Discretisation effects and the influence of walking speed in cellular automata models for pedestrian dynamics

TL;DR

This paper examines how discretisation and walking speed variations in cellular automata models affect pedestrian flow simulations, showing that higher speeds improve alignment with real-world data.

Contribution

It introduces methods to incorporate higher walking speeds into cellular automata models and analyzes their impact on pedestrian dynamics.

Findings

Increasing cell size introduces non-local conflicts.

Higher maximal walking speeds improve model accuracy.

Different model variants handle speed extensions effectively.

Abstract

We study discretisation effects in cellular automata models for pedestrian dynamics by reducing the cell size. Then a particle occupies more than one cell which leads to subtle effects in the dynamics, e.g. non-local conflict situations. Results from computer simulations of the floor field model are compared with empirical findings. Furthermore the influence of increasing the maximal walking speed $v_{{\rm max}}$ is investigated by increasing the interaction range beyond nearest neighbour interactions. The extension of the model to $v_{{\rm max}}>1$ turns out to be a severe challenge which can be solved in different ways. Four major variants are discussed that take into account different dynamical aspects. The variation of $v_{{\rm max}}$ has strong influence on the shape of the flow-density relation. We show that walking speeds $v_{{\rm max}}>1$ lead to results which are in very good agreement with empirical data.