Design and implementation of an agent-based social simulation model of energy related occupant behaviour in buildings

Energy consumption in buildings accounts for 20% of the total energy consumed worldwide in 2016. Furthermore, this value grows around 2% per year and is higher in place such as the EU, 41%, and US, 40%. Thus, there is a growing interest in technologies that enable intelligent monitoring and control of buildings' equipment to improve energy efficiency. In addition, the number of indoor workers, whose productivity is greatly affected by their comfort, is increasing. This has brought the notion of Smart Buildings, that combine the energy saving with the occupants' comfort goals. Energy consumption of buildings is proved to be largely influenced by the presence and behaviors of occupants. Most systems of control and operation of buildings in energy systems consider general models based on general occupants' schedules, behaviors and preferences that leads to large predictive and optimization errors. A novel energy efficiency model based on multi-agent occupancy simulation and methodologies application is proposed in this work. The behavior of each occupant is modeled by states according to schedules and Markov decision process. The electrical equipment, lighting system and air conditioning systems are modeled as passive and reactive agents. The simulated environment's operations are modeled by three distinct policies or control strategies involving different levels of intelligent operation. The obtained comfort and energy results are evaluated and valued. As result, a 13.28% reduction in energy consumption and a 11.03% improvement in occupant comfort are obtained as comparison between the traditional and most advanced policy