| Start Date: | 12/5/2018 | Start Time: | 10:00 AM |
| End Date: | 12/5/2018 | End Time: | 12:00 PM |
|
Event Description
Sensory-mediated neural oscillators have been proposed as controllers for robotic flexible fish-like foils. However little is known about how to design these types of oscillators to drive the motions of flexible foils. Specifically, how to design the sensory feedback architecture to produce a desired oscillation profile is an ongoing research question. This design task is complicated by the fact that the oscillator profile of the oscillator-driven system is affected by the mechanical properties of the foil, which may vary over time in order to modulate swimming forces.
In this thesis, the relationship between the sensory feedback architecture and the mechanical properties of a flexible fish-like foil are examined to determine the role of sensory feedback during swimming. It is hypothesized that the sensory feedback architecture is modulated as the stiffness of the foil increases in order to make the system stable to perturbations. To test this hypothesis a robotic fish-like foil with variable stiffness properties will be designed and fabricated. Analysis of the system’s Floquet multipliers will be performed to measure and quantify the stability of the oscillator-driven system as the mechanical properties and sensory feedback architecture are modulated. A neural-based control methodology will be proposed to determine how to drive foils at their time-varying resonant frequency. The results of this work will shed light on the role of sensory feedback during swimming and offer insights into how to design neural-based oscillators to be adaptive controllers for flexible fish-like foils. |
| |