Hahn Research Group

Answers to Frequently Asked Questions:

Why work in the systems area?

Essentially, research in the systems area studies the behavior of dynamic systems. Since every system found in nature exhibits dynamic behavior in one form or another it is important to develop detailed models of a process in order to make predictions about its behavior under different operating conditions. These models also allow us to design and operate processes more economically or in an environmentally friendlier way. Additionally, it enables us to study the behavior of processes that we were previously unable to study.
Students who work in the systems area will develop a good background in mathematics, a strong understanding for the process under investigation, and will develop computer skills needed to perform the investigation. In many cases the research will be carried out in collaboration with industry, which offers the possibility for summer internships. Graduates in the systems area are highly sought by industry, often accepting jobs more than one year prior to completion of their degrees.

Why is it important to study dynamic systems?

In today's economy the production of high purity speciality chemicals and other speciality products (pharmaceuticals, semiconductor devices, etc.) has increased dramatically. Many of these products are manufactured in batch processes, which can not be described by steady state models.
Furthermore, due to increasing energy costs as well as more stringent environmental and safety regulations, it is becoming ever more important to consider the transient behavior of processes as well.
Additionally, there are many emerging biomedical applications for which dynamic models are required. One example is the drug infusion problem. A patient's response to a drug infusion will never reach a steady state. Instead, it is important to guarantee that certain parameters will stay within desired bounds for the whole duration of the drug infusion, while one has to make sure that the patient will not be overdosed at any time. This problem can be investigated with dynamic models, whereas steady state models would not result in sufficient information to guarantee this.
The above mentioned reasons are just some of many for answering the question. Due to the fact that the available computational power is constantly increasing and due to the availability of powerful simulation and optimization algorithms, it is now possible to simulate and analyze many dynamic processes that were intractable only a few years ago.

How is your research related to what I learn in an undergraduate process control class?

In an undergraduate process control class you are introduced to studying the dynamic behavior of systems described by linear time invariant differential equations (if the system is described by nonlinear differential equations then it is usually linearized in order to transform it into the desired linear form). You also learn how you can design a single-loop PID controller for this type of system.
However, all processes that are found in nature are essentially nonlinear. In some cases, linearizing the system can be a good assumption, but this is not always the case. This is especially true for batch processes that are never operated at a steady state or for processes that are described by population balance models (crystallization, bioreactors, polymerization) because these unit operations tend to exhibit strongly nonlinear behavior. Studying the dynamic behavior of nonlinear systems and designing controllers for nonlinear processes is an important task that we address in our research group. Also, most plants have dozens to hundreds of control loops that are intertwined with one another. Designing single-loop controllers and thereby neglecting the interactions that exist between the control loops often results in poor performance and economical loss. Part of our research focuses on methods that facilitate designing controllers for such systems with mutiple inputs and multiple outputs.

What is different about research in your group compared to research groups at other universities?

In our research group we take an approach that integrates theoretical development with simulations, and experimental investigation, including collaboration with industry. The work that is performed in our group focuses on the modeling, control, and optimization related issues. However, we closely collaborate with other research groups that carry out the experimental aspects of these problems. We have strong ties with industry via several ongoing collaborations. This enables us to focus on our strengths, while it guarantees that the research results are directly applicable in practice.