Modeling of occupant energy consumption behavior based on human dynamics theory: A case study of a government office building

Abstract

The diversity and complexity of non-physical factors make it difficult to predict occupant behavior accurately. It's also the main reason for the great difference between actual and simulated energy consumption. In this paper, the human dynamics method is introduced to establish the occupant energy consumption behavior (OECB) models of multiple offices in a government office building, and the possible generation mechanism of the OECB is discussed to analyze its underlying reason for improvement of energy consumption system design. The results show the models of different single/multi-person office rooms can be well fitted by non-exponential distributions, such as truncated power-law distribution, stretched exponential distribution and log-normal distribution. Among them, the truncated power-law distribution model is the most suitable model for fitting the temporal statistical characteristics of OECB, and the power exponents of the models are close to 1. Furthermore, the generation mechanism of OECB in office rooms is well explained by Priority-based task queue model (PBTQM). The modeling results in this paper can be applied for occupancy classification with similar energy use habits, building performance simulation, energy consumption system design and energy conservation management.