/epuck-swarm-webots2023b

This repository maintains the simulation worlds used for the E-Puck Swarm Platform.

Primary LanguagePythonApache License 2.0Apache-2.0

swarm-epucks

This repository contains typical environments and demo codes for epucks swarm simulation.

Requirements

Ubuntu 20.04, Webots 2023b, python 3.8

Requirements of py-swirld installation refers to https://github.com/Lapin0t/py-swirld

pip install py-swirld

Install

git clone git@github.com:SICC-Group/swarm-hashgraph-webots2023b.git

Usage

object-searching:

  1. Launch the Hashgrpah process

cd object_searching/controllers/py-swirld-object-searching/
python swirld_object_searching2.py    
  1. Launch the Webots simulator by click the webots icon
  2. In Webots simulator, open the world object_searching2.wbt in object_searching/worlds/
  3. Click the run button

black-white-ratio-estimate:

  1. Launch the Hashgrpah process

cd black-white-ratio-estimate/controllers/py-swirld-black-ratio/
python swirld_black_white_ratio_estimate.py    
  1. Launch the Webots simulator by click the webots icon
  2. In Webots simulator, open the world epuck_gezi_twenty_48.wbt in black-white-ratio-estimate/worlds/
  3. Click the run button

Tips

Don't forget to change the absolute path of the document in the code to read your preset parameters and save your experimental results:

Replace all paths in the form of '/home/<>/<>.txt' with the paths on your own computer.

All txt files with names that appear in the code can be created as empty files or have 0 filled in on the first line.

Acknowledgements

Part of this repository is original from the hashgraph-based framework. The corresponding paper is here. The source code is here. Demo Video is here.