A unified multiscale vision of behavioral crowds

Bouchra Aylaj, Nicola Bellomo*, Livio Gibelli, Alessandro Reali

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


This paper proposes a multiscale vision to human crowds which provides a consistent description at the three possible modeling scales, namely, microscopic, mesoscopic, and macroscopic. The proposed approach moves from interactions at the microscopic scale and shows how the same modeling principles lead to kinetic and hydrodynamic models. Hence, a unified framework is developed which permits to derive models at each scale using the same principles and similar parameters. This approach can be used to simulate crowd dynamics in complex environments composed of interconnected areas, where the most appropriate scale of description can be selected for each area. This offers a pathway to the development of a multiscale computational model which has the capability to optimize the granularity of the description depending on the pedestrian local flow conditions. An important feature of the modeling at each scale is that the complex interaction between emotional states of walkers and their motion is taken into account.

Original languageEnglish
Pages (from-to)1-22
JournalMathematical Models and Methods in Applied Sciences
Issue number1
Early online date12 Dec 2019
Publication statusPublished - Jan 2020


  • Crowd dynamics
  • living systems
  • multiscale vision
  • social dynamics


Dive into the research topics of 'A unified multiscale vision of behavioral crowds'. Together they form a unique fingerprint.

Cite this