tammerb / mae298-optimal-control-rl

Group project for MAE 298: Optimization and Learning-based Control (Fall 2021)

We aim to use reinforcement learning to develop three- and four-armed robot control policies. The motivating use-cases for this research are autonomous robots for search and rescue, mobile robotic telesurgery and intravehicular robotics in future deep space habitats such as NASA's Gateway. All these applications present mobility challenges due to uncertainties in the environment.

Dependencies:

• pip install gym https://gym.openai.com/
• Mujoco https://www.roboti.us/index.html
• pip install stable-baselines3 https://stable-baselines3.readthedocs.io/
• pip install optuna https://optuna.readthedocs.io/