Abstract

False data injection attacks (FDIAs) have recently become a major threat to smart grids. Most of the existing FDIA detection methods have focused on modeling the temporal relationship of time-series measurement data but have paid less attention to the spatial relationship between bus/line measurement data and have failed to consider the relationship between subgrids. To address these issues, in this article, we propose a subgrid-oriented microservice framework by integrating a well-designed spatial–temporal neural network for FDIA detection in ac-model power systems. First, a well-designed neural network is developed to model the spatial–temporal relationship of bus/line measurements for subgrids. A microservice-based supervising network is then proposed for integrating the representation features obtained from subgrids for the collaborative detection of FDIAs. To evaluate the proposed framework, three types of FDIA datasets are generated based on a public benchmark power grid. Case studies on the FDIA datasets show that our method outperforms state-of-the-art methods for FDIA detection in these datasets.

Highlights

  • I N recent years, false data injection attacks (FDIAs) have drawn the attention of researchers to the vulnerability of cyber-physical smart grids [1]–[3]

  • The commonly used metrics, precision (P rec), recall (Rec) and F-1 score (F1), are utilized to evaluate the performance of the FDIA detection [11], [13], [14], which is formulated as follows: P rec = T P/(T P + F P ), Rec = T P/(T P + F N ), F1 = 2 × (P re × Rec)/(P re + Rec), where T P denotes the number of true FDIA samples predicted as FDIA data, F P denotes the number of normal samples predicted as FDIA data, T N denotes the number of true normal samples predicted as normal data, and F N denotes the number of FDIA samples predicted as normal data

  • We proposed an efficient sub-grid-oriented privacy-preserving microservice framework based on deep neural networks for FDIA detection in smart grids

Read more

Summary

INTRODUCTION

I N recent years, false data injection attacks (FDIAs) have drawn the attention of researchers to the vulnerability of cyber-physical smart grids [1]–[3]. The second-hand state values estimated from the measurement data may suffer from data noise These machine learning based methods have achieved some success in detecting FDIAs in AC-model systems, the accuracy can be further improved. The privacy of their local system data need to be protected To address these issues, we propose a sub-grid-oriented privacy-preserving microservice framework integrating a well-designed spatial-temporal neural network for FDIA detection in AC power systems. 2) Compared with most existing methods which focus on temporal relationship between the measurement data, we propose a novel spatial-temporal neural network to learn a sub-grid-level representative feature to represent the spatial-temporal relationship between time-series bus/line measurement data.

RELATED WORK
AC State Estimation
Residual-based Bad Data Detection
PROPOSED FRAMEWORK
False Data Injection Attack
Sub-grids and Measurement Data
MSsuTb
Msup: Supervising Microservice Architecture
Loss Function
Summary
Dataset for FDIA detection
Deployment of Proposed Framework
Evaluation Metrics
Performance of FDIA Detection
Methods
Advantages of Proposed Framework
Limitation and Future Research Direction
Findings
CONCLUSION
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call