Using Nonlinear Normal Modes for Execution of Efficient Cyclic Motions in Articulated Soft Robots

Abstract

Thanks to their body elasticity, articulated soft robots promise to produce effective and robust oscillations with low energy consumption. This in turn is an important feature which can be exploited in the execution of many tasks, as for example locomotion. Yet, an established theory and general techniques allowing to excite and sustain these nonlinear oscillations are still lacking. A possible solution to this problem comes from nonlinear modal theory, which defines curved extensions of linear Eigenspaces called Eigenmanifolds. Stabilizing these surfaces is equivalent to exciting regular hyper-efficient oscillations in the robotic system. This paper proposes a first experimental validation of the Eigenmanifold stabilization technique. It also proposes a simple yet effective means of injecting energy into the system, so to sustain the oscillations in presence of damping. We consider as experimental setups a single robotic leg, and a full soft quadruped. Preliminary locomotion results are provided with both systems.

Experimental Closed-Loop Excitation of Nonlinear Normal Modes on an Elastic Industrial Robot

“…The results are stable hyper-efficient natural oscillations. However, these techniques have had a primarily theoretical development and were validated on simple systems and exclusively in simulation -except for the preliminary work [26].…”

Comparison and experimental validation of predictive models for soft, fiber-reinforced actuators

“…Within first principles-based models, two subcategories emerge. Della Santina et al (2018a) and others (Bruder et al, 2017(Bruder et al, , 2018bDella Santina et al, 2018b;Satheeshbabu and Krishnan, 2017;Tondu and Lopez, 2000) show that lumped-parameter models can be used in design and control tasks. Continuum mechanical formulations have recently been demonstrated to be useful in soft system design and control (Coevoet et al, 2017;Connolly et al, 2017;Deimel and Brock, 2016;Goury and Duriez, 2018;Moseley et al, 2016;Polygerinos et al, 2015b;Uppalapati et al, 2018).…”

Finding the rhythm: Humans exploit nonlinear intrinsic dynamics of compliant systems in periodic interaction tasks

“…In previous research, a similar experiment as presented here had suggested that a bang-bang control law could represent the human behavior [25], which showed to work robustly and optimally efficient in robotic systems [23, 24, 26]. Thus, this control strategy is likewise considered as a possible excitation approach of the human and applied in simulation to obtain a third baseline for comparison.…”