Robust Defense Framework for Active Distribution Networks Under Multiple Attacks

Cong Cai

doi:10.1109/TIE.2025.3639750

Robust Defense Framework for Active Distribution Networks Under Multiple Attacks

2025-12-30

0

PRE

AI

DOI：10.1109/TIE.2025.3639750

Original

Original request for help

Save

Abstract

En 中文

This article proposes a robust defense framework for hierarchical active distribution networks under multiple malicious attacks including false data injections (FDIs) and line disconnections. The framework integrates an adaptive augmented observer and a Defender-Attacker-Optimizer (D-A-O) algorithm against different layers of attacks. The adaptive augmented observer reconstructs the FDI signals and sends them to the controller as compensation values to defend against controller FDI attacks. In addition, the observer transmits the observed system states which are used in the D-A-O algorithm. Due to the observed true states of the system, the observer can defend against interlayer FDI attacks from the physical layer to the cloud decision layer. The D-A-O algorithm, based on Column-and-Constraint Generation (CC&G), is designed to achieve lower cost of the system under line disconnection scenarios. The proposed framework demonstrates stronger attack defense and lower cost through simulation and hardware-in-the-loop (HIL) testing of an IEEE 33-bus system.

Keywords:

Active distribution networks (ADNs)

cyber-physical attacks

false data injection (FDI)

hardware-in-the-loop (HIL) simulation

robust optimization

Robust Defense Framework for Active Distribution Networks Under Multiple Attacks

Abstract

Journal

Organization