home :: departments :: rising stars

Rising Stars

Optimizing operations across engineering

Eugenio Schuster, assistant professor of mechanical engineering and mechanics and recipient of an NSF CAREER Award, can look at any type of system and visualize how it is being controlled.

“Controls are everywhere in our lives,” says Schuster. “People use them without even knowing it.”

As an example, Schuster cites the everyday cruise control system, which measures and monitors an automobile’s speed. If that speed dips below the rate set by the driver, a microcontroller signals a mechanical device to depress the accelerator and send more gas to the engine.

Dynamic models have improved the defense and aerospace industries, says Eugenio Schuster, and can do the same for energy applications.

It may be tempting to call Schuster a “control freak,” but “enabler” is a more accurate label. Schuster works on control systems that are far more complex than cruise control and that have applications in every engineering discipline.

The CAREER Award supports Schuster’s work with the nonlinear control of plasmas in nuclear fusion, a project that has been reported in previous issues of Resolve. Schuster is also interested in controls for other forms of energy, control of aerospace and mechanical systems, optimal control of large experimental physics devices such as particle accelerators, and magneto-hydrodynamic (MHD) flow control.

“In my field, we try to control the dynamics of a system, or how it behaves over time,” says Schuster. “Our goals are usually stabilization and performance. We want to control systems so they run and respond faster and more reliably to commands. We do most of our work mathematically, exploiting the availability of differential equation models that predict a system’s behaviors. Modeling a system’s dynamics and carrying out high-performance computer simulations are significant parts of our work.”

Schuster’s collaborative work with Carlos Romero of Lehigh’s Energy Research Center aims to control emissions using catalytic converters at coal-fired power plants. Under the auspices of the New York State Energy Research and Development Authority, Schuster and Romero are designing optimal controllers to reduce emissions and improve the overall economics of the plants. The controllers, driven by mathematical models, enable the catalytic converters to adapt to changes in the system, which in turn allows emissions to be minimized while power production is maximized.

By injecting ammonia into the catalytic converters, plant operators can lower emissions. But this generates a waste product called ammonia slip, which can cause damage to the power plant. Schuster’s model-based controllers allow for slightly higher but permissible emissions while lowering ammonia levels. The controllers also account for and react to real-time changes in the plant, such as coal quality, while maintaining optimal operating efficiency.

With today’s overreliance on foreign oil, says Schuster, control systems have the potential to save millions of dollars for the U.S. energy sector.

“Coal-fired power plants could save significant amounts of money,” he says, “but while there’s a revival of interest in optimized configurations, it remains to be seen how much controls are used. Unlike the nation’s defense and aerospace industries, the fossil-fuel industry is historically somewhat conservative when it comes to applying sophisticated controls.

“The use of dynamic models has been one of the reasons for the extraordinary progress of the defense and aerospace industries. We need to do the same in the energy industry. It’s a pity not to exploit the models we develop.”

Schuster notes that systems that depend only on time can be modeled by ordinary differential equations, but those with dynamics depending both on time and space require partial differential equation (PDE) models. Schuster is leading a group of researchers in Lehigh’s Laboratory for Control of Complex Systems on PDE control. He was invited to present the results of this work at an NSF workshop at UCLA in early 2009.

“We want to control systems so they respond faster and more reliably to commands. Our goals are usually stabilization and performance.” —Eugenio Schuster