Exit dynamics of occupant evacuation in an emergency

https://doi.org/10.1016/j.physa.2005.08.012Get rights and content

Abstract

A two-dimensional Cellular Automata model is proposed to simulate the exit dynamics of occupant evacuation. Concerning the exit width and the door separation, we put forward some useful standpoints: (1) exit width should be bigger than a critical value, and the door separation should be neither too small nor too big; (2) for single-exit door, with the increase of exit width, the flux per unit width will decrease but the total flux will always increase; (3) the total flux of the exit is an increasing nonlinear function of the exit width; (4) the optimal value of the door separation does not vary with the value of exit width; (5) the layout of exits should be symmetry. Those results are helpful in performance-based design of building.

Introduction

Along with the unceasing research on pedestrian flow and occupant evacuation, more and more researchers apply themselves to the research on individual behavior [1], [2] and a variety of phenomena resulting from collectivity behavior [3], [4] such as: jamming in traffic systems and granular materials [5], [6], [7], pedestrian behavior [8], [9], flocking in birds [10] and so on.

In order to find out the behavior of occupant evacuation at the exits of building, Ref. [11] used a simple rule of movement, namely, the occupant faces the vacant cell along the exit path and moves if L+R<B+ϕ where ϕ is a measure of his level of anxiety or panic (eagerness to move), otherwise he stays. L, R, and B are the total number of neighbors to the left, right and back directions of the cell, respectively. Other pedestrians who are separated from the cell by a vacant cell are not counted as neighbors.

As we know that evacuation process is quite complex and changeful, it is not suitable to simulate occupant evacuation with a deterministic rule. In order to approach the fact of evacuation, a particular two-dimensional Cellular Automata (CA) model based on the individual behavior is proposed.

Section snippets

Model description

The structure of the building is represented by a two-dimensional grid. Each cell can either be empty, occupied by an obstacle or occupied by one occupant. The size of a cell corresponds to 0.4×0.4 m2, which is the typical space occupied by an occupant in a dense crowd. In the model, each time step (ts) represents different real time of evacuation based on different walking velocity of occupant. For example, empirically the average velocity of an occupant in common condition is 1.0 m/s. Thus one

Simulation

In this paper, actual size of the side of each cell (0.4 m) is regarded as one unit, namely, the dimensionless size of the occupant is 1 and the dimensionless size of the room is the actual size divided by 0.4 m.

Generally, the range of occupants’ walking speed is between 0.5 and 1.5 m/s. It decreases with the increase of the density of occupants. Because of the dense distribution, the experiential value 0.8 m/s is regarded as the average walking speed. Therefore, one ts represents 0.5 s in this

Summary

In this paper, a two-dimensional CA random model is proposed to investigate the exit dynamics of occupant evacuation, which is regarded as a kind of collective motion caused by individual behavior. We have validated the phenomenon and conclusion proposed before: (1) there are optimal values of the exit width and the door separation; (2) arch, which indicates a jammed state will be formed at each exit door during evacuation. And the shape will change with the value of door separation. In

Acknowledgements

This paper was supported by the China NKBRSF project (no. 2001CB409603), National Natural Science Foundation of China (Grant no. 50276058) and the Important International Cooperate Project of NNSFC (no. 2003-50320120156). The authors deeply appreciate the supports.

References (20)

  • F. Ozel

    Saf. Sci.

    (2001)
  • P. Gipps

    Math. Comp. Simul.

    (1985)
  • C. Burstedde et al.

    Physica A

    (2001)
  • A. Kirchner et al.

    Physica A

    (2002)
  • A. Sekizawa et al.

    Fire Mater.

    (1999)
  • T. Vicsek

    Nature

    (2001)
  • H. Levine et al.

    Phys. Rev. E

    (2001)
  • D. Helbing et al.

    Phys. Rev. E

    (1999)
  • K. To et al.

    Phys. Rev. Lett.

    (2001)
  • H.M. Jaeger et al.

    Rev. Mod. Phys.

    (1996)
There are more references available in the full text version of this article.

Cited by (111)

  • Robust scheduling for large scale evacuation planning

    2020, Socio-Economic Planning Sciences
View all citing articles on Scopus
View full text