Bayesian Optimisation package for ROS
-
Updated
Nov 17, 2021 - Python
Bayesian Optimisation package for ROS
Stick Labs Pro: verified controller tuning (deadzones, curves, 8 kHz polling, calibration). Windows app and browser version.
On-premise open-source LLMs as reliable structural priors for coupled MIMO controller tuning (IEEE Access)
Bilingual desktop platform for servo controller optimization, built with PyQt5 and MATLAB/Simulink.
Code and compact results for a simulator-grounded benchmark of LLM controller performance judgment
Early stopping Bayesian optimization with prior-fitted networks for controller tuning (bachelor thesis, RWTH Aachen)
To associate your repository with the controller-tuning topic, visit your repo's landing page and select "manage topics."