Rank resilience of pattern matrices against structured perturbations with applications

DOI10.1016/J.AUTOMATICA.2025.112236MaRDI QIDQ6993267zbMATH OpenFDO

Authors Yuan Zhang, Yuanqing Xia, Gang Wang

Publication date 25 April 2025

Published in Automatica (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/J.AUTOMATICA.2025.112236

zbMATH Keywords

algebraic combinatorial approachgeneric propertieslow-rank matrix completionrank resiliencestructured systems

Mathematics Subject Classification ID

Controllability (93B05) Observability (93B07) Algebraic methods (93B25) Perturbations in control/observation systems (93C73)

Review text

This paper studies how the rank of a pattern matrix changes when some of its entries are modified in a structured way, meaning the pattern of zeros and nonzeros stays fixed while the values may vary. This is important for understanding how systems like networks or control systems can keep working even when some parts are affected by faults or attacks.\N\NThe authors introduce the idea of generic rank resilience (GRR): the ability of a matrix to keep a certain rank under almost all possible structured changes. They prove that rank resilience is a generic property, meaning it either holds in most cases or not at all. They also give combinatorial tests to check when a matrix loses rank and show how to use these ideas in structured systems, including systems with descriptor dynamics.\N\NApplications include checking the controllability and observability of systems and finding conditions for avoiding fixed modes that limit control design. The work combines algebraic geometry, graph theory, and matrix analysis, offering useful tools for researchers in control theory, network systems, and system robustness.

Cites work

• A sparse effective Nullstellensatz
• An overview of structural systems theory
• Generic properties and control of linear structured systems: A survey
• Invariant subspaces and invertibility properties for singular systems: The general case
• Minimal structural perturbations for controllability of a networked system: complexities and approximations
• Minimum jointly structural input and output selection
• Observability Preservation Under Sensor Failure
• Observability Robustness Under Sensor Failures: A Computational Perspective
• On real structured controllability/stabilizability/stability radius: complexity and unified rank-relaxation based methods
• On Structural Rank and Resilience of Sparsity Patterns
• On the stabilization of decentralized control systems
• Perturbation-tolerant structural controllability for linear systems
• Secure Estimation and Control for Cyber-Physical Systems Under Adversarial Attacks
• Solvability, controllability, and observability of continuous descriptor systems
• Structural controllability of multi-agent networks: robustness against simultaneous failures
• The algebraic combinatorial approach for low-rank matrix completion
• The Observability Radius of Networks
• The Power of Convex Relaxation: Near-Optimal Matrix Completion
• The robust minimal controllability problem
• Typical and generic ranks in matrix completion
• Uniqueness of low-rank matrix completion by rigidity theory

This page was built for publication: Rank resilience of pattern matrices against structured perturbations with applications

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6993267)