Little self learning racing car made in python and a genetic algorithm implementation. It's mainly made with Pygame and NEAT.
- Python 3
- Python package manager (pip)
Use the package manager pip to install the requirements.
pip install -r requirements.txt
- Head to simulation folder
- Run
python main.py
Quick note: The amount of sensors can only be: 3, 4, 5, 9 or 11
- Go to simulation folder
- Edit the file .env and change the value of NUM_SENSORES to how many sensors you'd like
- Save changes and close the file
- While still in the folder, edit the file config-feedforward.txt
- Go to line 48 and change the value of num_inputs to match the amount of sensors in step 2
- Save changes and close the file
- Run
python main.py
Quick note: The map can only be 1, 2 or 3. 1 is the easiest map, meanwhile 3 is the hardest.
- Go to simulation folder
- Edit the file .env and the value of MAP to the difficulty of the map you'd like
- Save changes and close the file
- Run
python main.py
- Oscar Juárez
- José Pablo Cifuentes
Universidad del Valle de Guatemala - 2020