SYNCHRONIZATION AND STABILIZATION OF BIPEDAL WALKING
Topic: Stabilization | All
Séance du jeudi 29 mars 2018, Salle L224, 11h00-12h00.
Christine CHEVALLEREAU, IRCCYN, Nantes, CNRS
Walking on flat ground is essentially a periodic phenomenon. The objective of this presentation is to highlight the main characteristics of a gait that can naturally induce stability in walking (stability in the sense of convergence towards periodic motion). For a robot moving in one plane, the importance of vertical oscillations on stability for a robot controlled via a virtual constraint approach (i.e. without time reference) was shown several years ago. In the case of a robot evolving in 3D with a movement in the frontal and sagittal planes, we will show that the synchronization between these 2 movements depends on the conditions chosen for the support foot changes and we will propose conditions to generate stable control law for a very simplified model of robot (3D LIP). These results will be extended to stability conditions by integrating either a robot speed measurement or vertical oscillations into the approach. We will then show how these results can be extended to realistic models of humanoid robots.
Christine CHEVALLEREAU, IRCCYN, Nantes, CNRS
Walking on flat ground is essentially a periodic phenomenon. The objective of this presentation is to highlight the main characteristics of a gait that can naturally induce stability in walking (stability in the sense of convergence towards periodic motion). For a robot moving in one plane, the importance of vertical oscillations on stability for a robot controlled via a virtual constraint approach (i.e. without time reference) was shown several years ago. In the case of a robot evolving in 3D with a movement in the frontal and sagittal planes, we will show that the synchronization between these 2 movements depends on the conditions chosen for the support foot changes and we will propose conditions to generate stable control law for a very simplified model of robot (3D LIP). These results will be extended to stability conditions by integrating either a robot speed measurement or vertical oscillations into the approach. We will then show how these results can be extended to realistic models of humanoid robots.