Integration of Fiber Optic Sensing and Sparse Grid Sensors for Accurate Fault Localization in Distribution Systems

Publication Date: 7/27/2025

Event: 2025 IEEE PES General Meeting

Reference: Austin, TX

Authors: Md. Zahidul Islam, NEC Laboratories America, Inc., New York University; Yangmin Ding, NEC Laboratories America, Inc.; Yue Tian, NEC Laboratories America, Inc.; Ting Wang, NEC Laboratories America, Inc.; Yuzhang Lin, New York University,

Abstract: Fault localization in power distribution networks is essential for rapid recovery and enhancing system resilience. While Phasor Measurement Units (PMUs or ?PMUs) providehigh-resolution measurements for precise fault localization, their widespread deployment is cost-prohibitive. Distributed Fiber Optic Sensing (DFOS) offers a promising alternative for event detection along power lines using collocated optical fiber; however, it cannot independently differentiate between events and pinpoint exact fault locations. This paper introduces an innovative framework that combines DFOS with sparsely deployed PMUs for accurate fault localization. The proposed approach first utilizes a Graph Attention Network (GAT) model to capture spatial and temporal correlations from synchronized PMU and DFOS measurements, effectively identifying fault zones. High-spatial- resolution DFOS measurements further refine the fault locationwithin the identified zone. Singular Value Decomposition (SVD) is applied to extract feature vectors from DFOS measurements, enhancing the convergence speed of the GAT model. Thisintegrated solution significantly improves localization accuracy while minimizing reliance on extensive deployment of PMUs.

Publication Link: