Occupancy of rooms in urban residential buildings by users in cold areas of China

Occupancy is used to represent the movements and locations of users among various zones of buildings, and it is the basis of all other daily energy consumption behaviors. This study investigated eight families in cold areas of China based on occupancy measurements obtained in four main rooms, i.e., living room, bedroom, kitchen, and bathroom. In particular, we analyzed the duration of user occupancy and hourly mean occupancy, and characterized their regular and random features. According to the results, we developed an event-based occupancy model using an inhomogeneous Markov chain, where the rooms were modeled and daily events were divided into three categories according to their randomness. We established a new method for conversion between event characteristic parameters and a transition probability matrix, as well as an overlap avoidance method for active events. The model was then validated using real data. The results showed that the model performed well in terms of two evaluation criteria. The model should improve the accuracy of simulations of occupancy.