Modeling boundedly rational route choice in crowd evacuation processes

Abstract

With the development of modern information technologies, advanced evacuation guidance systems have been widely installed to provide updated route information. However, although pedestrians may know the information of the building and its surrounding environment, they sometimes still use the non-optimal route, that is to say, their decision-making process is rather boundedly rational. To understand the dynamics underlying such behavior, we present a model where we consider two important factors affecting pedestrians’ route estimation, i.e., distance and congestion. The model is then validated, after which, detailed simulation studies have been performed to discuss the influence of those two factors. Results show that for small and open spaces, the influence of the free walking time on pedestrians’ route choice is higher than the congestion queuing time. Moreover, pedestrians prefer changing route in order to save the completion time. For a large virtual building, if pedestrians would like to use short route, pay less attention to congestion route and prefer keeping current route, the route usage can be quite different, some of the routes can be extremely crowded while others can be barely used. On the contrary, if pedestrians do not mind distance difference, prefer avoiding congestion and changing route, significant detour behaviors will be observed, which also increases the evacuation time. This boundedly rational route choice feature results in distinct importance rank of evacuation facilities when considering their failures. The present study deepens our understanding of pedestrian route choice considering bounded rationality and provides a basis for crowd management in building evacuation.