English

Approximating the nearest stable discrete-time system

Optimization and Control 2019-03-29 v3 Numerical Analysis

Abstract

In this paper, we consider the problem of stabilizing discrete-time linear systems by computing a nearby stable matrix to an unstable one. To do so, we provide a new characterization for the set of stable matrices. We show that a matrix AA is stable if and only if it can be written as A=S1UBSA=S^{-1}UBS, where SS is positive definite, UU is orthogonal, and BB is a positive semidefinite contraction (that is, the singular values of BB are less or equal to 1). This characterization results in an equivalent non-convex optimization problem with a feasible set on which it is easy to project. We propose a very efficient fast projected gradient method to tackle the problem in variables (S,U,B)(S,U,B) and generate locally optimal solutions. We show the effectiveness of the proposed method compared to other approaches.

Keywords

Cite

@article{arxiv.1802.08033,
  title  = {Approximating the nearest stable discrete-time system},
  author = {Nicolas Gillis and Michael Karow and Punit Sharma},
  journal= {arXiv preprint arXiv:1802.08033},
  year   = {2019}
}

Comments

15 pages, new title, accepted in LAA