Training Structured Mechanical Models by Minimizing Discrete Euler-Lagrange Residual
This work addresses a specific problem in robotics for more accurate model-based control, but it is incremental as it builds on existing SMM frameworks.
The paper tackles the problem of fitting Structured Mechanical Models (SMMs) to data for robotics by proposing a method that minimizes the discrete Euler-Lagrange residual, showing improved accuracy over conventional schemes in experiments on double-pendulum systems with and without noise.
Model-based paradigms for decision-making and control are becoming ubiquitous in robotics. They rely on the ability to efficiently learn a model of the system from data. Structured Mechanical Models (SMMs) are a data-efficient black-box parameterization of mechanical systems, typically fit to data by minimizing the error between predicted and observed accelerations or next states. In this work, we propose a methodology for fitting SMMs to data by minimizing the discrete Euler-Lagrange residual. To study our methodology, we fit models to joint-angle time-series from undamped and damped double-pendulums, studying the quality of learned models fit to data with and without observation noise. Experiments show that our methodology learns models that are better in accuracy to those of the conventional schemes for fitting SMMs. We identify use cases in which our method is a more appropriate methodology. Source code for reproducing the experiments is available at https://github.com/sisl/delsmm.