Indoor air quality and energy management in buildings using combined moving horizon estimation and model predictive control

We present a novel approach for energy-optimal control of indoor air quality in the presence of system-model mismatch. We develop a physics-based building model that predicts concentrations of indoor pollutants (ozone, formaldehyde, and particulate matter) as a function of time-varying outdoor concentrations and instantaneous indoor emissions. We use a combined moving horizon estimation (MHE) and model predictive control (MPC) approach for simultaneous control of indoor air pollutants and energy consumption related to a dedicated ventilation system (DVS). The impact of model inaccuracies on MPC performance is addressed by combining the MPC with an MHE to predict the model parameters at each time instant based on a series of past measurements. The control performance of the proposed framework is shown through a case study, that also considers the impact of location and seasonality.