Realistic Synchrophasor Data Generation for Anomaly Detection Using Cyber-Power Testbed

Anomaly detection in Phasor Measurement Unit (PMU) data requires high-quality, realistic, labeled datasets for algorithm training and validation. Obtaining real field-labeled data is challenging due to privacy, security concerns, and the rarity of certain anomalies, making a robust testbed indispensable. This paper presents the development and implementation of a Hardware-in-the-Loop (HIL) Synchrophasor Testbed designed for realistic data generation for testing and validating PMU anomaly detection algorithms. The testbed integrates several advanced components, including a Real-Time Digital Simulator (RTDS), hardware PMUs, a Real-Time Automation Controller (RTAC), the PingThings cloud platform, and the ns-3 network simulator. This comprehensive setup allows for emulating complex power system dynamics and injecting controlled events and anomalies under various operating conditions. Leveraging this testbed, high-fidelity, time-synchronized data that accurately represents real-world scenarios, facilitates the development, testing, and validation of robust anomaly detection algorithms. Data sets corresponding to scenarios such as normal operation, missing data, power events, and cyber-induced anomalies are created using the testbed. The proposed testbed is used to create multiple data sets to train and validate anomaly detection algorithms. The datasets created using the testbed are subsequently processed through an existing anomaly detection tool, and the results are validated. The data set developed in this work has been made public in the IEEE data port, and it has been used to conduct an anomaly detection competition at the International Conference on Smart Grid Synchronized Measurements & Analytics (SGSMA), Washington DC, May 2024.