Self driving car with crossover in genetic algorithms from scratch

Introduction

This project explores the implementation of a Genetic Algorithm to train a virtual self-driving car with a simple neural network, focusing on the crossover operator, entirely in vanilla JavaScript (no external libraries). The objective is for the car to navigate a simulated environment, avoid obstacles, and complete a track autonomously, leveraging the power of genetic optimization.

You can try it out right here.

Sample results

Technologies Used

• JavaScript: The sole programming language used for this project, utilizing standard browser APIs.
• HTML5 Canvas: Used to draw the simulated environment, car, track, and obstacles.
• No External Libraries: This project is built without any third-party libraries or frameworks, providing a deeper understanding of the underlying implementation.
• Genetic Algorithm:
  • Crossover: The crossover operator generates new car "genomes" by combining the "genetic" material of the most successful individuals. In this project, we use a custom crossover where the best left-turning car's genome is combined with the best right-turning car's genome.
  • Selection: The fittest cars are selected to pass their genes on to the next generation.
  • Mutation: Adds random variations to the car genomes, preventing premature convergence.

How to Run the Project

1. Clone the repository:

   git clone [repository-url]
   cd [repository-directory]

2. Open index.html in your browser: Simply double-click the index.html file or use a local server to view the simulation in your web browser.

Requirements

• No specific requirements beyond a web browser: Any modern web browser that supports HTML5 and JavaScript will be able to run the project. No additional software installations or libraries are necessary.

Contributing

Your contributions to this project are welcomed. Feel free to submit pull requests if you have improvements.