On the impacts of occupancy sensing on advanced model predictive controls in commercial buildings

摘要

Advanced optimization-based control, such as Model Predictive Control (MPC), has been shown to achieve increased energy savings and thermal comfort across different building types and climatic conditions. However, the success of such control algorithms is typically contingent on factors, such as sensor measurements used for the advanced control implementation. In this paper, we specifically investigate the role of occupancy sensors in improving the performance of MPC for a large office building situated in Chicago, IL weather conditions with VAV type HVAC system. Utilizing a detailed simulation-based study involving Occupancy-Based Model Predictive Control (OB-MPC), we infer that occupancy sensing could enable higher energy savings (-.5 on summer days) and improve thermal comfort relative to a baseline MPC that does not utilize occupancy information. Additionally, we investigate the impact of using occupancy presence sensors versus counting sensors, different levels of occupant density, and varying weather conditions on the control performance. Finally, we perform a systematic investigation of the impact of sensor non-idealities (specifically bias and latency errors) on the performance of OB-MPC algorithms. The study indicates that measurement bias could slightly degrade the realizable benefits of using occupant counting-based MPC (up to-.1 reduction in energy savings and up to 2X increase in thermal discomfort compared to ideal sensing). However, measurement latency may not impact the control performance.