Spring-Rod System Identification via Differentiable Physics Engine

TL;DR

This paper introduces a differentiable physics engine for system identification of spring-rod assemblies, enabling efficient, explainable parameter learning with less data compared to black-box methods.

Contribution

It presents a modular, physics-based differentiable engine that simplifies parameter estimation through linear regression, reducing data needs and enhancing interpretability.

Findings

Effective identification of spring and mass parameters

Reduced data requirements compared to black-box methods

Successful application to tensegrity systems like NASA's icosahedron

Abstract

We propose a novel differentiable physics engine for system identification of complex spring-rod assemblies. Unlike black-box data-driven methods for learning the evolution of a dynamical system \emph{and} its parameters, we modularize the design of our engine using a discrete form of the governing equations of motion, similar to a traditional physics engine. We further reduce the dimension from 3D to 1D for each module, which allows efficient learning of system parameters using linear regression. The regression parameters correspond to physical quantities, such as spring stiffness or the mass of the rod, making the pipeline explainable. The approach significantly reduces the amount of training data required, and also avoids iterative identification of data sampling and model training. We compare the performance of the proposed engine with previous solutions, and demonstrate its efficacy on tensegrity systems, such as NASA's icosahedron.