TSCW-GAN Based FDIAs Defense for State-of-Charge Estimation of Battery Energy Storage Systems in Smart Distribution Networks
Zhiying Liu, Yuancheng Li, Qingle Wang, Jingrong Li
Abstract
In smart distribution networks (SDNs), false data injection attacks (FDIAs) on the state of charge (SoC) estimation of battery energy storage systems (BESSs) can successfully escape bad data detection, making SDNs suffer serious security risks. Some valuable work has been done on FDIAs and the detection of FDIAs, but how to defend FDIAs is still tough. Considering the spatio-temporal characteristics of the measurement data in SDNs containing BESSs, a TSCW-GAN defense method is proposed in this article. The method consists of the following two parts: 1) a generator and 2) a discriminator. The generator combines the transformer with the extraction of valid path features to form path features, which are then used to represent the regularity of the data in terms of the expected values. In addition, the discriminator uses conditional signature Wasserstein-1 metrics, thus capturing the conditional joint law of the measurement data. In IEEE13 and IEEE33 bus systems, the experimental results show that the proposed method achieves good performance with 97.5% accuracy.