rlrobotics

Reinforcement Learning for Robotics multistable based legged robots.

How to run the code ?

There is 3 main categories.

1. Controller Place the mapping table in the controller folder as _all_sequences.pkl install requirements pip install -r requirements.txt Run main.py to start the Deep Q learning process or use Visual studio code debug mode and select Run AI
2. Energy Run energy.py to produce the plot
3. robot install requirements pip install -r requirements.txt Run the selected simulation you want or the mapping table generator with debug monde on Visual Studio code.

Code documentation

All the documentation is inside the code, especially for the Robot section where a full sphinx style documentation 
is created.

Reinforcement learning:

The following video shows how the robot learns to maneuver and reaches the selected targets using only two sequences of A and B.

Learning-A-B.mp4