QLMI

Matlab Implementation of the paper "Model-Free LQR Design by Q-function Learning."

Overview

This repository provides a Matlab implementation of model-free Linear Quadratic Regulator (LQR) controllers. It covers approaches such as:

• Policy Iteration
• Value Iteration
• SDP-Based Convex Optimization (introduced in the paper)

Different formulations of the SDP-based method are provided to design centralized and distributed controllers.

Paper and Citation

If you use this code in your research, please cite the following paper:

"Model-Free LQR Design by Q-function Learning"

• Authors: Milad Farjadnasab, Maryam Babazadeh
• Journal: Automatica
• DOI: https://doi.org/10.1016/j.automatica.2021.110060

Citation in BibTeX format:

@article{farjadnasab2022model,
  title={Model-free LQR design by Q-function learning},
  author={Farjadnasab, Milad and Babazadeh, Maryam},
  journal={Automatica},
  volume={137},
  pages={110060},
  year={2022},
  publisher={Elsevier}
}

License

This project is licensed under the MIT License - see the LICENSE file for details.