N. Lawrence Ricker,
Associate Chair
Professor of Chemical Engineering
Contact Information
| 365 Benson Box 351750 Seattle, WA 98195-1750 |
Phone: 206-543-8786 Fax: 206-685-3451 or 206-543-3778 E-mail: ricker@u.washington.edu |
Education
B.S., University of Michigan, 1970.
M.S., University of California (Berkeley), 1972.
Ph.D., University of California (Berkeley), 1978.
Research Interests
- Process control theory and applications
- Process monitoring and fault detection
- Nuclear fuel reprocessing material accountancy
Process Control and Optimization
As the chemical industry matures, companies are emphasizing waste reduction and the optimal use of raw materials and energy resources. Changes in process design are one way to improve efficiency. Other opportunities arise during normal operations.
For the last decade my group has been developing control algorithms for use in complex continuous and batch processes. Applications have included biological systems, municipal waste-treatment, and semiconductor materials production. The key idea is to develop a mathematical model that incorporates the dominant process features, then use the model directly in a control strategy. Such methods have come to be known as Model Predictive Control (MPC).
MPC offers significant improvements over conventional control methods. For example, the figure below compares a nonlinear version of MPC to a classical single-loop (SISO) strategy. The application is the Tennessee Eastman Industrial Challenge Process. The objective is to keep three variables within the limits shown as dashed horizontal lines. The strategies are equally good for product composition (%G and %H in the figure), but the conventional strategy (dotted lines) violates the limits on production. Also, it does a much poorer job of controlling reactor pressure, which is a critical variable from the point of view of safety and operating costs.
An outgrowth of this work is the MPC Toolbox for MATLAB (The MathWorks - co-authored with M. Morari), which is currently installed at over 1000 industrial and academic institutions world-wide. Current projects include the control and optimization of batch processes, and the use of "chemometric" techniques to derive the maximum benefit from process data.
Selected Recent Publications
Ricker, N. L. and Lee, J. H. "Nonlinear Model Predictive Control of the Tennessee Eastman Challenge Process," Computers Chem. Engng., 19, 961-981(1995).
Wu, C.; Danielson, J. D. S.; Callis, J. B.; Eaton, M.; Ricker, N. L. "Remote, in-line monitoring of emulsion polymerization of styrene by short wavelength near-infrared spectroscopy Part I: Performance during normal runs," Process Control and Quality, 8, 1-23 (1996); Part II: Performance in the face of process upsets," Process Control and Quality, 8, 25-40 (1996).
Pearsall, T. P.; Brown, N.; Ricker, N. L.; Johnson, M. "Flux monitoring and control in epitaxy by chemical vapor deposition," J. Crystal Growth, Vol. 188, 63-68 (1998).
Johnson, M. C., Poochinda, K.; Ricker, N. L.; Rogers, J. W. Jr.; and Pearsall, T. P.; "In situ monitoring and control of multicomponent gas-phase streams for growth of GaN via MOCVD," J. of Crystal Growth, Vol. 212, 11-20 (2000).
Srinivasan, B.; Primus, C. J.; Bonvin, D.; and Ricker, N. L.; "Run-to-run optimization via control of generalized constraints," accepted, Control Engineering Practice, 2001.
de Jong, S.; Wise, B. M.; and Ricker, N. L.; "Canonical partial least squares and continuum power regression," J. Chemometrics, Vol. 15, 85-100 (2001).
Recent M.S. Theses
Recent Ph.D. Dissertations
