English

Revisiting LQR Control from the Perspective of Receding-Horizon Policy Gradient

Optimization and Control 2024-02-02 v3 Artificial Intelligence Machine Learning Systems and Control Systems and Control

Abstract

We revisit in this paper the discrete-time linear quadratic regulator (LQR) problem from the perspective of receding-horizon policy gradient (RHPG), a newly developed model-free learning framework for control applications. We provide a fine-grained sample complexity analysis for RHPG to learn a control policy that is both stabilizing and ϵ\epsilon-close to the optimal LQR solution, and our algorithm does not require knowing a stabilizing control policy for initialization. Combined with the recent application of RHPG in learning the Kalman filter, we demonstrate the general applicability of RHPG in linear control and estimation with streamlined analyses.

Keywords

Cite

@article{arxiv.2302.13144,
  title  = {Revisiting LQR Control from the Perspective of Receding-Horizon Policy Gradient},
  author = {Xiangyuan Zhang and Tamer Başar},
  journal= {arXiv preprint arXiv:2302.13144},
  year   = {2024}
}