Development of floor field cellular automaton model for pedestrian dynamics: Incorporating empirical acceleration mechanisms

Abstract

Pedestrian movement during emergency evacuations involves frequent and rapid speed changes. However, most existing simulation models – including the widely used Floor Field Cellular Automaton (FFCA) – do not realistically account for acceleration and deceleration. These models often assume an instantaneous transition from rest to maximum speed within a single timestep. This simplification reduces their accuracy in high-speed or high-density situations. To address this limitation, we propose a fine-discrete FFCA model that explicitly integrates empirically derived acceleration mechanisms. Controlled experiments were conducted to identify triggers for acceleration and deceleration, collecting data across a broad range of pedestrian speeds. These behaviors were integrated into the FFCA framework through dynamic rules governing movement initiation, adjustment, and interaction. The model was validated by comparison with the classic FFCA model and empirical data from the 2008 Wenchuan Earthquake evacuation, as well as conducted bottleneck evacuation experiments. In validation using earthquake evacuation data, the developed model more accurately replicates pedestrian dynamics, producing smooth acceleration/deceleration profiles and flow rates consistent with empirical observations. Notably, it reduced the root mean standard error of cumulative passing interval distribution by 77.6%. In the controlled experiment validation, the model predictions closely matched experimental results in evacuation timing, pedestrian trajectories, and spatial speed distributions. These improvements significantly enhance the FFCA model’s applicability in emergency evacuation simulations and supporting more effective safety assessments.