Occupancy Detection Using a Temperature-Sensitive Adaptive Algorithm

Abstract

By using a liquid crystal shutter to periodically modulate the infrared energy received by the on-board passive infrared (PIR) sensor, the synchronized low energy electronically chopped PIR sensor recently developed by our team can detect stationary occupancy. However, the fixed threshold value-based detection accuracy is largely dependent on environmental settings, such as the room and floor surface temperature. In this letter, we first developed a thermoelectric model to quantify the environmental impact on the threshold value and then created an adaptive algorithm to generate temperature-sensitive threshold values to compensate environmental effects. We validated our thermoelectric model and tested our adaptive algorithm in two uncontrolled environmental settings, and results show that the detection accuracy reaches over 92%, which is much higher than that using the fixed threshold approach.