Occupancy-driven intelligent control of HVAC based on thermal comfort

Abstract

Nowadays, the building sector is a substantial consumer of world’s energy. The dominant energy share of Heating, Ventilation and Air-Conditioning (HVAC) systems, makes it the focus of research for saving energy. Current air conditioning systems often rely on maximum occupancy assumptions and fixed schedules to maintain sufficient comfort level. Having information regarding occupancy situation may lead to significant energy-savings. On the other hand, focusing on the reduction of energy only, may lead to sacrificing the thermal comfort of the occupants in a building. Moreover, due to the difference of preference of thermal comfort of individuals, particularly in a shared space, a fixed set point for HVAC systems, can cause discomfort. Therefore, a comprehensive technique is required to save energy while maintaining thermal comfort. The present research proposes an occupancy-driven HVAC control system based on thermal comfort analysis. A ZigBee-based indoor localization system is developed to monitor the location of occupants inside the buildings. Algorithms are used to improve the accuracy of positioning system, which include Near Neighbour Area (NNA), Principle Component Analysis (PCA) and Exponential Moving Average algorithms (EMA). Computational Fluid Dynamics (CFD) is used to simulate the thermal comfort through modelling the indoor air distribution and flows. Wind velocity and temperature are simulated in several scenarios and the Predicted Mean Vote (PMV) and the Predicted Percentage Dissatisfied (PPD) are computed. The simulation results are verified through a survey asking for occupants’ real feelings and consequently thermal comfort zones are identified with associated occupants, which are used for possible energy saving while providing satisfied level to all the occupants. To investigate different satisfaction feeling of occupants, a personalized thermal profile is created for individuals inside the test bed area. A fuzzy based approach is used to develop a fuzzy map of each occupant and as a result, a personal thermal preference profile is created. Based on the present occupants in the room, the minimum and maximum preferred temperatures are estimated and used for controlling the HVAC system. The Semi-hidden Markov chain method is used to create the occupants’ behavioural pattern which can reduce the frequencies of turning ON or OFF the HVAC systems. The real-time locations of the persons, estimated based on the NNA and MA localization method, are combined with their behavioural patterns and thermal preference profiles and their comfort zones to control the corresponding HVACs. The proposed method has been implemented to a shared office occupied by nine users and equipped with two individual air conditioners. The comparison of different control strategies show that the proposed intelligent control has a significant potential of saving energy and at the same time maintaining occupants in a reasonable thermal comfort range.