An improved social force model based on nucleus force theory to simulate building evacuation in different visibility conditions

Abstract

Accurately simulating pedestrian behaviour under different visibility conditions is crucial for reducing casualties during emergency fire evacuations. Current research employing social force models typically simulates pedestrian behaviour based on interaction forces under specific visibility conditions. However, these studies often inadequately capture the dynamic effects of visibility changes on pedestrian interaction forces. Based on the nucleus force theory, this study developed an improved social force model (NSFM), incorporating environmental visibility parameters, establishing corresponding pedestrian movement rules. Additionally, we investigated the interaction between pedestrians and walls to determine the optimal parameters. The model’s accuracy was then validated by comparing its simulations under specific visibility conditions from previous visibility-based evacuation experiments and results from other models. Furthermore, we conducted simulations under different visibility conditions, the results show that reduced visibility intensifies wall-following behaviour and herd effects, leading to more detour behaviour, slower movement velocity, and longer evacuation times. As visibility increases, the impact on evacuation gradually diminishes. Finally, we investigated the impact of the number and location of exits and discovered that increasing the number substantially reduces evacuation time, while changes in exit locations can notably affect evacuation efficiency. The numerical simulation results demonstrate that the NSFM has significant potential for simulating pedestrian evacuation behaviour and processes under different visibility conditions, providing a scientific basis for designing more effective evacuation strategies in the future.