Skip to content

Attempt to build and balance an rotary inverted pendulum using reinforcement learning

License

Notifications You must be signed in to change notification settings

ludobouan/inverted-pendulum

Repository files navigation

Inverted Pendulum

Custom implimentation of Q-Learning to try and balance a custom-built rotary inverted pendulum. Algorithm successfully tested on OpenAI's cartpole enviroment. On the real system we were unable to converge to the optimal solution.

The report is available in French.

About

Attempt to build and balance an rotary inverted pendulum using reinforcement learning

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages