Observer-based Detection And Localization Against False Data Injection Attacks In Smart Grid

With the rapid development of modern science and technology,information communication technology is permeating deeply into modern power system.Information exchange between physical power networks and information networks is becoming more frequent.The traditional power system is deeply integrated with modern advanced sensing technology,control method,and computer communication technology to form a smart power grid system.However,because of the openness and complexity in smart grid system,it is easy to become the target of malicious information attacks.Therefore,the security performance of smart grid has been widely concerned.Aiming at the information security of smart grid,this paper considers using observers to detect and locate the attack of false data injection in smart grid system.The main works of this paper are as follows:Firstly,for the security problem of smart grid system with facing false data injection attack,a detection algorithm for false data injection attack is proposed.For a distributed power system,unknown input observers are designed for each region based on the interconnectedness,modeling disturbance,and uncertainty.On this basis,an adaptive threshold is designed by considering the initial error and model uncertainty.The effectiveness of the algorithm is verified by a simulation example.Secondly,an attack detection and localization algorithm based on interval observer is designed to solve the problem of false data injection attack detection against smart grid system.A state space model of smart grid is established and the limitations of the traditional static estimation mechanism with facing the false data injection attack are analyzed.Then,in order to avoid the complex calculation of adaptive threshold and shorten the detection time,an interval observer,which effectively reduces the detection time,is designed to detect the false data injection attack by considering system interference and modeling uncertainty.On this basis,the logical location matrix is proposed to accurately locate the sensor which is affected by the attack.Based on logical location matrix and interval observer,an attack detection and localization algorithm against false data injection is proposed.Finally,for the security problem of smart grid system with facing false data injection attack,the nonlinear state space model of smart grid is established to better reflect the real state of the system.For a large-scale smart grid,the method of region division is given.Considerring the existence of interval observer,the transformation matrix is used to guarantee the existence of interval observer gain.The nonlinear interval observer is designed for the transformed system and the convergence condition is given.Using the IEEE 36 bus smart grid system,the effectiveness of the regionalization method and the attack detection method based on the designed interval observer are verified.