Anomaly Detection on Smart Meters Using Hierarchical Self Organizing Maps

With the growing environmental and technological demands in the energy market, changes in the electrical systems are inevitable. To accommodate these demands advanced metering infrastructures and smart meters are employed. In this paper, we explore a data driven unsupervised learning approach for anomaly detection on smart meters. To this end, we employ and evaluate a hierarchical self organizing map on real world smart meter data. Our results show that different types of anomalies could be detected with an F1-score of over 90 %.