Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Deep Learning-Enabled System Diagnosis in Microgrids: A Feature-Feedback GAN Approach

Swetha Rani Kasimalla, Kuchan Park, Junho Hong, Young-Jin Kim, HyoJong Lee

TL;DR

The paper tackles the dual challenge of detecting internal faults and adversarial false data injection (FDI) attacks in inverter-based microgrids. It introduces a two-stage framework where an unsupervised Feature Feedback GAN (F2GAN) discriminates real faults from cyber anomalies using feature matching, followed by supervised fault localization with classifiers such as SVM, KNN, DT, and ANN. Empirical results in a MATLAB/Simulink microgrid with 1,097 labeled fault samples show that F2GAN achieves higher AUC and accuracy than a conventional GAN, and that ANN achieves near-perfect fault classification accuracy. The approach enhances microgrid resilience by robustly diagnosing faults and cyber distortions, with practical implications for real-time protection and control under zero-day attack conditions.

Abstract

The increasing integration of inverter-based resources (IBRs) and communication networks has brought both modernization and new vulnerabilities to the power system infrastructure. These vulnerabilities expose the system to internal faults and cyber threats, particularly False Data Injection (FDI) attacks, which can closely mimic real fault scenarios. Hence, this work presents a two-stage fault and cyberattack detection framework tailored for inverter-based microgrids. Stage 1 introduces an unsupervised learning model Feature Feedback Generative Adversarial Network (F2GAN), to distinguish between genuine internal faults and cyber-induced anomalies in microgrids. Compared to conventional GAN architectures, F2GAN demonstrates improved system diagnosis and greater adaptability to zero-day attacks through its feature-feedback mechanism. In Stage 2, supervised machine learning techniques, including Support Vector Machines (SVM), k-Nearest Neighbors (KNN), Decision Trees (DT), and Artificial Neural Networks (ANN) are applied to localize and classify faults within inverter switches, distinguishing between single-switch and multi-switch faults. The proposed framework is validated on a simulated microgrid environment, illustrating robust performance in detecting and classifying both physical and cyber-related disturbances in power electronic-dominated systems.

Deep Learning-Enabled System Diagnosis in Microgrids: A Feature-Feedback GAN Approach

TL;DR

The paper tackles the dual challenge of detecting internal faults and adversarial false data injection (FDI) attacks in inverter-based microgrids. It introduces a two-stage framework where an unsupervised Feature Feedback GAN (F2GAN) discriminates real faults from cyber anomalies using feature matching, followed by supervised fault localization with classifiers such as SVM, KNN, DT, and ANN. Empirical results in a MATLAB/Simulink microgrid with 1,097 labeled fault samples show that F2GAN achieves higher AUC and accuracy than a conventional GAN, and that ANN achieves near-perfect fault classification accuracy. The approach enhances microgrid resilience by robustly diagnosing faults and cyber distortions, with practical implications for real-time protection and control under zero-day attack conditions.

Abstract

The increasing integration of inverter-based resources (IBRs) and communication networks has brought both modernization and new vulnerabilities to the power system infrastructure. These vulnerabilities expose the system to internal faults and cyber threats, particularly False Data Injection (FDI) attacks, which can closely mimic real fault scenarios. Hence, this work presents a two-stage fault and cyberattack detection framework tailored for inverter-based microgrids. Stage 1 introduces an unsupervised learning model Feature Feedback Generative Adversarial Network (F2GAN), to distinguish between genuine internal faults and cyber-induced anomalies in microgrids. Compared to conventional GAN architectures, F2GAN demonstrates improved system diagnosis and greater adaptability to zero-day attacks through its feature-feedback mechanism. In Stage 2, supervised machine learning techniques, including Support Vector Machines (SVM), k-Nearest Neighbors (KNN), Decision Trees (DT), and Artificial Neural Networks (ANN) are applied to localize and classify faults within inverter switches, distinguishing between single-switch and multi-switch faults. The proposed framework is validated on a simulated microgrid environment, illustrating robust performance in detecting and classifying both physical and cyber-related disturbances in power electronic-dominated systems.
Paper Structure (20 sections, 15 equations, 4 figures, 3 tables, 1 algorithm)

This paper contains 20 sections, 15 equations, 4 figures, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Contributions
  4. Paper Outline
  5. Microgrid Framework and Threat Model
  6. Microgrid framework and Vulnerability Zones
  7. Communication Layer
  8. Microgrid Central Controller Layer
  9. Physical Layer (DERs and Local Controllers)
  10. Inverter Fault Scenario
  11. Mathematical Modeling and Evaluation Methods of Proposed Framework
  12. Mathematical Modeling of F2GAN
  13. Modelling of FDI Attack
  14. Mathematical Modeling of Fault Classification
  15. Support Vector Machine (SVM)
  16. ...and 5 more sections

Figures (4)

  • Figure 1: Multi-layer microgrid framework illustrating DER integration used to simulate inverter faults. The resulting data is collected for deep learning-based fault and anomaly analysis.
  • Figure 2: The F2GAN architecture distinguishes between real inverter faults and FDI attacks and is coupled with supervised learning to classify and localize the real faults into specific single or multiple switch fault categories.
  • Figure 3: Performance comparison and system-level visualization across models and data layers.
  • Figure : Two-Stage Framework: F2GAN-based Detection and Supervised Fault Classification