Gaussian Process-Based Model Predictive Control

This lecture covers Gaussian Process-Based Model Predictive Control (MPC) applied to a Hospital HVAC System. It explains the use of Gaussian processes to model phenomena, Bayesian statistics for modeling, training, and predictions, and the NMPC formulation. The instructor discusses crafting models from static functions to dynamical systems, advanced training details, and the NMPC formulation with non-linear equations. The lecture concludes with the benefits of using Gaussian processes for learning dynamics with uncertainty estimates, especially in optimizing HVAC plant operations and responding to time-varying electricity prices.