|
Biological inspiration: a route for locomotion| old_uid | 9288 |
|---|
| title | Biological inspiration: a route for locomotion |
|---|
| start_date | 2010/11/22 |
|---|
| schedule | 14h-16h |
|---|
| online | no |
|---|
| location_info | Tour 65, couloir 65/66, etage 3, salle de réunion 304 |
|---|
| summary | Generating efficient movement for legged robots is an engineering challenge because it requires the production of signals with the right frequencies, phases and amplitudes for multiple degrees-of-freedom (DOFs) and the skills to coordinate these multiple DOFs. Further, if efficiency and adaptivity is desired, it requires the processing of multi-modal and multi-dimensional sensory inputs; the ability to deal with dynamic, complex and unpredictable environments and adaptivity to changes in body properties. It is our motivation and belief, that nature provides us with some solutions to these problems, that is, to control a lot of elements, to form specific walking patterns and provides high autonomous adaptation to various environments.
We propose a bio-inspired architecture for adaptive and autonomous generation of complex motor behaviours, with application to articulated robots with many DOFs. We are particularly interested on autonomously generating trajectories for robotic systems considering autonomous adaptive robotic locomotion. These goals demand a multidisciplinary undertaking, combining insights of dynamical systems theory, computational neuroscience and robotics. We apply a mathematically rigorous
formalism, i.e. nonlinear dynamical systems for modeling how behaviours related to locomotion are programmed in the oscillatory feedback systems of Central Pattern Generators (CPGs) in the central nervous system. The Nonlinear dynamical systems approach has proven to be successful in many robotic applications.
In this presentation, I will focus on the generation of timed trajectories for a wheeled robot and adaptive quadruped and hexapod locomotion. |
|---|
| responsibles | <not specified> |
|---|
| |
|