A Generalized Accumulation and Detection Method for False Data Injection Attacks on Smart Grid

Abstract

The invention discloses a generalized accumulation and detection method for false data injection attacks of smart grids. The discrete-time dynamic model of smart grid state estimation is established by using the obtained grid topology, historical system state estimation values ​​and line parameter data, and iterative least squares is adopted. The estimation method is used as an estimation method to estimate the state of the system in real time, and then combined with the estimated value of the power grid at the past time, the state characteristics of the system at the next time are predicted by the fractional calculus theory. An efficient generalized accumulation-sum detection method to detect whether the system is attacked by false data injection in real time. The invention can detect the false data injection attack of the smart grid, has higher detection accuracy and faster detection speed than the prior art, and ensures the safety control and stable operation of the power grid.

Description

technical field [0001] The invention belongs to the field of smart grid information security, and in particular relates to a generalized accumulation and detection method for false data injection attacks on smart grids. Background technique [0002] The smart grid achieves the goal of reliability, safety, economy, and efficiency of the grid through advanced measurement and sensing equipment and advanced control methods. However, in recent years, frequent vicious network attacks at home and abroad have exposed the vulnerability of smart grids, which has sounded the alarm for strengthening the information security construction of smart grids. In particular, fake data injection attacks covertly tamper with the measured values ​​and state variables of the power grid by exploiting the loopholes in the bad detection methods in the power grid, which seriously endangers the safe and stable operation of the power grid. Therefore, how to design effective detection of false data injec...

AI Technical Summary

Problems solved by technology

[0003] At present, the mainstream methods of false data injection attack detection in smart grid mainly include: (1) The detection method based on the principle of residual detection and the comparison with the set threshold, but the detection accuracy is low, the detection efficiency is low, and the false alarm rate is high. Defects; (2) Attack detection methods based on artificial intelligence such as deep learning and reinforcement learning, but the adjustable parameters of the artificial intelligence method itself are relatively complex, the calculation efficiency is low, and a large amount of data is required for training.

None of the existing technologies can fully meet the requirements for fast and accurate detection of false injection attacks on smart grids

[0004] The basic principle of the generalized accumulation and detection method is based on the sequential probability ratio test, using the current and recent process data to test the signal sequence with little change in the mean or variance, and has been relatively successful in network abnormal traffic detection and fault detection applications, but rarely used in smart grid false data injection attack detection

Images