Filling in the blanks: Liberzon to study observers for reconstructing variables in nonlinear systems
Elise King, Coordinated Science Lab
- ECE Associate Professor Daniel Liberzon is part of a research team that received a grant to study robust nonlinear observers.
- Nonlinear observers can reconstruct missing or hidden variables in nonlinear systems.
- Liberzon will focus on the quantized control, which is something that converts information from analog to digital, such as a camera.
The National Research Foundation of Korea (NRFK) recently awarded ECE Associate Professor Daniel M. Liberzon, a researcher at the Coordinated Science Laboratory (CSL), and Hyungbo Shim from Seoul National University a three-year, approximately $300,000 grant to research and develop robust nonlinear observers.
Shim is the principal investigator (PI) and Liberzon is the co-PI of the project titled “A New Approach to Design of Nonlinear Observers Robust to Measurement Disturbances, with Applications to Quantized Feedback Control.”
Nonlinear observers can reconstruct missing or hidden variables in a nonlinear system. In complicated processes, such as flying a plane, you can measure different quantities of that process, such as speed and velocity. However, sometimes variables can be missing -- that’s where an observer comes in. “This is a classic theory of control,” said Liberzon, a member of CSL's decision and control group.
Through their research, Liberzon and Shim aim to make nonlinear observers robust to measurement disturbances. Sometimes an observer has to reconstruct something based off of incorrect or imprecise measurements, Liberzon said. For example, in an eight megapixel camera, the picture may not be perfectly clear, and therefore all of the information may not be completely precise. A robust observer would be able to accurately reconstruct hidden variables even if it was given incorrect measurements.
Shim will focus on the robustness of the observer, while Liberzon will focus on the quantized control. A quantizer device is something that converts information from analog to digital, such as a camera. This process is where errors can occur, and that is why it must be controlled.
The grant officially began on October 1; however, Liberzon said that the collaboration with Shim has been ongoing for a while. Shim was a visiting professor at the University of Illinois in 2010, and spent a year at the University working with CSL's decision and control faculty.
When Liberzon decided he wanted to do this project, “I needed to go to an expert,” he said. So he went to Shim. After agreeing to do the project, Shim, then back in Korea, decided it would be best to apply for a grant through the National Research Foundation of Korea, or NRFK. “He led the proposal submission and writing effort,” Liberzon said.
NRFK has a global research network, Liberzon said, that supports researchers in Korea working with researchers abroad. While this project addresses the development of a robust nonlinear observer, the international collaboration also advances NRFK's goal of bringing Korean and American researchers together.