Multi agent decision making algorithms for a team game of pacman. Project made in the context of an artificial intelligence class at the UTBM. Project made by:
- Cloarec Florian
- Mann William
- Dewasmes Oscar
- Dreano Hermance
- Le Guilly Erwann
Cloner le projet
git clone https://github.com/kalharko/pAIcman
Go into the project
cd pAIcman
Create a python virtual environnement :
python -m venv env
Enter the virtual environnement:
source env/bin/activate
or (windows powershell)
.\env\Scripts\Activate.ps1
Install the necessary dependencies:
pip install -r requirements.txt
At the project's root
python main.py <map_path> <team1_decision_algo> <team2_decision_algo>
With map_path an optional argument that default to maps/original.txt.
With team1_decision_algo and team2_decision_algo defaulting to utility and strategy_triangle
Here are some extracts of the presentation the team made to our professors.
MAS = Multi Agent System
We implemented 2 team strategies, the first one beeing a utility algorithm. Following are examples of the different parameters taken into account by the utility algorithm.
The second team strategy implemented is a collection of behavior algorithms that are performant in specific situation. An overarching algorithm attributes the best behavior to each members of the team.
