Gradient-Based Trajectory Optimization With Learned Dynamics
This work addresses the challenge of model inaccuracies in robotics for tasks requiring precise control, though it is incremental as it combines existing trajectory optimization and model learning techniques.
The paper tackled the problem of performing dynamic robotic tasks without accurate analytical dynamics models by using a neural network to learn differentiable dynamics from data, achieving accurate modeling of highly nonlinear behaviors over large time horizons with only 25 minutes of interaction data on two robots.
Trajectory optimization methods have achieved an exceptional level of performance on real-world robots in recent years. These methods heavily rely on accurate analytical models of the dynamics, yet some aspects of the physical world can only be captured to a limited extent. An alternative approach is to leverage machine learning techniques to learn a differentiable dynamics model of the system from data. In this work, we use trajectory optimization and model learning for performing highly dynamic and complex tasks with robotic systems in absence of accurate analytical models of the dynamics. We show that a neural network can model highly nonlinear behaviors accurately for large time horizons, from data collected in only 25 minutes of interactions on two distinct robots: (i) the Boston Dynamics Spot and an (ii) RC car. Furthermore, we use the gradients of the neural network to perform gradient-based trajectory optimization. In our hardware experiments, we demonstrate that our learned model can represent complex dynamics for both the Spot and Radio-controlled (RC) car, and gives good performance in combination with trajectory optimization methods.