Simulating human thermal comfort and resulting usage patterns of urban open spaces with a Multi- Agent System

Attractive public open spaces can serve as key design elements for enhancing life quality in urban areas and to diminish the negative effects of urbanisation. In order to fulfil this task, open spaces must meet the expectation of potential users and offer the right mix of environmental, economic and social conditions. It is well known that the local microclimate belongs to the main factors driving the individual perception and assessment of an outdoor environment. In order to simulate the thermal comfort of pedestrians in urban open spaces, the method of Multi-Agent Systems (MAS) is used. In which each agent represents a virtual pedestrian with individual physiological properties and different routes through the model environment. While walking through the domain, these agents are virtually exposed to different climate conditions. The impact of these climate environments on the individual thermal comfort are simulated using a dynamic model of the human thermoregulatory system. Depending on their individual thermal comfort level, the agents can adjust their routing decisions and also consider using optional facilities such as benches or restaurant chairs. Using the simulation results from a few thousand virtual pedestrians, it is then possible to generate different kinds of comfort or frequentation maps not only representing the local microclimate conditions, but also considering the space-time links that establish through pedestrian motion and individual decisions.