Learning Sampling Distributions for Model Predictive Control

Jacob Sacks, Byron Boots

Sampling-based methods have become a cornerstone of contemporary approaches to Model Predictive Control (MPC), as they make no restrictions on the differentiability of the dynamics…