TOMSK, Apr 11 – RIA Tomsk. Radiophysics of Tomsk State University (TSU) are developing a
system which builds a 3D model for evacuating people in emergency situations.
As it is reported on the website of the university, the program will allow to
check evacuation plans for various premises, with taking into account possible
panic, specificity of structures and materials.
Scientists plan to create a working prototype of the
system by 2020. Divisions of the Ministry of Emergency Situations will be able
to use it during check of places of mass use, it will significantly reduce the
number of victims in case of emergencies, in particular, during fires.
Employees and students of the Department of
Information technologies in the study of discrete structures of the Faculty of
Radiophysics of TSU work on the project. The project manager – is the employee
of the department, Candidate of Technical Sciences Anton Ermakov.
To find a problem
In the process of evacuation in emergency situations,
in particular, in case of fire, an important factor is the time during which
people are able to leave the dangerous territory, is explained in the message.
Both the structure of the building and the condition of the people themselves
at the moment of danger can affect the speed of evacuation.
The created system will help to reveal hidden
drawbacks of building structures, doorways, ventilation.
The idea of radiophysics is to upload building and space
plans into the system, and then, using emergency modeling, check their
evacuation plans or develop new ones. The model will take into account the
materials from which the walls, floor and ceiling are made – the speed of fire
propagation, smoke, smoke toxicity and other indicators depend on it. The
system also simulates psychophysical parameters of people in the conditions of
panic, predicts the number of fallen, wounded and dead.
“Now there are some methods for modeling of people’s
behavior in emergency situations, but in most cases they do not take into
account the third dimension, volume, that is, they analyze only two-dimensional
space, which significantly reduces the reliability”, – Ermakov is quoted in the
message.
He adds that these methods also "greatly simplify
the human model": "Eliminate physical parameters and interaction with
other people, reduce their movement to the movement of a multitude of particles
or two-dimensional figures. The process is similar to the flow of liquid, and fallen
or dead people-particles just disappear".
TSU radiophysicists create a system for modeling the evacuation of people during emergencies
To
save a bot
Now the authors of the project can already upload a 3D
model of the room into the system, distribute people (bots) and set an
evacuation point where the bots should run in case of fire. At the time of
launch, each of the bots is counting on the shortest evacuation route and is
trying to leave the premises, overcoming obstacles.
“We conducted experiments when the bots left the
premises in a calm state, that is, they did not participate in fights, did not
fall, did not panic. In this case, the evacuation time is minimal. But we want
to put in model psychophysical parameters and physical interaction of people
among themselves, because if, for example, one was damaged, then it changes
speed, detains others", – says Ermakov.
In the model being developed, the behavior of each bot
will be described by an expanded automaton using self-learning models
(artificial intelligence). This approach will allow to carry out effective and
realistic modeling with the possibility of 3D visualization.
© сайт Томского государственного университета
“We plan to involve psychologists in the process of
work, because they know typical behavioral reactions, for example, that a child
will not run to the evacuation exit, and will look for a mother, and so on”, – notes
the head of the Information technologies in the study of discrete structures
department of the Faculty of Radiophysics Stanislav Torgayev.
Radiophysicists are also working on creation of a base
of materials for interior decoration of premises where their fire prevention
characteristics will be taken into account – flame spreading rate, smoke level,
toxicity, and others. The presence of such a database will improve the adequacy
of the simulation in each particular room.