Welcome to the Maze-Solver Robot project! This project leverages reinforcement learning, specifically Q-learning, to navigate and solve mazes. The learning algorithm runs on a laptop, which sends instructions to the robot in real-time. The robot then moves accordingly to solve the maze.
- Q-learning Algorithm: Utilizes reinforcement learning to teach the robot how to solve mazes efficiently.
- Real-Time Control: Sends instructions from the laptop to the robot to navigate the maze in real-time.
- OpenAI Gym Environment: Uses the gym-maze environment for training the model before deploying it on the robot.
- Microcontroller: ESP32
- Motors: 2 DC motors with wheels
- Motor Driver: L298N Motor Driver
- Sensors: 3 infrared sensors for obstacle detection
- Power Supply: Powerbank
- Chassis: Robot chassis kit
- Communication Module: Built-in Wi-Fi module on the ESP32 for communication between the laptop and the robot
- Python 3.6+
- pip (Python package installer)
-
Clone the Repository:
git clone https://github.com/Sharukesh3/Maze-solver-using-q-learning
-
Install the Requirements:
pip install -r requirements.txt
-
Clone the gym_maze Environment:
git clone https://github.com/MattChanTK/gym-maze cd gym-maze python setup.py installNote: The model will be downloaded during the first run.
-
Run the Code:
python qlearning.py
- Operating System: Windows, macOS, or Linux
- RAM: Minimum 4GB
- Dependencies: Listed in requirements.txt
Your feedback is valuable for improving this project. Feel free to contribute or report any issues you encounter.
This project is a prototype for educational purposes and may not work perfectly in all environments. For best results, ensure the maze setup is stable and the robot is properly calibrated.
This project is licensed under the MIT License. See the LICENSE file for details.
Thank you for using the Maze-Solver Robot project! We hope it enhances your understanding of reinforcement learning and robotics.
For any inquiries, please contact us at:
This README file provides a comprehensive guide for users to understand and utilize the Maze-Solver Robot project effectively.