MLPro 2.1.0 released
We are pleased to announce today that we have expanded MLPro with two new sub-frameworks related to closed-loop control:
MLPro-BF-Control: classic closed-loop control
MLPro-OA-Control: online-adaptive closed-loop control
As part of this development, we created an online-adaptive PID controller that tunes itself using a reinforcement learning algorithm.
The online documentation is not yet complete, but initial howtos are available. We have also described the key innovations and the RLPID architecture in our paper “Online-adaptive PID control using Reinforcement Learning”. The experiment for the paper can be found in the GitHub repository fhswf/paper-da-ieee-codit-2025. Have fun trying it out! We welcome your feedback.
For a detailed list of changes, please refer to the changelog on GitHub.