This paper proposes a novel method for detecting anomalies in Phasor Measurement Units (PMUs) data leveraging quantum computing techniques. As the quality of PMUs data has become a fundamental prerequisite for power system analysis, anomaly detection is of critical importance. Among various approaches, Generative Adversarial Networks (GANs)-based methods are attracting widespread attention due to their unsupervised learning capability. However, they inevitably suffer from the problem of an excessive number of trainable parameters. To address this issue, we introduce for the first time a hybrid Quantum Generative Adversarial Networks (QGANs) for anomaly detection in PMUs data. Specifically, the generator in classical GANs is constructed using Quantum Neural Networks (QNNs). Benefiting from the unique properties of quantum mechanics, QNNs exhibit stronger expressivity while requiring fewer trainable parameters than their classical counterparts. Simulation results validate the effectiveness of the proposed method.

anomaly detection,PMUs data,Quantum Generative Adversarial Networks,Quantum Neural Networks