Wireless sensor system for detection of occupants to increase building energy efficiency

Building climate control efficiency can be improved through accurate occupancy information. Schedules alone, particularly at a university with cancelled classes, may provide inaccurate occupancy information and climate control. This work proposes system of occupancy detection using wireless sensors to detect energy from cellular devices used universally by occupants. Cellular signals at the sensors are detected using Empirical Mode Decomposition (EMD), and the Received Signal Strength (RSS), time, channel and sensor information is sent to a fusion server. This server aggregates this information and applies a generalized version of Improved Prior Measurement Comparison (PMC+) to estimate occupant locations in the building. These locations are used directly to estimate occupancy. This occupancy information is sent to a Building Control system, which fuses the estimated occupancy information with schedule information to provide an occupancy estimate of improved accuracy. An improvement from 40% error with schedule alone to 15% error with fused information was observed, a significant improvement in occupancy accuracy.