Quantifying time delays in environmental, behavioral, and energy consumption data to understand occupant space usage patterns in residential building

For occupant-centric building operations, it is crucial to understand individual occupants’ space usage patterns. For example, during summer season, if the outdoor temperature rises, the indoor temperature increases as well. Occupants, upon perceiving this change, may turn on the air conditioning, leading to increased energy consumption and eventually a decrease in indoor temperature. The interactions between the environment, occupants, and energy consumption repeats over time, and each occupant has different indoor preferences. These temporal factors should be considered when operating a building that is centered on occupant.

The study aims to quantify the time delays between environmental, behavioral, and energy consumption data in residential buildings to understand the space usage patterns of individual occupants. In this study, data on the environment (indoor temperature), occupants (occupancy rate), and energy consumption (air conditioning power usage) were collected at 10-minute intervals over three weeks during the summer in detached houses in Korea. To identify time delays between the data, the Dynamic Time Warping (DTW) algorithm was utilized, adjusting the time axes of the two monitoring data sets to find the minimum distance.

This method allowed us to determine which specific time points in one monitoring dataset correspond to which time points in the other dataset. After optimal alignment, the time delays were calculated by matching each time point’s displacement. Analysis of data from a space occupied by a single person showed that the average time displacement between indoor temperature and air conditioning power usage was 6.54(approximately 65 minutes = 6.54 * 10 min), between occupancy rate and indoor temperature was 5.32(approximately 53 minutes = 5.32 * 10 min), and between air conditioning power usage and occupancy rate was 4.8(approximately 48 minutes = 4.8 * 10 min). These results can serve as indicators for quantifying the time lag between environmental changes and occupants’ perception and behavior, providing a basis for understanding individual occupants’ space usage patterns.

In conclusion, utilizing DTW can quantitatively identify the time delays between environmental, behavioral, and energy data. This study serves as a fundamental basis for occupant-centric building operations by quantifying the time delays between environmental, behavioral, and energy data related to individual occupants, helping to figure out specific space usage patterns and broader occupant tendencies. Future research will aim to develop behavioral prediction models for occupant-centric building operations based on the identified occupant behavior patterns.