Analysis, Identification, and Control of Rhythmic Dynamic Systems: From Robots to Humans
M. Mert Ankaralı, Johns Hopkins University
Date: Friday, January 16, 2015
Place: EA 409
This talk covers two recent projects on analysis and system identification of rhythmic dynamic systems. (1) Neglecting evident characteristics of a system can certainly be a modeling convenience, but it can also produce a better, more predictive model. I examined the consequences of neglecting (or not) bilateral asymmetries during human walking. Indeed, I showed that there are statistically significant asymmetries in the dynamics of human walking in healthy participants (N=8), but that by ignoring these asymmetries, I arrive at a more consistent and predictive model of human walking. (2) Rhythmic hybrid dynamic behaviors can be observed in a wide variety of biological and robotic systems. Powerful analytical and numerical tools exist in order to control and analyze such systems. Analytic modeling tools are limited to the case when we have a full (and preferably simple) mathematical model that can accurately describe the system dynamics. On the other hand, system identification provides a powerful complement for modeling and analyzing general dynamical systems. However, few tools exist for identifying the dynamics of rhythmic systems from input–output data. In order to advance the field, I proposed a new formulation for rhythmic hybrid dynamic systems using discrete time harmonic transfer functions that enables us to perform input–output system ID in frequency domain.
M. Mert Ankaralı received his B.Sc. degree in Mechanical Engineering and minor certificate in Mechatronics from the Middle East Technical University, Turkey in 2007, and 2008, respectively. He received his M.Sc. degree from the Electrical & Electronics Engineering Department of Middle East Technical University, Turkey in 2010. Throughout his M.Sc. studies at METU, he worked on developing high-performance control algorithms, physically realistic simulations, and analytical models for dynamically capable robotic platforms. He is currently in his last year as a PhD candidate in the Dept. of Mechanical Engineering, Johns Hopkins University. His research at Hopkins has been devoted to discovering the mechanisms by which the human nervous system controls rhythmic dynamic behaviors, which is essential for improving the quality of life of individuals suffering from motor deficits. He was also named to the 2015 class of Siebel Scholars, which is awarded annually for academic excellence and demonstrated leadership to 85 top students from world’s leading graduate schools.