Embedded single-board controller for Double Skin Facade : a co-simulation virtual test bed

Operational building performance of Double Skin Facades (DSFs) is highly sensitive on their control strategies. Due to the complexity of their operations, model-based controls could represent an opportunity to improve further the potential energy savings and occupant comfort. The present work explores the potential and applicability of rule and model-based control strategie, implemented on a facade-embedded single-board controller. This is demonstrated by means on a virtual experiment, carried out by means of a co-simulative framework implemented on the controller itself, to replicate the real world implementation. The embedded controller has different tasks, from signal processing from low-cost sensors, to performing parametric on-line simulation to feed in the control decision making algorithm, to sending the control signal to the different DSF actuators. Two different control strategies are developed for the implementation, and their performance compared by means of co-simulation on the single board controller: a) a Rule-Based Control strategy, based on ifelse rules depending on environmental conditions; b) a Model-Based Control strategy, aims at minimising the heating and cooling loads of the environment adjacent of the DSF, while maximising the autonomous Useful Daylight Illuminance. The results show that, compared to benchmark DSF configurations and control strategies, the MBC implemented on the embedded single board controller would be able to reduce total (heating, cooling and lighting) energy use between 11% and 18%, while improving useful daylight availability by 13% without introducing potential glare risk.