Co-simulation of a building energy system integrated with an agent-based model of individuals' behaviour and thermoigrometric comfort

In the context of residential environments, thermal comfort has emerged as a fundamental consideration in architectural design, impacting both the well-being of occupants and the energy efficiency of buildings. The growing emphasis on sustainability and consumption reduction has driven research to develop increasingly accurate models to describe human thermal perception and predict user behaviour. International standards, such as ASHRAE 55 and UNI EN ISO 7730, provide established guidelines for assessing ideal environmental conditions. However, these standards often ignore the variety of people and their daily habits. Consequently, it is imperative to identify more flexible and personalised methods to understand how individuals perceive and react to heat and cold. The purpose of this work is to explore precisely this aspect: determining the thermal comfort of a user in a domestic context by comparing different models for calculating thermal sensitivity. The following models were considered: Fanger's PMV, Gagge's two-node model and a modified version of the PMV that takes body mass index (BMI) into account. In order to achieve this objective, an agent-based model was developed in order to simulate the interaction between three principal actors. The first component is the Occupant agent, which represents the user and their daily routines. The second component is the BDI agent, an agent based on the Belief-Desire-Intention architecture. The BDI agent is capable of predicting user behaviour and choosing the most appropriate actions using Bayesian probabilities. It is also able to evaluate the effectiveness of past choices and the level of environmental awareness. The third component is the HVAC agent, which manages air conditioning devices (air conditioner, radiators, dehumidifier/humidifier and fan) and estimates their energy consumption. The proposed approach facilitates the dynamic evaluation of the impact of diverse thermal sensitivity models and behavioural strategies on energy efficiency and perceived comfort, offering a useful tool for the design of more sustainable and user-centred air conditioning systems.