Control of naturally ventilated buildings: a model predictive control approach
This project is studying possible techniques for improving the control of naturally ventilated buildings. Control of naturally ventilated buildings can be complex and requires automation of windows and vents. Due to the use of natural driving forces such as wind and buoyancy to drive flow, and typically high amounts of thermal mass to stabilise indoor temperatures, conditions are hard to predict accurately and buildings can be slow to respond. Additional problems such as poorly placed sensors, changes in occupancy and lack of proper calibration can result in unsuitable internal conditions and excess energy usage.
This project aims to investigate the potential advantages of a model predictive control (mpc) strategy. By using an empirical model of the building to predict the internal conditions at future time instants an optimal control sequence can be determined which minimises energy consumption and maintains a suitable internal environment for occupants.