Incorporating occupancy data in scheduling building equipment: A simulation optimization framework

Avinash Pallikere, Robin Qiu, Parhum Delgoshaei, Ashkan Negahban

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


A common energy-conservation method involves specifying a schedule in the building automation system (BAS) so that equipment enter a sleep mode or low power mode during predetermined off-shift hours generally defined based on the expected or perceived occupancy schedule. This paper proposes a binary programing optimization model that incorporates actual occupancy patterns for different zones in the building as well as equipment interdependence to systematically determine the optimal schedule for each equipment while maintaining a minimum required service level to meet occupant needs. The model is then integrated into a simulation optimization framework, where historical or simulated occupancy data are used to determine the optimal frequency of schedule updates and the best design option for new buildings or retrofit projects. Through a real-world case study of a university building, we illustrate how the proposed approach harnesses historical occupancy data to select the best option for re-purposing the zones/spaces in the building. The results provide important practical insights by showing the significant potential to improve common practice that typically uses the exact same schedule for all equipment in the BAS. The results further show that the choice of the design option depends on the minimum required service level. The paper also illustrates the importance of updating equipment schedules in response to possible seasonal changes in occupancy patterns throughout the year. All codes and datasets are made available in an open repository to facilitate adoption by practitioners and enable reproducibility of the results for researchers.
Original languageEnglish (US)
JournalEnergy and Buildings
StatePublished - 2020


Dive into the research topics of 'Incorporating occupancy data in scheduling building equipment: A simulation optimization framework'. Together they form a unique fingerprint.

Cite this