Symbolic Learning of Interpretable Reduced-Order Models for Jumping Quadruped Robots

TL;DR

This paper introduces a method combining autoencoders and symbolic regression to automatically learn interpretable, low-dimensional models for quadruped robot jumping, outperforming handcrafted models in simulation and hardware.

Contribution

It presents a novel approach that automatically derives task-specific, interpretable models directly from data using a hybrid autoencoder and symbolic regression, tailored for quadruped jumping.

Findings

Symbolic models outperform handcrafted aSLIP baseline.

Method works across multiple robots and jumping styles.

Models are interpretable and physics-aligned.

Abstract

Reduced-order models are central to motion planning and control of quadruped robots, yet existing templates are often hand-crafted for a specific locomotion modality. This motivates the need for automatic methods that extract task-specific, interpretable low-dimensional dynamics directly from data. We propose a methodology that combines a linear autoencoder with symbolic regression to derive such models. The linear autoencoder provides a consistent latent embedding for configurations, velocities, accelerations, and inputs, enabling the sparse identification of nonlinear dynamics (SINDy) to operate in a compact, physics-aligned space. A multi-phase, hybrid-aware training scheme ensures coherent latent coordinates across contact transitions. We focus our validation on quadruped jumping-a representative, challenging, yet contained scenario in which a principled template model is especially valuable. The resulting symbolic dynamics outperform the state-of-the-art handcrafted actuated spring-loaded inverted pendulum (aSLIP) baseline in simulation and hardware across multiple robots and jumping modalities.