Abstract: Full state feedback (FSF) controllers for series elastic actuators bridge the gap between impedance and admittance controllers. For robot manipulators-which must stably accomplish both stiff and soft behaviors-FSF controllers are ideal candidate joint controllers. Although previous work on FSF gain tuning has derived necessary and sufficient conditions on controller passivity with time-delay and derivative filtering, they fail to achieve their target impedance behavior due to the large nonlinear friction found in the high gear-reduction actuators of robot manipulators. In this talk, I will discuss the history, advantages, and disadvantages of adaptive friction models and friction observers. I will then present a novel, back-drivable friction compensator that combines an adaptive Karnopp friction model in parallel with dual disturbance observers for model training and disturbance attenuation. Low-speed time-domain demonstrations show an improvement in trajectory tracking for the combined compensator over separate modeled or unmodeled approaches.
Read more about the lab at https://stories.tamu.edu/news/2024/01/10/engineering-leads-collaboration-for-in-space-operations-technologies/
Park at Alamo Stadium - https://map.trinity.edu/?id=1858#!ct/48724,71654,71656,71657,71659,71660?m/991683?s/?mc/29.46316899999998,-98.48067500000002?z/18?lvl/0?share
Add Event to Calendar
Biography:
Zachary Moyer
PhD Student in Mechanical Engineering
