[Paper Review] Robust periodic stability implies uniform exponential stability of Markovian jump linear systems and random linear ordinary differential equations
This paper establishes that robust periodic stability implies uniform exponential stability for both discrete-time Markovian jump linear systems and random linear ODEs driven by semiflows with the closing by periodic orbits property. Using a novel Gel'fand-Berger-Wang formula derived via Lyapunov exponent approximation by periodic orbits and Shantao Liao’s perturbation techniques, the authors prove equivalence between robust periodic stability and uniform exponential stability, resolving long-standing questions on spectral finiteness and robustness.
In this paper we show that if a linear cocycle is robustly periodical stable then it is uniformly stable.
Motivation & Objective
- To establish the equivalence between robust periodic stability and uniform exponential stability in discrete-time Markovian jump linear systems.
- To extend this equivalence to random linear ODEs driven by semiflows with the closing by periodic orbits property.
- To resolve open questions regarding spectral finiteness and robustness conditions in linear cocycle theory.
- To develop a Gel’fand-Berger-Wang formula for Markovian jump systems using periodic orbit approximation of Lyapunov exponents.
- To construct counterexamples to the spectral finiteness conjecture and the robustness condition in linear cocycle theory.
Proposed method
- Derives a Gel’fand-Berger-Wang formula for Markovian jump linear systems by approximating Lyapunov exponents using periodic orbits.
- Applies Shantao Liao’s perturbation technique from differentiable dynamical systems to analyze stability in random linear ODEs.
- Uses semi-uniform ergodic theorems to relate the joint spectral radius to the behavior on nonwandering sets and periodic orbits.
- Employs ergodic theory to analyze the stability of linear cocycles over dynamical systems with the closing by periodic orbits property.
- Constructs counterexamples to the spectral finiteness of linear cocycles and to the robustness condition in stability analysis.
- Establishes equivalence between uniform exponential stability and robust periodic stability via topological and measure-theoretic arguments on symbolic spaces and semiflows.
Experimental results
Research questions
- RQ1Does robust periodic stability imply uniform exponential stability in discrete-time Markovian jump linear systems?
- RQ2Can the equivalence between robust periodic stability and uniform exponential stability be extended to random linear ODEs driven by semiflows with the closing by periodic orbits property?
- RQ3Is the spectral finiteness property valid for general linear cocycles over dynamical systems with periodic orbit structure?
- RQ4Can a Gel’fand-Berger-Wang formula be established for Markovian jump systems without assuming irreducibility or aperiodicity of the transition matrix?
- RQ5What are the limitations of the robustness condition in stability analysis of linear switching systems?
Key findings
- A discrete-time Markovian jump linear system is uniformly exponentially stable if and only if it is robustly periodically stable, under no additional assumptions on the Markov transition matrix.
- For random linear ODEs driven by a semiflow with the closing by periodic orbits property, uniform exponential stability is equivalent to robust periodic stability.
- A new Gel’fand-Berger-Wang formula is proven for Markovian jump systems by approximating Lyapunov exponents via periodic orbits.
- The joint spectral radius of a linear cocycle over a system with closing by periodic orbits property equals the infimum over max norms on periodic orbits.
- Counterexamples are constructed showing that spectral finiteness does not hold in general for linear cocycles.
- The robustness condition—uniform boundedness of spectral radii over periodic orbits—is insufficient to guarantee uniform exponential stability, highlighting a subtle distinction in stability criteria.
Better researchstarts right now
From reading papers to final review, dramatically reduce your research time.
No credit card · Free plan available
This review was created by AI and reviewed by human editors.