RACE: Recovery Analysis For Cascading Events In Complex Networked Systems

Author

• Sanfan Liu
• Sahra Sedigh Sarvestani, Missouri University of Science and TechnologyFollow
• Ali R. Hurson, Missouri University of Science and TechnologyFollow

Abstract

This paper proposes a stochastic model that elucidates the race between failure propagation and recovery actions in a networked system. The basis of the model is a time-dependent Markov decision process model that predicts the failure and recovery of components and enables comparison of different recovery schemes with respect to one or more aspects of system survivability. Efficacy of the proposed method is illustrated through case studies on smart grids based on the IEEE-14 and IEEE-30 test systems, using high-fidelity digital twins that accurately reflect component states and cascading dynamics. The results demonstrate that the model supports objective comparison of recovery strategies under consistent survivability criteria.

Recommended Citation

S. Liu et al., "RACE: Recovery Analysis For Cascading Events In Complex Networked Systems," Proceedings 2026 IEEE 50th Annual Computers Software and Applications Conference Compsac 2026, pp. 1240 - 1251, Institute of Electrical and Electronics Engineers, Jan 2026.

The definitive version is available at https://doi.org/10.1109/COMPSAC69091.2026.00167

Department(s)

Electrical and Computer Engineering

Second Department

Computer Science

Keywords and Phrases

cascading failure; decision support; failure propagation; Markov decision process; stochastic model; survivability

Document Type

Article - Conference proceedings

Document Version

Citation

File Type

text

Language(s)

English

Rights

© 2026 Institute of Electrical and Electronics Engineers, All rights reserved.

Publication Date

01 Jan 2026