Skip to content

This repository contains the experiments developed in the T6.3.2 of REXASI-PRO in Navground simulator.

Notifications You must be signed in to change notification settings

Cimagroup/ExperimentsT6.3.2_L1

Repository files navigation

Experiments T6.3.2 - D6.3. Topology-based optimization of robot fleet behavior. Detection of stable topological patterns using persistent entropy

This repository contains data and experiments associated to "D6.3. Topology-based optimization of robot fleet behavior" performed for the European Project REXASI-PRO (REliable & eXplainable Swarm Intelligence for People with Reduced mObility) (HORIZON-CL4-HUMAN-01 programme under grant agreement nº101070028), in concrete to subtask 6.3.2 and line 1. It has been created by the CIMAgroup research team at the University of Seville, Spain.

Repository structure

  • ExploratoryAnalysisWithouTopology Folder: It contains an exploratory analysis for behavior comparation.
  • RealScenarios Folder: It contains a realistic simulation environment for different scenarios: Cross and corridor.
  • T6-3-2-Experiments Folder: It contains the experiments that have developed for the specific purpose. It contains a folder for Cross Scenario and another one for Corridor Scenario. Each of them contains different notebooks, for different behaviors, comparing them, and for predicting collisions. In this folder, it is also a folder called IlustrationDeliverable, where images for the deliverable have been created.
  • TrajectoryAnalysis Folder: It contains experiments analyzing robot trajectories and their relation with collisions.
  • Two_or_more_type_agents: It contains a simulation example with two type of agents.
  • function.py: Contains some functions that are useful and will be used in the rest of files.

Usage

  1. Clone this repository and create a virtual enviromment:
python3 -m venv entorno python=3.10.12
  1. Activate the virtual enviromment:
source entorno/bin/activate
  1. Install Navground (we have used version 0.3.3, so using other version may vary or produces error):
pip install navground[all]
  1. Install the necessary dependencies:
pip install jupyter notebook matplotlib scipy multiprocess gudhi plotly scikit-learn pandas ripser seaborn tqdm

Ok, you can now run experiments! :)

Note: In case of issues doing that in WSL, reinstall using the following distribution and reinstall Navground with the steps mentioned earlier:

  1. Remove the entire Ubuntu distribution: wsl --unregister distribution

  2. Install the WSL distribution: wsl --install -d Ubuntu-22.04

  3. sudo apt-get update

  4. wsl --set-default-version 2

About

This repository contains the experiments developed in the T6.3.2 of REXASI-PRO in Navground simulator.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published