ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 2111.03772
25
9

Dynamic Regret Minimization for Control of Non-stationary Linear Dynamical Systems

6 November 2021
Yuwei Luo
Varun Gupta
Mladen Kolar
ArXivPDFHTML
Abstract

We consider the problem of controlling a Linear Quadratic Regulator (LQR) system over a finite horizon TTT with fixed and known cost matrices Q,RQ,RQ,R, but unknown and non-stationary dynamics {At,Bt}\{A_t, B_t\}{At​,Bt​}. The sequence of dynamics matrices can be arbitrary, but with a total variation, VTV_TVT​, assumed to be o(T)o(T)o(T) and unknown to the controller. Under the assumption that a sequence of stabilizing, but potentially sub-optimal controllers is available for all ttt, we present an algorithm that achieves the optimal dynamic regret of O~(VT2/5T3/5)\tilde{\mathcal{O}}\left(V_T^{2/5}T^{3/5}\right)O~(VT2/5​T3/5). With piece-wise constant dynamics, our algorithm achieves the optimal regret of O~(ST)\tilde{\mathcal{O}}(\sqrt{ST})O~(ST​) where SSS is the number of switches. The crux of our algorithm is an adaptive non-stationarity detection strategy, which builds on an approach recently developed for contextual Multi-armed Bandit problems. We also argue that non-adaptive forgetting (e.g., restarting or using sliding window learning with a static window size) may not be regret optimal for the LQR problem, even when the window size is optimally tuned with the knowledge of VTV_TVT​. The main technical challenge in the analysis of our algorithm is to prove that the ordinary least squares (OLS) estimator has a small bias when the parameter to be estimated is non-stationary. Our analysis also highlights that the key motif driving the regret is that the LQR problem is in spirit a bandit problem with linear feedback and locally quadratic cost. This motif is more universal than the LQR problem itself, and therefore we believe our results should find wider application.

View on arXiv
Comments on this paper