Resarch On Attack Detection Of Dynamic Load Altering Attack In Cyber Physical Power Systems

With the advent of the Internet age,information and data have become two factors that have a huge impact on human beings in the new era.In order to better study the impact of informatization on industrial production,the concept of cyber-physical system（CPS）is proposed.With the deepening of research,CPS has been combined with various industries to form different CPS systems.They all have CPS structures,but they have different characteristics.The cyber physical power system（CPPS）is a combination of CPS and traditional power production.An important research field formed by the combination of technologies,the research and application of CPPS has brought great convenience to human life.Smart grid is an application form of CPPS,which embodies all the characteristics of CPPS to a certain extent.Smart grid is the result of the high coupling of information and physical equipment.All attacks that may occur in the information network will occur in the smart grid.The smart grid itself has a certain degree of anti-interference ability,but the network attacks are diversified.Any link may be attacked,and its safe operation affects all aspects of human beings.In order to ensure its security,it is necessary to detect possible attacks on the smart grid.This paper takes the attack detection method in CPPS as the research direction,and studies the detection method of a new form of CPPS attack-dynamic load altering attack（D-LAA）.The main research content of this paper mainly includes the following three parts:First,the attack detection method based on adaptive Kalman filter（AKF）is studied.Considering the smart grid under D-LAA attack,the AKF is applied to the power system state estimation with noise.Based on AKF,a distance ratio detection algorithm is proposed to detect D-LAA,and the attack is detected by comparing the preset threshold with the distance ratio.Then,the feasibility of the proposed attack detection method is verified by simulation.Secondly,the attack detection and localization research based on maximum correntropy robust two-stage Kalman kilter（MCRTSKF）.Aiming at the power system with non-Gaussian noise attacked by closed-loop D-LAA,an attack detection and localization method based on MCRTSKF is proposed for the problem of D-LAA attack detection and localization.Combining the two-stage Kalman filter algorithm（RTSKF）with the principle of maximum entropy Kalman filter,the MCRTSKF algorithm is designed,and the detection and localization of D-LAA is realized based on MCRTSKF.The effectiveness of the attack detection method under different attack situations is verified by digital simulation.Finally,the attack detection method based on Kalman/H_∞hybrid filter is studied.Aiming at the attack detection problem of D-LAA in smart grid with Gaussian noise,a new power system model is established,and a Kalman/H_∞hybrid filter with robust performance is designed for the model.Based on the designed filtering algorithm,a detection method of D-LAA is proposed.Then,the feasibility of the proposed attack detection algorithm is verified by digital simulation experiments.