[Zurück]

M. Bruckner, S. Tauböck, N. Popper, S. Emrich, S. Alkilani, B. Rozsenich:
"A Combined Cellular Automata and DEVS";
Vortrag: MATHMOD 2012 - 7th Vienna Conference on Mathematical Modelling, Wien; 14.02.2012 - 17.02.2012; in: "Preprints Mathmod 2012 Vienna - Full Paper Volume", F. Breitenecker, I. Troch (Hrg.); Argesim / Asim, 38 (2012), S. 350 - 351.

Summary: In autumn of 2006, at the Vienna University of Technology the project TU Univercity 2015 was
launched, which includes the modernization of the buildings and the concentration of the faculties on maximum
two locations. In the context of this project, the Institute
for Analysis and Scientific Computing developed a
model to optimize the classroom assignment and
simulate the processes in place at the university. This
model called MoreSpace was realized mostly in the
simulation language Enterprise Dynamics (ED), with
exception of the part that calculates the time that the
students need to change between lecture rooms. This
outsourced part of the simulation is the main topic of
this paper. The model is implemented in the objectoriented
programming language Java and connected via
TCP/IP with Enterprise Dynamics. To model the
dynamic behavior of single individuals an agent-based
system was chosen in which the individuals move on a
discrete grid. The cell size is 0.125 x 0.125m, so that one
m2 consists of 64 cells. Each student takes 4 x 4 cells, or
0.5m x 0.5m. The forward movement of people in a
building depends on several interrelated factors. Some of these e.g. are the density of people in a group, the
maximum speed which varies for each individual. Of course these factors are conditioned by the environment; For
example moving into a room area or a staircase makes a noticeable difference. It is also of crucial importance,
whether a person is facing multiple other individuals. All this is relevant for the speed and thus for the required
time which the students need to switch from location A to another location B.
The Model: Basically, this simulation consists of two simulators: On the one hand Enterprise Dynamics, a
commercial Simulation Software from INCONTROL based on the DEVS (Discrete Event System Specification)
formalism in which the optimization of the room utilization is modeled. On the other hand a proprietary
development in JAVA which provides the time needed by students to changed between lecture rooms. This part is
designed to ensure that the time between two teaching units that may be held in different buildings, is sufficiently
dimensioned. In order to implement the task a cellular automats (CAs) model was used [1, 2]. The main
components of this model is a discrete plane and the individual agents, or as in our case, the individual students
who are moving on this grid, and whose decision on their further action depends on the behaviour of the
surrounding agents. Because of the fact that the university area is too big to display on only one of such grids, the
buildings are divided in logically coherent parts that are connected at several points to give the agent the
possibilities to change between these discrete planes. In order to allow the students to move through the huge
number of planes in the shortest possible way a combination of graph theory and a kind of scalar field is used
The Elevator: An Important Part of this simulation was the implementation of the Elevators. This component
creates a lot of problems for example the elevator controller for grouped elevators and the decision making for the
students to find her best (with consideration of way length and walking time) personal way.
Persons with physical handicap: This Program also regards the needs of people with physical handicap. This
means that it is possible to adjust the probability that a random generated person is physical handicapped and is
rely on a wheel chair. Some consequences of this are the inability of using stairs, an increased required space, and
a special speed distribution and so on.
This project is supported by "ZIT - Die Technologieagentur der Stadt Wien GmbH".
[1] Victor J. Blue, Jeffrey L. Adler. Cellular automata microsimulation for modeling bi-directional pedestrian
walkways. Transporation Research Part B 2001
[2] Fang Weifeng, Yang Lizhong, Fan Weicheng. Simulation of bi-direction pedestrian movement using a
cellular automata model. Physica A 2003