Thermal dynamic modelling and temperature controller design for a house

Heat consumption management and effective temperature control strategies to meet heat demand in residential and office buildings have become an important aspect within energy management. A thermal dynamic model of a building is not only necessary to estimate the energy consumption under different operating conditions but also to design effective controllers. This paper presents a classical control approach for the indoor temperature regulation of buildings. State-space and transfer function models of house thermal behaviour are developed. These are obtained from first principles of heat transfer and their analogy with electrical systems. To capture a realistic behaviour of heat transfer, the proposed models consider parametric uncertainties. A frequency response-based approach is used to obtain a reduced order system that facilitates control system design. The models have been implemented in MATLAB/Simulink and a PI controller has been designed to maintain a comfortable indoor temperature in the building. Simulation results show that the controller effectively regulates temperature despite system disturbances. An energy saving of around 8% comparing the proposed controller to a traditional on/off controller is achieved