Distributed fault detection and isolation resilient to network model uncertainties.

Area of Science:

• Control Systems Engineering
• Networked Systems Analysis
• Fault Tolerance

Background:

• Resilient control systems require maintaining state awareness amidst unexpected errors and threats.
• Large networked systems with uncertain models present significant challenges for fault detection and isolation (FDI).

Purpose of the Study:

• To develop a distributed fault detection and isolation (FDI) scheme for large networked systems with uncertain models.
• To enhance the resilience of FDI methods against model uncertainties and external system changes.

Main Methods:

• Propose a distributed scheme for joint fault detection and isolation in both subsystems (nodes) and interconnections (edges).
• Analyze scheme behavior under model uncertainties arising from dynamic network topology (edge addition/removal).
• Develop a novel distributed FDI scheme utilizing local models and measurements resilient to external changes.

Main Results:

• A distributed FDI scheme capable of detecting and isolating faults in nodes and edges of networked systems.
• Demonstrated resilience of the FDI scheme to model uncertainties and dynamic network changes.
• Achieved complexity reduction by identifying minimal information requirements and reducing monitoring nodes.

Conclusions:

• The proposed methods enable a scalable and resilient distributed FDI architecture.
• The architecture achieves local FDI even with unknown external system modifications.
• Validated through numerical experiments on the IEEE 118-bus power network.