Modern power systems are vulnerable to false data injection attacks (FDIAs) due to the widespread applications of two-way communication networks for system operation and control. Diagnosing such malicious attacks are of great significance for the resilient operations of power systems. Yet, in literature, the majority of studies in FDIA detection focus on power transmission systems. In this paper, a novel FDIA detection scheme is proposed for three-phase distribution systems based on zero-sequence voltage (ZSV). From the voltage and power measurements, the bus voltages are estimated, and then the estimated ZSV is calculated as the sum of the estimated bus voltages on the three phases to represent the degree of unbalance of the distribution system. Via mathematical analysis of the linear distribution system state estimation (DSSE) model, the distribution of the estimated ZSV under the normal condition is derived, based on which a whitening process is adopted on the estimated ZSV to weaken the effect of measurement noises. The $\mathcal {L}_{2}$ -norm of the whitened ZSV vector is then compared with a predefined threshold for FDIA detection. Moreover, the probability of false alarm of the proposed scheme is derived, which can be utilized to determine the detection threshold for a desired tolerance of false alarm rate. The proposed scheme is validated on several IEEE Test Feeders and simulation results show the effectiveness of the proposed scheme in detecting FDIAs in three-phase distribution systems.

中文翻译：

---

一种新的基于 ZSV 的多相配电系统 FDIA 检测方案

由于用于系统操作和控制的双向通信网络的广泛应用，现代电力系统容易受到虚假数据注入攻击（FDIA）。诊断此类恶意攻击对于电力系统的弹性运行具有重要意义。然而，在文献中，大多数 FDIA 检测研究都集中在电力传输系统上。在本文中，针对基于零序电压（ZSV）的三相配电系统提出了一种新的 FDIA 检测方案。根据电压和功率测量值，估算母线电压，然后将估算的 ZSV 计算为三相上估算的母线电压之和，以表示配电系统的不平衡程度。通过线性配电系统状态估计（DSSE）模型的数学分析，推导出正常情况下ZSV估计值的分布，在此基础上对ZSV估计值进行白化处理，减弱测量噪声的影响。这 $\数学{L}_{2}$ 然后将白化 ZSV 向量的范数与 FDIA 检测的预定义阈值进行比较。此外，推导了所提出方案的虚警概率，可用于确定虚警率期望容差的检测阈值。所提出的方案在几个 IEEE 测试馈线上得到验证，仿真结果表明所提出的方案在检测三相配电系统中的 FDIA 方面的有效性。