Matson, S. A. and Ramirez, W. F., "Optimal Operation of an Electric Arc Furnace, " Elec. Furnace Conf. Proc. 57, 719-730 (1999).
Tholudur, A., Ramirez, W. F., and McMillan, J. D., "Mathematical Modeling and Optimization of Cellulase Protein Production Using Trichoderma reesei RL-P37, " Biotech. Bioeng. 66, 1-15 (1999).
Thelen, T. V. and Ramirez, W. F., "Monitoring, Modeling, and Control Strategies for Expanded-bed Adsorption Processes, " Bioseparation 8, 11-31 (1999).
Lu, S. X., Ramirez, W. F., and Anseth, K. S., "Photopolymerized, Multilaminated Matrix Devices with Optimized Nonuniform Initial Concentration Profiles to Control Drug Release, " J. Pharm. Sci. 89, 45-51 (2000).
Klein, E. J. and Ramirez, W. F., "Consideration of Local Shadowing and Ion Beam Voltage Effects in the Prediction of a Surface Evolving Under Ion Milling, " J. Vacuum Sci. & Tech. A 18, 166-175 (2000).
Tholudur, A., Ramirez, W. F., and McMillan, J. D., "Interpolated Parameter Functions for Neural Network Models, " Computers and Chemical Engineering 24, 2545-2553 (2000).
Narayan, A. P. and Ramirez, W. F., "Fault Diagnosis for Air Contamination Events in Three-dimensional Environments, " Int. J. Control 73, 1592-1604 (2000).
Klein, E. J. and Ramirez, W. F., "State Controllability and Optimal Regulator Control of Time-delayed Systems, " Int. J. Control 74, 281-289 (2001).
Klein, E. J., Ramirez, W. F., and Hall, J. L., "A Common-path Heterodyne Interferometer for Surface Profiling in Microelectronic Fabrication, " Rev. Sci. Instr. 74, 2455-2466 (2001).
Klein, E. J., and Ramirez, W. F., ?Consideration of Local Shadowing and Ion Beam Voltage Effects in the Prediction of Surface Evolving Under Ion Milling, ? J. Vac. Sci. Tech. A, 18, 166-175 (2001).
Klein, E. J., and Ramirez, W. F., ?On-Line Control of Ion Milling Etch Depth Using a Heterodyne Interferometer and Optimal Parameter Estimation, ? A.I.Ch.E. Journal, 48, 311-323 (2002).
W.F. Ramirez, Process Control and Identification, Academic Press (1994).
W.F. Ramirez, Computational Methods for Process Simulation, Butterworths, 2nd edition (1997).
Optimal Control and Identification of Industrial Processes. Current projects are on Biotechnology Processes, Microelectronic Fabrication, Environmental Air Quality, and Enhanced Oil Recovery.
Optimal control and on-line identification studies are being conducted on lumped and distributed parameter systems. Specific projects include fermentation reactors, microelectronic processes, and pollution prevention using process optimization techniques.
Control of Biotechnology Processes: We have been applying optimal control and identification theory to bioreactor operation. We have developed optimal feeding policies for fed-batch reactors which show operation on multiple singular arcs. These singular conditions represent the multiple operational capabilities of biological organisms. We are modeling and optimizing the biomass system of ethanol production with NREL and hemoglobin production with Baxter Inc. We have developed a new neural network modeling technique of Parameter Function Modeling that allows us to use neural networks and experimental data to develop dynamic models that can be optimized using dynamic program-ming.
Control of Microelectronic Fabrication: We have studied the on-line control of three microelectronic manufacturing processes, namely optical lithography and wet and dry etching. For the optical lithography process we have developed on-line methods for the control of the line width for the final development phase of the process. For the wet etching process we have developed a new design concept that allows for the simultaneous control of the line width and the side wall angle. For dry etching we have implemented on-line optimal stopping times using a new interferometric measuring technique. These studies have been in collaboration with local industry.
Pollution Prevention: Working under a combined NSF and EPA grant we have used dynamic optimization to develop new operating strategies for electric arc furnaces used to recycle steel. Our strategy is to minimize the production of carbon dioxide, minimize the production of the by product iron oxide, while minimizing batch processing time. Results to date indicate that significant improvements are feasible. We have also started a project to develop new pollution prevention strategies for steady state processes based upon mixed integer optimization concepts. This project has started with the collaboration of the University of Newcastle in Australia during a sabbatical leave.
