Development of a Dynamic Impatience Model for Building Evacuation Simulation

Project: Research

View graph of relations


In cases of compartment fire, the common objective of evacuees is to get out of the firecompartment. The exit door or corridor becomes a bottleneck area when the flow rate ofthe evacuees surpasses the throughput capacity of the door or corridor. This is a commonproblem in the course of evacuation. Microscopic evacuation models have been widelyadopted to identify the bottleneck areas of building layouts. Such models simulate themovements of all evacuees and the interactions between them. They also providebuilding designers with detailed information on evacuations, such as overall evacuationtime, flow pattern and bottleneck area locations, etc., which allows them to adjust theirdesigns to provide efficient evacuation routes. However, researchers have reported thatlocal effects (e.g. interactions between evacuees) have strong influence on macroscopicquantities such as flow pattern and total evacuation time. Therefore, it is very criticalthat a microscopic evacuation model be able to closely mimic the realistic interactionsbetween evacuees and produce a reasonable simulation.Interactions between evacuees are especially important in congested situations.Conflicted movement may occur during an evacuation when evacuees are competing forvacant space. Some models that simply apply the probabilistic approach ignorepsychological factors and randomly select one evacuee as the winner of the competition.Other models use desired velocity to simulate evacuees’ impatience without consideringthe dynamic growth of impatience over time. Such models only relate impatience tomoving velocity. Furthermore, impatience is kept constant and does not grow with timewhen an evacuee is blocked by others. These models do not truly reflect the dynamicnature of impatience.This project will develop a dynamic impatience model to determine the impatience levelof evacuees using their velocities and travelling times. If there is a movement conflictbetween a few evacuees, the one with the highest level of impatience will be picked tooccupy the vacant space in the next time step. The basic features of the proposedimpatience model have been investigated and developed to enable the dynamic change ofimpatience with time. A mathematical model that satisfies those features will bedeveloped in the project. That model will be implemented in a cellular automata modelfor evacuation simulation. Experimental data will be used to calibrate the model’sparameters. We will also benchmark the performance of the model with othercommercial models that are widely adopted in the fire engineering industry.


Project number9042482
Grant typeGRF
Effective start/end date1/01/1831/12/21