Reinforcement-Learning-Based-Vehicle-Pathfinding-Simulation

Description

A pathfinding simulation using reinforcement learning, where a car navigates a grid to reach a goal while avoiding obstacles like barriers and cones.

Technologies Used

• Python 3.9
• Pygame
• Numpy
• Matplotlib
• Pandas

Setup Instructions

1. Clone the Repository:

   git clone https://github.com/yourusername/graduation-project.git

2. Install the Required Dependencies:

   pip install pygame numpy matplotlib pandas
   

3. Run the Main Script:

   python main.py

How It Works

This project implements a reinforcement learning-based agent that learns to navigate a grid with obstacles. The car uses Q-learning to explore the environment, gradually improving its decision-making by learning from rewards and penalties.

Car: The agent that moves within the grid. Goal: The target that the agent must reach. Obstacles: Barriers and cones that the car must avoid.

Key Features

• Q-learning Algorithm: Used to learn the best path to the goal.
• Pygame Visualization: Real-time graphical representation of the agent's movement.
• Reward System: The agent receives rewards or penalties based on its actions.

Q-Table Heatmap

Below is the Q-table heatmap, which shows the learned Q-values for each state-action pair and indicates the agent's learned policy.

Reward Table

The reward table displays the total reward accumulated by the agent over each episode and shows how it improves over time.