[Paper Review] Analysis and Design of Secured/Resilient Closed-loop Control Systems.

Sérgio Pequito, Farshad Khorrami+2 more|arXiv (Cornell University)|Nov 10, 2015

Smart Grid Security and Resilience28 references3 citations

TL;DR

This paper introduces resilient fixed-modes free systems to ensure stability and performance in industrial control systems under cyber-physical attacks or failures. By leveraging graph-theoretic representations and convex optimization, it designs minimal, resilient actuation-sensing-communication architectures—demonstrated on power grids—ensuring almost always the absence of destabilizing fixed modes under structural compromise.

ABSTRACT

In this paper, structural properties of distributed control systems and pairing of sensors and actuators are considered to generate architectures which are resilient to attacks/hacks for industrial control systems and other complex cyber-physical systems. In particular, we consider inherent structural properties such as internal fixed modes of a dynamical system depending on actuation, sensing, and interconnection/communication structure for linear discrete time-invariant dynamical systems. We aim to attain stability and performance objectives under disruptive scenarios such as attacks by a malicious agent on actuators, sensors, and communication components and natural failures. The main contributions of this paper are fourfold: (i) introduction of the notion of resilient fixed-modes free system that ensures the non-existence of fixed modes when the actuation-sensing-communication structure is compromised; (ii) a graph-theoretical representation that ensures almost always the non-existence of resilient fixed modes; (iii) a solution to the problem of designing the minimum actuation-sensing-communication structure that ensures a system without resilient fixed modes almost always; and (iv) determine, for a parametrized system, the gain satisfying the sparsity of a given information pattern by resorting to convex optimization tools. The different strategies to deploy actuators, sensors, and establishing communication between these are provided such that the system is resilient fixed-modes free; in particular, in order to show to show the efficacy of the proposed methodologies these results are applied in the context of the power electric grid.

Motivation & Objective

• To address the vulnerability of distributed control systems to malicious attacks on actuators, sensors, and communication links.
• To identify structural properties—specifically resilient fixed modes—that compromise system stability under component failures or cyberattacks.
• To design minimal actuation-sensing-communication topologies that are resilient fixed-modes free, ensuring robustness against structural compromise.
• To determine optimal sparse information patterns via convex optimization for achieving resilience with minimal control effort.
• To validate the framework in a real-world context through application to power electric grids.

Proposed method

• Introduces the concept of resilient fixed-modes free systems, where no fixed modes exist even when actuation, sensing, or communication structures are compromised.
• Develops a graph-theoretic representation to model actuation, sensing, and communication structures, enabling analysis of fixed-mode existence.
• Uses this graph model to derive conditions under which resilient fixed modes almost surely do not exist, ensuring structural resilience.
• Proposes an algorithm to compute the minimal actuation-sensing-communication structure that guarantees resilience by avoiding resilient fixed modes.
• Applies convex optimization techniques to determine gain matrices that satisfy a given sparsity pattern in the information structure, minimizing control complexity.
• Validates the framework through a case study on a power electric grid, demonstrating resilience under attack scenarios.

Experimental results

Research questions

• RQ1 Under what structural conditions does a linear discrete-time-invariant system remain stable when subjected to attacks on actuators, sensors, or communication links?
• RQ2 How can one design an actuation-sensing-communication architecture that is resilient to fixed modes arising from structural compromise?
• RQ3 What is the minimal configuration of sensors, actuators, and communication links that ensures the absence of resilient fixed modes almost surely?
• RQ4 How can sparse information patterns be optimized to maintain resilience while minimizing control complexity?
• RQ5 Can the proposed framework be effectively applied to real-world cyber-physical systems such as power grids?

Key findings

• The proposed resilient fixed-modes free system ensures that no fixed modes exist when the actuation-sensing-communication structure is compromised, guaranteeing stability under disruptive scenarios.
• The graph-theoretic representation enables almost sure non-existence of resilient fixed modes under generic structural perturbations, enhancing system robustness.
• A minimal actuation-sensing-communication architecture can be systematically designed to eliminate resilient fixed modes, reducing system complexity without sacrificing resilience.
• Convex optimization tools successfully determine gain matrices that satisfy a given sparsity pattern in the information structure, enabling efficient and resilient control design.
• The framework is validated on a power electric grid model, demonstrating effective resilience against attacks on sensors, actuators, and communication components.
• The results show that structural resilience can be achieved without requiring additional control effort, by proper placement of sensors and actuators and communication links.