English

Scalable regret for learning to control network-coupled subsystems with unknown dynamics

Systems and Control 2021-08-19 v1 Artificial Intelligence Systems and Control Optimization and Control

Abstract

We consider the problem of controlling an unknown linear quadratic Gaussian (LQG) system consisting of multiple subsystems connected over a network. Our goal is to minimize and quantify the regret (i.e. loss in performance) of our strategy with respect to an oracle who knows the system model. Viewing the interconnected subsystems globally and directly using existing LQG learning algorithms for the global system results in a regret that increases super-linearly with the number of subsystems. Instead, we propose a new Thompson sampling based learning algorithm which exploits the structure of the underlying network. We show that the expected regret of the proposed algorithm is bounded by O~(nT)\tilde{\mathcal{O}} \big( n \sqrt{T} \big) where nn is the number of subsystems, TT is the time horizon and the O~()\tilde{\mathcal{O}}(\cdot) notation hides logarithmic terms in nn and TT. Thus, the regret scales linearly with the number of subsystems. We present numerical experiments to illustrate the salient features of the proposed algorithm.

Keywords

Cite

@article{arxiv.2108.07970,
  title  = {Scalable regret for learning to control network-coupled subsystems with unknown dynamics},
  author = {Sagar Sudhakara and Aditya Mahajan and Ashutosh Nayyar and Yi Ouyang},
  journal= {arXiv preprint arXiv:2108.07970},
  year   = {2021}
}

Comments

12 pages

R2 v1 2026-06-24T05:12:40.055Z