Sparse Control for Dynamic Movement Primitives

TL;DR

This paper introduces a novel sparse control approach for Dynamic Movement Primitives (DMPs), enabling efficient modulation of rhythmic behaviors and transitions, with theoretical analysis and practical validation on robotic walking tasks.

Contribution

It presents a new framework using contraction theory for sparse modulation of DMPs, including the development of sparsely-inhibited rhythmic DMPs (SI-RDMPs) for controlling rhythmic primitives.

Findings

SI-RDMPs effectively manage start-stop transitions in robotic walking.

The framework provides analytical insights into oscillator coupling with diverse frequencies.

Experimental results demonstrate improved control efficiency and flexibility.

Abstract

This paper describes the use of spatially-sparse inputs to influence global changes in the behavior of Dynamic Movement Primitives (DMPs). The dynamics of DMPs are analyzed through the framework of contraction theory as networked hierarchies of contracting or transversely contracting systems. Within this framework, sparsely-inhibited rhythmic DMPs (SI-RDMPs) are introduced to both inhibit or enable rhythmic primitives through spatially-sparse modification of the DMP dynamics. SI-RDMPs are demonstrated in experiments to manage start-stop transitions for walking experiments with the MIT Cheetah. New analytical results on the coupling of oscillators with diverse natural frequencies are also discussed.