Review on occupancy detection and prediction in building simulation

Energy simulation results for buildings have significantly deviated from actual consumption because of the uncertainty and randomness of occupant behavior. Such differences are mainly caused by the inaccurate estimation of occupancy in buildings. Therefore, the error between reality and prediction could be largely reduced by improving the accuracy level of occupancy prediction. Although various studies on occupancy have been conducted, there are still many differences in the approaches to detection, prediction, and validation. Reports published within this domain are reviewed in this article to discover the advantages and limitations of previous studies, and gaps in the research are identified for future investigation. Six methods of monitoring and their combinations are analyzed to provide effective guidance in choosing and applying a method. The advantages of deterministic schedules, stochastic schedules, and machine-learning methods for occupancy prediction are summarized and discussed to improve prediction accuracy in future work. Moreover, three applications of occupancy models-improving building simulation software, facilitating building operation control, and managing building energy use-are examined. This review provides theoretical guidance for building design and makes contributions to building energy conservation and thermal comfort through the implementation of intelligent control strategies based on occupancy monitoring and prediction.

Automated summary

The study focuses on the deviation of energy simulation results for buildings from actual consumption, attributing it mainly to inaccurate estimation of occupancy. Improving the accuracy level of occupancy prediction can notably reduce the error between reality and prediction. Despite various studies on occupancy, differences in detection, prediction, and validation methods still exist.