Encoding Physical Constraints in Differentiable Newton-Euler Algorithm
This work addresses the issue of unrealistic learned dynamics for robotics control, offering an incremental improvement by adding structure to parameters for better physical plausibility.
The paper tackled the problem of learning physically implausible dynamics parameters from data using a differentiable Newton-Euler Algorithm by incorporating physical constraints into the learning process, resulting in improved training speed and generalization for real-time inverse dynamics control on a 7-DOF robot arm in simulation and real-world experiments.
The recursive Newton-Euler Algorithm (RNEA) is a popular technique for computing the dynamics of robots. RNEA can be framed as a differentiable computational graph, enabling the dynamics parameters of the robot to be learned from data via modern auto-differentiation toolboxes. However, the dynamics parameters learned in this manner can be physically implausible. In this work, we incorporate physical constraints in the learning by adding structure to the learned parameters. This results in a framework that can learn physically plausible dynamics via gradient descent, improving the training speed as well as generalization of the learned dynamics models. We evaluate our method on real-time inverse dynamics control tasks on a 7 degree of freedom robot arm, both in simulation and on the real robot. Our experiments study a spectrum of structure added to the parameters of the differentiable RNEA algorithm, and compare their performance and generalization.