Stochastic model of self-driven two-species objects in the context of the pedestrian dynamics

TL;DR

This paper introduces a stochastic model for two-species pedestrian dynamics, capturing fluctuations and stop-and-go waves through biased random walks and simulations, revealing non-regular diffusion and crowd effects.

Contribution

It develops a novel combined PDE and Monte Carlo framework to model self-driven particles against an opposing crowd, incorporating interaction delays based on local concentrations.

Findings

Particles exhibit non-regular, long-tailed diffusion distributions.

Opposing crowd particles increase dispersion of target particles.

Model effectively captures stop-and-go wave phenomena in pedestrian flow.

Abstract

In this work we propose a model to describe the statistical fluctuations of the self-driven objects (species A) walking against an opposite crowd (species B) in order to simulate the regime characterized by stop-and-go waves in the context of pedestrian dynamics. By using the concept of single-biased random walks (SBRW), this setup is modeled both via partial differential equations and by Monte-Carlo simulations. The problem is non-interacting until the opposite particles visit the same cell of the considered particle. In this situation, delays on the residence time of the particles per cell depends on the concentration of particles of opposite species. We analyzed the fluctuations on the position of particles and our results show a non-regular diffusion characterized by long-tailed and asymmetric distributions which is better fitted by some chromatograph distributions found in the literature. We also show that effects of the reverse crowd particles is able to enlarge the dispersion of target particles in relation to the non-biased case ($\alpha =0$) after observing a small decrease of this dispersion