An Evacuation Simulation Based on Boltzmann Distribution - Advanced Research on Computer Education, Simulation and Modeling - page 327

An Evacuation Simulation Based on Boltzmann
Distribution
Hao Lian, Ang Li, Yang Tian, and Ying Chen
No 10, Xi Tu Cheng Road, Haidian District
Beijing University of Posts and Telecommunications
100876 Beijing China
Abstract. The exited popular ways of evacuation simulation can't satisfy us in
the computational efficiency and the calculation accuracy at the same time. In
order to get acceptable efficiency and accuracy, we applied the related concept
of statistical physics to the specific circumstances of evacuation and built an
evacuation model based on Boltzmann Distribution. We also came up with the
exit attraction function and direction judging principles to depict people's exit
choosing and obstacles avoiding behaviors. At last, we proved our model is ap-
plicable and reasonable by comparing the classical models with ours.
Keywords: Evacuation, Boltzmann distribution, Exit Attraction Function.
1 Introduction
When the accidents (such as fires) happen in the large buildings nowadays, the loca-
tion design of the emergency exits and peoples' escaping routes both have crucial
influence on the final evacuation time. If we analyze the evacuation data from the
simulation to find the choke point of the evacuation process, then we can evaluate the
buildings' design quality and put forward the improvement suggestions.
The classical evacuation models either go against engineering applications because
of large amount of calculation (such as Social Force Model [1]-[6]), or can't provide
reliable evacuation results because of lacking authenticity (such as Cellular Automata
model [7]-[13]). Thus, how to balance the accuracy and the time of the simulation has
already been the main trend of the evacuation study.
We try to use a totally new method (the Boltzmann Distribution way) to handle the
evacuating problems and our goal is to make the efficiency and the accuracy both
reach an acceptable range.
2 The Evacuation Model Based on Boltzmann Distribution
2.1 Attraction Indicators
In many classical models, people are seen as particles in a specific space and the in-
fluence of emergency exits and the people around are treated as a kind of attractions