Stochastic scheduling for commercial building cooling systems: considering uncertainty in zone temperature prediction

This paper presents the first attempt to address uncertainty in zone temperature prediction with stochastic optimization. The uncertain is a process and has not been considered existing optimization for building control. To fill this gap, we proposed novel formulation of handle Specifically, examined accuracy typical linear model predicting temperature. We then formulated scheduling cooling system as problem over 24-hour look-ahead period minimize electricity cost studied system. After that, applied load (SLS) direct expansion (DX) that serves medium office building. Through simulation detailed energy software, EnergyPlus, evaluated operational thermal comfort compared deterministic scheduling. operation was found vary level uncertainty. SLS can mitigate impacts predictions on both comfort. evaluation results indicate works better when more significant.