Active Control of Indoor Environment Quality

This paper presents a methodology for the integrated control of Indoor Environmental Quality. The most important aspect of the methodology is the treatment of the task as a multi-objective optimization problem. Appropriate control actions that achieve reasonable tradeoffs among conflicting objectives are selected through multi-objective optimization in which predictive models are trained using machine learning techniques. The methodology is currently being implemented and tested in the Field Environmental Chambers (FEC) of the Department of Building, National University of Singapore.