Reinforcement Learning - guidosassaroli/controlbasics GitHub Wiki

Reinforcement Learning (RL) is emerging as a powerful framework for designing controllers in systems with complex, nonlinear, and uncertain dynamics. Unlike traditional model-based control methods, RL does not require an accurate system model and instead learns optimal control policies through interaction with the environment. This makes RL particularly attractive for real-world applications where system dynamics are difficult to model. For a deeper exploration of RL in general, I recommend the book by Sutton and Barto \cite{Sutton1998}. For a focus on RL in control I recommend this article \cite{BUSONIU2018}.

RL Scheme

Example

Work in progress.

In the meantime, consult the following repository: https://github.com/guidosassaroli/RL-cartpole.git.

In this example, a Reinforcement Learning (RL) agent is designed to control the Cart Pole environment: https://www.gymlibrary.dev/environments/classic_control/cart_pole/.