Occupancy modeling in buildings for multi-occupant single zone scenario

• Main Author: Tam, Jia Liang
• Other Authors: Soh Yeng Chai
• Format: Final Year Project
• Language: English
• Published: 2015
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/64365
• Institution: Nanyang Technological University

Energy usage in buildings contributes to 41% of the total energy consumption apart from the industry and transportation sector and heating, ventilation and air conditioning (HVAC) systems consume approximately half of the total energy used by buildings. Knowledge of occupancy in a building is important as it has great influence on the energy consumption of the building. However, accurate prediction of occupancy in a building is challenging due to the random nature of human behavior. There has been a growing interest in modeling occupancy in buildings and various techniques have been developed to predict occupancy in a single zone and multiple zones scenarios. This project explores the use of an inhomogeneous Markov chain to predict occupancy in a multi-occupant single zone (MOSZ) scenario. The MOSZ scenario used in this project is confined to a lab in NUS. The states of the Markov chain used are the number of occupants in the lab. Simulations are conducted using MATLAB to predict occupancy based on actual occupancy measurements of the lab. The model’s performance is measured based on the mean occupancy profile, time of first arrival, cumulative occupied duration and the number of occupied and unoccupied transitions. The performance of the model is quantified using the normalized root mean square error (NRMSE). The model was seen to track the actual mean occupancy closely. The trend of the time of first arrival was also well followed except for some discrepancy. However, the trends of the cumulative occupied duration and the transitions between occupied and unoccupied states were not well followed. These could be remedied with a larger dataset.