Structured learning of rigid-body dynamics: A survey and unified view from a robotics perspective
This survey addresses the challenge of creating accurate and data-efficient dynamics models for robotics researchers and practitioners by unifying structured learning approaches.
This paper surveys supervised regression models that integrate rigid-body mechanics with data-driven techniques to improve data efficiency and physical integrity in dynamics modeling. It analyzes latent functions and operators in rigid-body mechanics to provide a unified view on combining data-driven models with analytical priors.
Accurate models of mechanical system dynamics are often critical for model-based control and reinforcement learning. Fully data-driven dynamics models promise to ease the process of modeling and analysis, but require considerable amounts of data for training and often do not generalize well to unseen parts of the state space. Combining data-driven modelling with prior analytical knowledge is an attractive alternative as the inclusion of structural knowledge into a regression model improves the model's data efficiency and physical integrity. In this article, we survey supervised regression models that combine rigid-body mechanics with data-driven modelling techniques. We analyze the different latent functions (such as kinetic energy or dissipative forces) and operators (such as differential operators and projection matrices) underlying common descriptions of rigid-body mechanics. Based on this analysis, we provide a unified view on the combination of data-driven regression models, such as neural networks and Gaussian processes, with analytical model priors. Further, we review and discuss key techniques for designing structured models such as automatic differentiation.