debryu / autonomous-software-agents

Using BDI model to run multiple agents that have to compete against other opponents in a simulated world. The agents uses Ant Colony Optimization for pathfinding.

The authors are Nicola Debole and Mateo Rodriguez. This work is the project for the Autonomous Software Agents course at the University of Trento.

Usage

First clone the repository

git clone https://github.com/debryu/Autonomous-Software-Agents.git

Then move inside the folder and run

npm install

find the file mainframe.js located in lib/deliverooClient/agent_folder/mainframe.js and in the same path from the terminal run

node mainframe.js "CHOOSE A NAME FOR THE AGENT" "TOKEN"

Example you can run separately this two commands:

node mainframe.js AgentSmith_0 eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpZCI6IjhkZTYzMmJiNjYwIiwibmFtZSI6IkFnZW50U21pdGhfMCIsImlhdCI6MTY4NTk3NzExMn0.S9_nuZTVO5k6yQXyMNz4ia8YuJODnJJ8RqF36Mo56XE

node mainframe.js AgentSmith_2 eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpZCI6ImRlNjMyYmI2NjA0IiwibmFtZSI6IkFnZW50U21pdGhfMiIsImlhdCI6MTY4NTk3NzE3MX0.obnPhigqA-PDznKikl33TIuHdDYg4kvmrkQUIXH12KY

---

The token can be retrieved by connecting to https://deliveroojs.onrender.com/ and filling with the name of the agent when asked.

You need to run two different agents otherwise the first one will just wait endlessly for the other.

Flowchart

A flowchart of a simplified version of our code to better understand it at first glance.

Features

• Ant Colony Optimization (ACO)
• PDDL
• Beliefs-Desire-Intention (BDI)