An Anomaly Detection Method for Smart Power Grid: A Federated Learning Framework

As the scale and complexity of power grid continue to increase, some abnormal events pose a significant threat to the security and stability of the power grid. In this paper, we propose an anomaly detection method for power grid based on a federated learning framework, where the distributed data and model training are utilized, thus ensuring the privacy protection and realizing the anomaly detection. This method is comprised of multiple Edge Devices and a central server. Edge Devices perform local model training and upload the model parameters to the central server for parameter aggregation. Through the process of federated learning parameter aggregation, the global model is progressively improved, and the anomaly detection of power grid (without sharing raw data) can be realized. With this method, the power grid operators can promptly identify the abnormal events happening in the power grid and take appropriate measures to ensure the secure and stable operations. The simulation results demonstrate that the federated learning-based anomaly detection method for power grid can provide an innovative solution to achieve the distributed, privacy-preserving, and efficient real-time monitoring for power grid.

Paper

Full text

PDF

An Anomaly Detection Method for Smart Power Grid: A Federated Learning Framework

Semantic Scholar · Engineering · 2023

Abstract

As the scale and complexity of power grid continue to increase, some abnormal events pose a significant threat to the security and stability of the power grid. In this paper, we propose an anomaly detection method for power grid based on a federated learning framework, where the distributed data and model training are utilized, thus ensuring the privacy protection and realizing the anomaly detection. This method is comprised of multiple Edge Devices and a central server. Edge Devices perform local model training and upload the model parameters to the central server for parameter aggregation. Through the process of federated learning parameter aggregation, the global model is progressively improved, and the anomaly detection of power grid (without sharing raw data) can be realized. With this method, the power grid operators can promptly identify the abnormal events happening in the power grid and take appropriate measures to ensure the secure and stable operations. The simulation results demonstrate that the federated learning-based anomaly detection method for power grid can provide an innovative solution to achieve the distributed, privacy-preserving, and efficient real-time monitoring for power grid.

Similar papers

© 2026 NYSGPT2525 LLC