Snakeeyy: A Turtle-Based Snake Game with Pathfinding and Q-Learning

Overview

Snakeeyy is a Python-based Snake game built with the turtle module. It features a grid-based environment (20x20 grid) with wrap-around boundaries and an AI-driven snake using breadth-first search (BFS) pathfinding and Q-learning elements to navigate toward food while avoiding collisions. The game includes dynamic speed control, score tracking, and a reward system to guide the snake’s behavior.

Features

• Gameplay: Classic Snake game where the snake grows by eating food (light coral circles) and avoids self-collision or excessive steps (max 2000 steps per episode).
• AI: Combines BFS for safe pathfinding to food with a partial Q-learning approach for action selection (up, left, down, right).
• Controls:
  • Up/Down arrows: Adjust movement speed (0.01–0.2 s/step).
  • Space: Pause/resume the game.
• Rewards: Food collection (+20), wall crossings (+0.3), high score (+50), step penalty (-0.1), turn penalty (-0.1), collision penalty (-1000).
• Score System: Tracks current and high scores, saved to Highscore.txt.

Technologies

• Python Libraries:
  • turtle: Renders the 700x750 game interface with a light green background.
  • time: Controls game speed via delays.
  • random: Generates random food positions, avoiding snake overlap.
  • math: Computes distances for collision detection.
  • collections.deque: Implements BFS for pathfinding.
• AI Components:
  • BFS: Finds collision-free paths to food (find_safest_path).
  • Q-Learning: Partial implementation with a Q-table, learning rate (alpha=0.2), discount factor (gamma=0.9), and exploration rate (epsilon=0.3 with decay), but lacks Q-table updates.

Installation

1. Prerequisites: Python 3.6+ with standard libraries (turtle, time, random, math, collections).
2. Clone or download the repository:

   git clone https://github.com/adi-devv/Snake-Game-RL.git
   cd Snake-Game-RL