A General Framework for Structured Learning of Mechanical Systems
This addresses the need for flexible modeling in robotics, offering a method that balances bias and variance, but it is incremental as it builds on existing gray-box concepts.
The authors tackled the problem of learning accurate dynamics models for robotic control by proposing a gray-box approach that combines neural networks with prior knowledge, showing improved data-efficiency and performance in model-based reinforcement learning on a simulated double pendulum.
Learning accurate dynamics models is necessary for optimal, compliant control of robotic systems. Current approaches to white-box modeling using analytic parameterizations, or black-box modeling using neural networks, can suffer from high bias or high variance. We address the need for a flexible, gray-box model of mechanical systems that can seamlessly incorporate prior knowledge where it is available, and train expressive function approximators where it is not. We propose to parameterize a mechanical system using neural networks to model its Lagrangian and the generalized forces that act on it. We test our method on a simulated, actuated double pendulum. We show that our method outperforms a naive, black-box model in terms of data-efficiency, as well as performance in model-based reinforcement learning. We also conduct a systematic study of our method's ability to incorporate available prior knowledge about the system to improve data efficiency.