Training Efficient Controllers via Analytic Policy Gradient
Nina Wiedemann,Valentin Wüest,Antonio Loquercio,Matthias Müller,Dario Floreano,Davide Scaramuzza,Nina Wiedemann,Valentin Wüest,Antonio Loquercio,Matthias Müller,Dario Floreano,Davide Scaramuzza
Control design for robotic systems is complex and often requires solving an optimization to follow a trajectory accurately. Online optimization approaches like Model Predictive Control (MPC) have been shown to achieve great tracking performance, but require high computing power. Conversely, learning-based offline optimization approaches, such as Reinforcement Learning (RL), allow fast and efficien...
