Collective-Behaviour-22-23-Group-C

Repository for group work in Collective Behaviour 22/23

We aim to quantitively evaluate and compare different psychological crowd modelling algorithms through the systematic exploration of state of the art Crowd Simulation models. Our first fuzzy logic-based model, replicating OCEAN personality crowds, published in 2020 by https://www.mdpi.com/2073-8994/9/10/239. The Model we compare it with is based upon the BioCrowds algorithm and adds a Personality layer https://repositorio.pucrs.br/dspace/bitstream/10923/14148/2/Simulating_Crowds_with_OCEAN_Personality_Traits.pdf. Each model will be analysed on both a conceptual level and a practical level. All models will perform the same tasks and be evaluated on similar metrics, thus creating both a systematic review and an quantitative result comparison. All development and experimentation are conducted using Vadere. Vadere is an open- source framework for the simulation of pedestrian and crowd dynamics. Its GUI pro- vides a toolset to create topographies, alter hyperparameters and output measure- ments of its simulations. Quantitatively, Vadere can measure the overlaps of the pedes- trians, their wait times and other values. The simulation setups and validation metrics are described below in Simulation and Validation. The reviewed models are implemented inside Vadere. Its core is a simula- tion loop that calculates the movement of several pedestrians depending on the given targets, topography, and a locomotion model. Vadere integrates several locomotion models, which typically consist of a module that extends upon Vadere’s default pedes- trian object definition and a module that moves the pedestrians called the "steering model".

The following is the documentation of Vadere from https://gitlab.lrz.de/vadere/vadere:

Vadere

Vadere is an open source framework for the simulation of microscopic pedestrian and crowd dynamics. Vadere provides generic model classes and visualisation and data analysis tools for two-dimensional systems. A series of pedestrian locomotion models are already implemented in the framework that are also the basis for scientific publications. Namely the optimal steps model, the gradient navigation model, and the social force model are available for simulation studies. In addition to pedestrians, other systems including cars and granular flow can be introduced into the framework.

This software runs on Windows, OS X and Linux.

Vadere has been developed by Prof. Dr. Gerta Köster's research group at the Munich University of Applied Sciences at the department for Computer Science and Mathematics. However, we welcome contributions from external sources.

The Vadere framework includes a mesh generator for unstructured high-quality 2D meshes called EikMesh which is described here.

If you are using Vadere for your publication, please cite: http://dx.doi.org/10.17815/CD.2019.21.

Vadere Units

All measurements in Vadere are in SI units. That means, meters for the positions in the GUI and Topography, and meter/second for speeds.

Contact

If you have questions about Vadere, have found a software bug or have a suggestion for improvements, please feel free to either

• open an issue (Note: you require an LRZ account), or
• send an e-mail with your enquiry to our mailing list vadere@lists.lrz.de

Pipeline Status

Branch	Pipeline Status	Coverage
master
stable

Installation

Dependencies

• Java 11 (OpenJDK recommended -> see here for more information) (not tested for > 11)
• OpenCL (optional but recommended -> see the install instructions for details)

Pre-Built Releases

Latest build of master:

• Windows
• Linux

Stable releases and selected branch-builds:

• https://www.vadere.org/download/

The ZIP file contains:

• README.md - this README file.
• vadere-gui.jar - provides the GUI version of Vadere.
• vadere-console.jar - provides the command line version of Vadere and allows easy integration into other applications.
• Scenarios - contains test scenarios for pedestrian locomotion models. Note: The tests are also useful for a "getting started" (see below "Run Built-In Examples" for details).

Getting Started

Either run the Vadere simulator by

• Double-clicking vadere-gui.jar (after installing Java)
• Entering path/to/openjdk/java -jar vadere-gui.jar

Scenario Files

You can find some example scenarios under Scenarios/Demos

For more info see Scenario File Specification

Run Built-In Examples

With the following steps, you can run a simulation with one of the built-in examples from Scenarios:

• Start Vadere
• Click Project > Open
• Choose vadere.project of one of the test projects, e.g. TestOSM and click open
• Select the scenario on the left and press run selected scenario

Build from Source

Dependencies

• Java 11 or above (OpenJDK recommended)
• Maven 3.0
• Git
• OpenCL (optional but recommended)

Note: Please, ensure that the Git executable can be found in the PATH variable of your operating system.

Build Instructions

1. git clone https://gitlab.lrz.de/vadere/vadere.git
2. cd vadere
3. mvn clean
4. mvn -Dmaven.test.skip=true package

Changelog

See CHANGELOG.md for a list of changes.

JavaDoc

• state

Contribution

See CONTRIBUTING.md for how to set up the development environment and the coding guidelines.

License

This software is licensed under the GNU Lesser General Public License (LGPL).

For more information: http://www.gnu.org/licenses/lgpl.html