[Paper Review] A Fluid Dynamic Model for the Movement of Pedestrians

Dirk Helbing|arXiv (Cornell University)|May 18, 1998

Evacuation and Crowd Dynamics22 references432 citations

TL;DR

This paper develops a fluid dynamic model for pedestrian movement based on a Boltzmann-like gaskinetic approach, treating pedestrians as a multi-component system with distinct intended velocities. It derives coupled equations for density, mean velocity, and velocity variance, revealing that pedestrian jams and lane formation arise from avoidance maneuvers and directional asymmetry, with viscosity increasing with density and wave propagation governed by reaction time.

ABSTRACT

A kind of fluid dynamic description for the collective movement of pedestrians is developed on the basis of a Boltzmann-like gaskinetic model. The differences between these pedestrian specific equations and those for ordinary fluids are worked out, for example concerning the mechanism of relaxation to equilibrium, the role of ``pressure'', the special influence of internal friction and the origin of ``temperature''. Some interesting results are derived that can be compared to real situations, for example the development of walking lanes and of pedestrian jams, the propagation of waves, and the behavior on a dance floor. Possible applications of the model to town- and traffic-planning are outlined.

Motivation & Objective

• To develop a fluid dynamic model for pedestrian movement that captures collective behavior without assuming momentum and energy conservation like in ordinary fluids.
• To explain emergent phenomena such as pedestrian jams, walking lanes, and wave propagation in crowds using a physics-based approach.
• To account for pedestrian-specific dynamics such as intended velocity, directional preferences, and avoidance behavior in a fluid-like framework.
• To provide a theoretical basis for applications in urban and traffic planning by modeling crowd behavior under varying densities and interaction rules.

Proposed method

• The model classifies pedestrians into types μ based on intended direction, using a gaskinetic framework to describe the distribution function ρ̂μ(𝐱,𝐯μ,𝐯μ⁰,t).
• Continuity equations are derived for the distribution function, incorporating relaxation to intended velocity, directional changes, and interaction effects.
• From these, macroscopic fluid-like equations are derived for spatial density ⟨ρμ⟩, mean velocity ⟨vμ⟩, and velocity variance ⟨(δvμ,i)²⟩.
• The model introduces a non-equilibrium relaxation mechanism driven by intended velocity, leading to density-dependent viscosity and internal friction.
• Wave propagation is analyzed using the mean reaction time ζμ, showing that density waves travel at a velocity cμ dependent on ζμ.
• The model distinguishes 'temperature' as velocity variance due to intended velocity spread, and 'pressure' as a driving force from intended motion, not from thermal motion.

Experimental results

Research questions

• RQ1 How can pedestrian movement be modeled as a fluid-like system without assuming conservation of momentum and energy?
• RQ2 What causes the spontaneous formation of walking lanes and pedestrian jams in dense crowds?
• RQ3 How do avoidance maneuvers and directional asymmetry influence crowd flow and internal friction?
• RQ4 What determines the propagation speed of density waves in pedestrian flows?
• RQ5 How do 'temperature' and 'pressure' in pedestrian systems differ from those in ordinary fluids?

Key findings

• Pedestrian jams emerge from avoidance maneuvers, especially when velocity variance is high, and are exacerbated by high pedestrian density.
• Walking lanes form due to asymmetrical probabilities of avoiding others to the left or right, reducing conflict frequency.
• Viscosity ημ increases with pedestrian density, unlike in ordinary fluids, due to the relaxation mechanism toward intended velocity.
• Wave-like density propagation occurs with a phase velocity cμ that depends on the mean reaction time ζμ, not on diffusion or pressure gradients.
• The 'temperature' θμ of the system is defined by the variance of intended velocities, not by thermal motion, and can differ between groups on a dance floor.
• Stationary velocity profiles are hyperbolic rather than parabolic, as the intended velocity drives flow, not pressure gradients.