Pramod P. Khargonekar

University of Michigan
Claude E. Shannon Professor of Engineering Science and Chair,
Dept. of Electrical Engineering and Computer Science,
NSF Young Investigator
IEEE Fellow

Mail: Department of EECS, University of Michigan, 1301 Beal Avenue, Ann Arbor, MI 48109-2122, USA
Tel.: (734) 764-4328
Secretary Tel.: (734) 764-3317
Fax: (734) 763-1503


Research Activities

My major research interests cover a broad range of topics in the general area of systems and control. Given the rapidly changing scenario in research, I believe that a healthy mix of theory and applications is perhaps the most appropriate strategy. Currently, I am most actively working with my students on the following topics:

In a major interdisciplinary research thrust, a group of faculty from the control systems laboratory and the solid state electronics laboratory has started a new research program on applications of control technology to semiconductor and display manufacturing processes. We have several projects in this area. I am the Center Director for the MURI Center on Intelligent Electronics Manufacturing. The main research thrusts of this center include modeling, sensing, and control of plasma etching and deposition processes. I am also a Thrust Group Leader in the Center for Display Technology and Manufacturing and a participating faculty member in the SRC Center of Excellence on Automated Semiconductor Manufacturing. These activities are also focused on applications of modern control and systems theory to plasma based microelectronics manufacturing processes. In our initial work, we have focused on real time and run-to-run control of plasma reactive ion etching process which is a major unit process step. Starting from a conceptual control strategy, we have designed and implemented a real time control system for this process on an Applied 8300 Hexode reactor. Experimental results have shown that real time feedback control could have major impact on the performance characteristics of the reactive ion etchers. More recently, we have developed nonlinear estimators for calculating etch rate from single/multiple wavelength laser interferometers. We have also developed a combined run-to-run and real-time controller for the RIE process.

In control theory, our research is analysis and synthesis of control systems that give optimal performance given an approximate model of the physical system to be controlled. This is an extremely important problem since in practice one rarely has an exact model for the physical system. Modeling uncertainty arises due to incomplete and inexact experimental data, simplifying approximations, neglected high frequency dynamics, etc. Our research has focused on the problem of robust stabilization for a variety of possible modeling uncertainties. Our results have yielded necessary and sufficient conditions for the solvability of robust stabilization problems and constructive algorithms for solving these problems. This research is also closely connected to our work on H-infinity optimal control. A new development in our research on robust control is the analysis and synthesis of robust sampled-data control systems. Here we have focused on developing analytical tools for robustness analysis and synthesis of sampled-data systems in H-two and H-infinity norm.

Models are the basis for much of control systems analysis and design. In a new major effort, we have begun research in empirical model building, i. e., system identification. We are exploring frequency domain techniques for building system models which are suitable for robust control analysis and design. We have applied some of this recent work to the reactive ion etching problem as well as to the JPL example. More recent work has focused structured nonlinear models, such as Hammerstein systems, from empirical data.

We are also actively pursuing applications of modern control and systems techniques to automotive systems. Our initial work focused on an idle speed fluctuation reduction problem where a novel nonlinear control strategy has given promising results on nonlinear simulations. More recently, we have focused on modeling and control of single and double transition shift transmissions. This work is being done in collaboration with engineers from General Motors.

In collaboration with engineers at Xerox and Professor Dan Koditschek, we have started a new project on control of the color xerography process.

For more details, you are invited to read our publications on these topics.

Selected Publications

Research Groups

Revised: October 1997 by Beth Olsen, November 2000 by djh

University of Michigan College of Engineering