Skip to content

Latest commit

 

History

History
38 lines (27 loc) · 1.42 KB

README.md

File metadata and controls

38 lines (27 loc) · 1.42 KB


Racing Towards Reinforcement Learning based control of an Autonomous Formula SAE Car

Instructions To Run The Code

Please refer to instructions.md for in-depth instructions on how to run the code and replicate our experiments.

State-Space

Represented by x (lateral) and z (forward) displacements of 6 closest (sorted by z displacement) Aruco markers relative to robot position.

Action-Space

Linear Speed = Constant 0.5 m/s for both models.

DQN - Discrete Action Space +/- 0.2 rad/s rotatational speed. TD3 - Continuous Action Space from +0.4 rad/s to -0.4 rad/s rotational speed.

Results

Both models trained and tested in simulation. Example of simulation POV of robot shown below.

Results of training:

Video

A video showing the robot executing the tasks during the training process can be found at this youtube link.

Citation

If you use either the code or data in your paper, please kindly star this repo and cite our paper Cite this paper as: Coming soon

Contact

Please feel free to contact us or open an issue if you have questions or need additional explanations.