Building occupancy estimation using microwave Doppler radar and wavelet transform: Fid-Bau Portal

Building occupancy estimation using microwave Doppler radar and wavelet transform

Islam, Shekh M.M. / Droitcour, Amy / Yavari, Ehsan / Lubecke, Victor M. / Boric-Lubecke, Olga

Abstract Room-level occupancy estimation is a critical input for real estate management, including building utilization optimization, energy-efficient building control, building security, and occupant health and well-being. Knowing how many people are using a room at a time can enable businesses to use their real estate, energy and human capital more efficiently, reducing operating expenses as well as carbon footprint. Radar-based occupancy estimation is attractive because it is unobtrusive and does not introduce the privacy issues brought with video imaging-based sensors. In this paper, we present an occupant estimation approach based on continuous wave Doppler radar and wavelet-based signal processing techniques. Theoretical background provides a rational for using a wavelet-based time-frequency approach, and comprehensive simulation and experimental results demonstrate the effectiveness of the proposed approaches on a data set that includes 1–10 occupants. These results indicate that Doppler radar with wavelet-based signal processing may be an effective tool for occupant count in smart building applications.

Highlights • Non-contact, radar-based sensor with two novel algorithms estimates occupant count. • Simulation models radar-signal intermodulation with different numbers of occupants. • Algorithms are optimized to improve linearity of output with occupant count. • Both algorithms generate values that monotonically increase with number of occupants. • A single radar-based sensor can be used to obtain room-level occupant count.