This repository contains the code and documentation for my undergraduate thesis on the traffic environment complexity calculation model. The thesis discusses the importance of traffic environment complexity in autonomous vehicle testing and safety evaluation, and proposes a novel method for calculating both static and dynamic traffic environment complexity.

[Include the abstract from your thesis here]

• code/: Contains the source code for the complexity calculation models and RL environment.
  • static_complexity.py: Static complexity calculation code.
  • dynamic_complexity.py: Dynamic complexity calculation code.
  • traffic_env.py: Custom traffic environment for reinforcement learning.
  • train_rl_model.py: Script to train RL model using TF-Agents.
• data/: Contains sample data used in the models.
• docs/: Contains the thesis and related documentation.

1. Ensure you have Python 3.6+ installed.
2. Install the necessary Python packages:

   pip install tensorflow tf-agents gym numpy matplotlib

1. To run the static complexity calculation, use the following command:

   python code/static_complexity.py

1. To run the dynamic complexity calculation, use the following command:

   python code/dynamic_complexity.py

1. To test the custom traffic environment, use the following command:

   python code/traffic_env.py

1. To train the RL model using TF-Agents, use the following command:

   python code/train_rl_model.py

• static_complexity.py: This script calculates the static complexity of the traffic environment using information entropy and grey relational analysis.
• dynamic_complexity.py: This script calculates the dynamic complexity of the traffic environment using a gravitational model to analyze dynamic elements.
• traffic_env.py: This script defines a custom Gym environment for the traffic scenario, incorporating both static and dynamic complexity calculations.
• train_rl_model.py: This script uses TF-Agents to train a reinforcement learning agent in the custom traffic environment.

1. After training the RL model, the script will visualize the training results using matplotlib.

This project is licensed under the MIT License - see the LICENSE file for details.