# MATLAB FS12 – Research Plan

> * Group Name: anti-panic group
> * Group participants names: Flöry Nikolaus, Omeradzic Amir, Zengerle Thomas
> * Project Title: crowd flow optimization

## General Introduction

Emergency evacuation is a very important topic in many fields of our daily lifes. We find examples from small scale evacuations of a building, up to to a panic scenario in a stadium, in demonstrations or parades.
History has shown how dangerous such situations can get with many people getting injured or even killed by people pushing each other.


## The Model

We are using a cellular automata model, where each person is represented by a 3x4 matrix containing the coordinates and the information about the interactive environment.
The simulation is performed on a large matrix, which simulates our environment (room, corridor with exit, different width of exits, a variety of forms of obstacles etc).
The basic principle is to use a potential-like field, which gives every matrix entry a specific value, decreasing with the distance to the exit.
Therefore a person`s matrix contains the values (potential, obstacles) around him. Combining these values with some variables determines the next step.  

What we try is to focus on some specific variables:
-the desired velocity, people want to reach. This is of course strongly related to the level of panic.
-the realized velocity, which we can measure.

We will test our model in situations, where there is danger (for example a fire) on one side of a room, und people want to excape in the opposite direction.
Examples for such situations are cinemas, theaters or shopping centers.

## Fundamental Questions

Can we optimize the flux and efficiency of a evacuation?
Is there a "perfect" geometry of a room, that optimizes an evacuation?
	A cone-shaped room should get better results than a normal rectangle room.
Can a evacuation route positively be manipulated by placing obstacles in the room?	
What is the influence of the desired velocity on the efficiency of a evacuation? Faster is Slow?
	How do desired and realized velocity depend on each other? 
	What is the resulting average speed?


## Expected Results

We expect to enhance a evacuation by placing obstacles in the route. This could reduce the velocity and improve the efficiency of the whole evacuation.
Also we thnk we can find a correlation between desired and realized velocity.


## References 


Simulating dynamical features of escape panic	Dirk Helbing, Illes Farkas, Tamas Vicsek
Modelling crowd turbulence by many-particle simulations		Wenjian Yu, Anders Johansson
CA Approach to Collective Phenomena in Pedestrian Dynamics	Andreas Schadschneider, Ansgar Kirchner, Katsuhiro Nishinari


## Research Methods

self driven many particle systems simulation, cellular automata


## Other