Symbolic Learning of Interpretable Reduced-Order Models for Jumping Quadruped Robots
This work addresses the need for simplified yet accurate dynamic models for motion planning and control in quadruped robots, specifically for jumping tasks, representing an incremental improvement over existing methods.
The paper tackles the problem of deriving interpretable reduced-order models for jumping quadruped robots by introducing a learning architecture that combines Sparse Identification of Nonlinear Dynamics with physical priors, achieving superior accuracy compared to the traditional actuated Spring-loaded Inverted Pendulum model in simulations and hardware experiments.
Reduced-order models are essential for motion planning and control of quadruped robots, as they simplify complex dynamics while preserving critical behaviors. This paper introduces a novel methodology for deriving such interpretable dynamic models, specifically for jumping. We capture the high-dimensional, nonlinear jumping dynamics in a low-dimensional latent space by proposing a learning architecture combining Sparse Identification of Nonlinear Dynamics (SINDy) with physical structural priors on the jump dynamics. Our approach demonstrates superior accuracy to the traditional actuated Spring-loaded Inverted Pendulum (aSLIP) model and is validated through simulation and hardware experiments across different jumping strategies.