Improved Occupancy Detection Accuracy using PIR and Door Sensors for a Smart Thermostat

Abstract

It is important to investigate the occupancy levels in buildings in order to achieve accurate and meaningful building simulations results. Thus, research on buildings occupants’ behavior has been conducted in various institutions, including the International Energy Agency (IEA) annex 66. In this paper, we studied the use of passive infrared (PIR) sensors along with a door sensor to ensure the accuracy of occupancy detection in the built environment. This experimental study was conducted in a testbed setting to determine the appropriate locations and number of PIR sensors needed to improve the accuracy of occupancy detection. An occupancy detection algorithm was suggested to improve the performance of the developed system. The study result indicated that a PIR sensor located on a wall provides more accurate detection rate compared to those located on the ceiling or on the wall beside the door. In terms of the optimal number of PIR sensors to be used, the accuracy of detection improved as the number of sensors increased. It is found that an accuracy rate of 91.5% (number of occupants: 1 to 6) was obtained when using all four sensors together. When the suggested occupancy algorithm was applied, an improved occupancy detection accuracy rate of 99.8% was achieved (number of occupants: 1 to 6), while using one PIR sensor located on the wall and a door sensor together in this specific study. Thus, applying the algorithm allowed to obtain accuracy rates better than when four sensors were used together.