Modeling occupancy distribution in large spaces with multi-feature classification algorithm

Abstract

Occupancy information enables robust and flexible control of heating, ventilation, and airconditioning (HVAC) systems in buildings. In large-sized spaces, multiple HVAC terminals are usually installed to provide cooperative services for different thermal zones. Occupancy information for large spaces, such as people counts, determines the cooperation among terminals. However, a people count at room-level is not adequate to optimize HVAC system operation due to occupant movement within the room leading to uneven distribution of loads. Without an accurate knowledge of occupants' spatial distribution, uneven distribution of occupants often results in under-cooling/heating or over-cooling/heating in some thermal zones. Therefore, the lack of high-resolution occupancy distribution is often perceived as a bottleneck for future improvement in HVAC operation efficiency. To fill this gap, this study proposes a multi-feature based k-Nearest-Neighbors (k-NN) classification algorithm to extract occupancy distribution through reliable and low-cost Bluetooth Low Energy (BLE) networks. To demonstrate the proposed methods, an on-site experiment was conducted in a typical office of an institutional building. To validate the detection accuracy, the experiment outcomes were examined in three case studies, and one method based on City Block Distance (CBD) is used to measure the distance between detected occupancy distribution and ground truth and assess the results of occupancy distribution. The results show that the accuracy of CBD = 1 is over 71.4% and accuracy of CBD = 2 can reach 92.9%.

Modeling occupant behavior in buildings

“…The description of the localization of occupants in real-time is fundamental for a large variety of smart buildings services; specifically, energy management and indoor environmental control [73]. For example, the proper load calculation due to occupants and their spatial distribution could avoid over-heating/cooling or under-heating/cooling of areas which is of a major importance especially for large public spaces [72,74]. Furthermore, these models could help to track and learn inhabitant's daily routine unobtrusively with the aim to optimize energy usage without affecting occupants' comfort [75].…”

“…The second task is usually performed with machine-learning algorithms that are able to learn representation from the data and use them to forecast, simulate and model the occupants' presence in rooms and their movements [74,75,90,91]. Some studies solve the simulation and forecasting via stochastic models, due to the lack of surveys and statistical information with proper detail [72,92].…”

Occupant Activity Detection in Smart Buildings: A Review

“…A beacon (or iBeacon for Apple) is a device able to emit BLE signals, which can be captured by mobile applications. Hence, in a smart building scenario, the use of occupants mobile devices (with BLE-based beacons technology) as a source of information represents an effective solution to accurately detect occupancy, with energy efficient methods [61,62,63,64]. The work presented in [65] proposes a modification of the iBeacon protocol to change the way the beacons advertise the region associated with them.…”

Cross-source sensing data fusion for building occupancy prediction with adaptive lasso feature filtering

“…CO2 [7], lighting [8], PIR [9], Bluetooth [10,11], Wi-Fi [12,13], and so on. However, with development of sensor and information technologies, to improve the accuracy and robustness of occupancy detection and prediction, occupancy estimation with multiply sensors/parameters fusion is a significant trend instead of by a single parameter [14][15][16].…”