Improving Thermal Comfort in Smart Buildings by an Artificial Immune System

Nowadays most HVAC (Heating, Ventilation and Air Conditioning) systems for residential buildings employ a singlezone, two-position control system which is simplistic. According to statistical studies, people spend 80% of their lives in buildings. Corresponding to the increasing demands for environment, comfort, energy, and productivity, advanced methods are applied for improving thermal conditions in residential buildings thanks to the dramatically rapid development of information and artificial intelligence technologies. Certainly, this kind of control system is a classic example of CyberPhysical Systems (CPS), which are integrations of computation with physical processes, where embedded computers and networks monitor, control and affect the physical processes and vice versa [4]. In this work, based on the artificial immune system (AIS) theory [1, 2] and its advantageous properties, we try to build an intelligent system to keep the inside air temperature comfortable while reducing energy consumption [5]. The contribution of this work is threefold : – An AIS architecture for intelligent thermal control of residential buildings is proposed, based on which it is more appropriate to build a CPS. – An adaptive reward function to update the affinities between antibodies and antigens is designed. – The AIS approach is compared with the traditional baseline control method, namely two-position control, and their experimental results are analysed.