Development of Deep Reinforcement Learning Policy in a Parking Simulated Environment

Overview

In this project, I've developed an autonomous parking system using the Deep Deterministic Policy Gradient (DDPG) algorithm within the Carla API environment. DDPG empowers the system to effectively learn and optimize its parking policies, resulting in improved performance and increased success rates in navigating challenging parking scenarios. By integrating DDPG with the Carla API, we've created a stable and efficient solution capable of real-time adaptation and continuous learning, contributing to the advancement of autonomous driving research and applications.

Getting Started

To run this program, follow these steps:

Clone the Repository: Clone this project into a new folder within the PythonAPI folder provided by the Carla Simulator. Make sure you have Carla Simulator version 0.9.10 installed and configured.
```
git clone https://github.com/your-username/your-repo.git
```
Start Carla Simulator: Run CarlaUE4.exe, the simulator provided by Unreal Engine.
Run the Program: Open a Command Prompt and navigate to your project folder, then execute the program using Python 3.7.
```
py -3.7 main.py
```

Main Program

The core of this project is located in main.py. This script contains the implementation of the autonomous parking system using the DDPG algorithm.

Video Demonstration

For a visual overview of the project's best behaviors, you can find demonstration videos in the best_behavior folder.

Contributions and Collaborations

If you're interested in contributing to this project or collaborating on further developments, please follow the guidelines in the "Contributing" section of this README. I welcome contributions and ideas from the community to enhance and expand the capabilities of this autonomous parking system.

Acknowledgments

I would like to acknowledge the Carla Simulator development team and the creators of the Deep Deterministic Policy Gradient (DDPG) algorithm for their valuable contributions to this project.

Feel free to adapt and expand this README further as your project evolves or based on the feedback and requirements of your audience. It's a great start to inform users and potential collaborators about your work!