A Review of Building Evacuation Models - NIST Virtual Library

A.6 PedGo
Developer: TraffGo
Purpose of the model: To model crowd movement, to simulate the evacuation of pedestrians
from buildings, ships, aircraft, and other kinds of public transportation systems 22-24 .
Availability to the public for use: Yes, there are software licenses available for the PedGo
model via the company, TraffGo.
Modeling method: Movement/Partial Behavioral Model
Structure of model: This is a fine network model that divides the floor plan into – 0.4 x 0.4 m
grid cells – represent the space taken up by a person. The walls, furniture and any other
obstacles are represented by cells which are occupied at all times throughout the simulation.
Perspective of model: This model is labeled as a microscopic model, meaning that each person
is represented individually. Therefore the perspective of the model is individualistic. And, since
the user can specify egress routes for the occupants, the occupants’ perspective of the building is
also individualistic.
Occupant behavior: Implicit behavior. The model begins to offer implicit behavioral inputs,
such as the individual inputs of pre-evacuation delays, patience, reaction, dawdle, and sway.
This set of parameters is used for characterization of behavior and are assigned to individuals in
the simulation according to a normal distribution. Two of these parameters, delay time and sway
are stochastic. These parameters, as shown in Figure A.8 (reference 24 below), are the
following:
• Maximum walking speed (cells)
• Patience (s) – the time a person is willing to wait until choosing another escape route
• Look (cells) – a factor describing visual perception of the environment
• Reaction (s) – a factor describing the inertia of a person’s movement
• Dawdle (%) – stopping for one timestep (stochastic)
• Sway (%) – deviation from a straightforward path to the exit (stochastic).
Occupant movement: Cellular automata model.
One set of rules is applied to all occupants in the model regarding their movement. Individual
differences affect the person’s behavior. The six parameters characterizing an occupant’s ability
are the following: maximum velocity (cells), patience (s), reaction (s), dawdle (%), sway (%),
and inertia. Each parameter is given a minimum, maximum, and mean value and standard
deviation to distribute over occupants in the simulation.
A-16