Toward a Knowledge-Based Anomaly Identification System for Detecting Anomalies in the Smart Grid

State Estimation (SE) supports Situation Awareness for Cyber-Physical Systems (CPS). Time-synchronized Phasor Measurement Unit (PMU) measurements are used for SE in a Smart Grid. Data injection attacks on PMU measurements can lead to incorrect estimation of a system’s state and decrease the trustworthiness of SE. Anomaly detection and root cause analysis can help apply appropriate mitigations to the power system. Executing anomaly identification on measurements before estimating the states of a system makes the estimated states trustworthy. In this work, we propose a Knowledge-Based Anomaly Identification System (KBAIS) for detecting an anomaly and identifying its root cause. An ontology, a knowledge graph, and context are used for analysis. We also present a case study - applying the KBAIS to Ecole Polytechnique Federale de Lausanne (EPFL) electrical network. It illustrates anomaly identification in PMU measurements in the context of the EPFL PMUs network.