Neural Parameter Estimation of RC Thermal Building Models for Model Predictive Control
The Pretrained Estimator is proposed, a novel neural parameter estimation approach that embeds the physical equations into a neural network's training process to estimate RC parameters and achieves the best prediction performance among all RC-based methods and yields the lowest and least variable MPC costs across build...
F. Raisch, T. Germann, Sang-woo Ham et al.
· 3 citations