guoyaq / tensegrity

Tensegrity Leg Control Using Reinforcement learning

This Project is developed as the final course project for the course ECEN 689 603, Reinforcement learning, Texas A&M University.

The project includes Mujoco Model of a tensegrity leg and its corresponding gym environments using openai gym along with these we provide training files to train the tensegrity leg model to walk using stable baselines PPPO2 algorithm.

Chekcout the youtube video of Tensegrity Leg walking using PPO2.

Installation

Dependencies

• Python 3.6
• mujoco-py.
• stable baselines.

Install tensegrity

After installing the dependencies install tensegrity-

pip install -e tensegrity

Training

 python tensegrity/env/leg_train.py