The goal of this project is to implement Monte Carlo Localization (MCL) to enable a robot to estimate its position within a given environment by integrating sensors.
- Agent Control: Uses Q-Learning algorithm to make decisions and learn which action is the most beneficial. (see
agent.py). - Helper Utilities: Display the agent's performance in real time, including the training progress, score and mean scores over multiple games. (
helper.py). - Model Management: Defines the neural network model and the training process for the reinforcement learning agent which uses Q-Learning (
model.py,modeldirectory). - Robotics Motion Simulation: Simulate and control the robot movments, interaction with obstacles and readings of its surrounding. (
robot_motions.py).
- Clone this repository: git clone https://github.com/nicolas-bock/ROBa_Project.git
- Navigate to the project directory: cd ROBa_Project
- Install the required dependencies: pip install -r requirements.txt Note: Ensure you have Python 3.8 or higher installed.
-
Run the main agent script using Python 3.8 or higher versions : python agent.py
If you the 'python' command alias is not configured, run the main agent using : python3 agent.py