Systems Overview
A.Dynamic modeling via conservation and accounting principles
B. Linearization and state-space model representation,
C. Transfer functions and frequency response review.
II. Analysis and design of analog SISO closed-loop systems
A. Objectives of feedback control; H_2 and H_infinity optimality criteria
B. Rigorous Internal Model Control design (feedback and feedback/feedforward) procedure for continuous systems.
C. IMC design using low-order models and relationships to PID control.
VII. Analysis and design of digital SISO closed-loop systems
A. Representation of discrete-time systems via z-transforms.
B. Performance and robustness of digital control systems; classical techniques
C. Internal Model Control design procedure for discrete-time systems
IV. Multi-input, multi-output systems
A. Variable selection and pairing for decentralized control.
B. Decoupler design and analysis.
C. IMC design procedure for MIMO first-order with delay plants.
V. Model Predictive Control
A. Moving horizon philosophy and state-space formulation of MPC
B. Unconstrained 2-norm MPC and constrained MPC via Quadratic Programming
C. MPC implementation aspects
VI. System Identification Overview