Control-Coupled Structural Decomposition and Physics-Informed Anomaly Detection for Resilient Inverter-Dominated Microgrids

Nasr Esfahani, Fatemeh and Wu, Yueqi and Galeela, Mohamed and Ma, Xiandong (2026) Control-Coupled Structural Decomposition and Physics-Informed Anomaly Detection for Resilient Inverter-Dominated Microgrids. In: ICIEA 2026- The 21st IEEE Conference on Industrial Electronics and Applications :. IEEE, ITA. (In Press)

Text (ICIEA_2026)
2026204784_1_.pdf - Accepted Version
Restricted to Repository staff only until 2 August 2026.
Available under License Creative Commons Attribution.
Download (1MB)

Abstract

High penetration of inverter-based resources (IBRs) is increasing the complexity of disturbance propagation and resilience analysis in microgrids. This paper presents a control-coupled structural model decomposition (SMD) framework for inverter-dominated microgrids, enabling partition-aware disturbance analysis, anomaly detection, and localisation across electrically and control-defined regions. A modified CERTS-based PSCAD benchmark is developed with grid-forming (GFM) and grid-following (GFL) converters, distributed loads, and detailed converter control dynamics. The proposed framework extends conventional structural decomposition by integrating dq-domain electrical interfaces with converter control pathways, including phase-locked loop (PLL) synchronisation, droop regulation, and DC-link dynamics. Interface residuals are formulated to detect boundary inconsistencies, assess disturbance severity, discriminate anomaly classes, and localise dominant disturbed interfaces. Simulation studies under structural faults, load disturbances, and control-layer anomalies demonstrate that different disturbance classes produce distinct residual signatures and propagation patterns. Results confirm accurate anomaly detection, scenario discrimination, and partition-aware localisation, with structural faults producing sharp transient residuals and control anomalies producing persistent oscillatory residuals.