Modelling pedestrian evacuation in airport terminals accounting for movable obstacles

With increasing passenger volumes at airports, especially when travelers carry luggage, trolleys, and other movable obstacles, the efficiency and safety of evacuation are significantly affected. This study develops a microscopic evacuation model that explicitly incorporates pedestrian-movable obstacle interaction mechanisms to improve the simulation of evacuation behaviors in complex, high-density environments. We integrate exit attractiveness, local density, and the time-to-collision (TTC) risk index, and introduce three typical behaviors (pushing, detouring, and crossing) using a multinomial logit (MNL) framework for behavioral decision-making. Through simulation analyses in selected areas of Terminal 2 at Baiyun Airport, we systematically examine the impacts of movable obstacles and behavioral strategies on evacuation efficiency and safety. The results indicate that both the spatial size distribution and the behavioral strategies related to movable obstacles have substantial effects on the evacuation process. Specifically, movable obstacles located near check-in islands or exits can markedly prolong evacuation time and increase collision risk. Moreover, pushing large obstacles can easily trigger severe congestion, whereas pushing small obstacles can partially improve pedestrian flow. The direction of push is also a key factor. A mixed strategy based on TTC and movable obstacle size performs best in complex, high-density scenarios. However, when the proportion of luggage-carrying passengers exceeds 30%, overall evacuation efficiency drops sharply. These findings inform airport emergency management and obstacle-control policies.