A Method for model-reduction of nonlinear building thermal dynamics

We propose a method for model-reduction of a class of non-linear models that are relevant to modeling thermal dynamics of multi-zone buildings. These models can have large state-space dimension even for a moderate number of zones. Reduced order models of building thermal dynamics can be useful to model-based control for improving energy efficiency, especially to computationally intensive ones such as Model Predictive Control (MPC). Although there are a number of well-developed techniques for model reduction of LTI systems, the same cannot be said about non-linear systems. The method we propose exploits the linear portion of the model to compute a transformation (by using balanced realization) and a specific sparsity pattern of the non-linear portion to obtain the reduced order model. Simulations are presented with a four zone building model, which show that the prediction of the zone temperatures and humidity ratios by the reduced model is quite close to that from the full-scale model, even when substantial reduction of model order is specified.