False Data Cyber Attack Against Power Systems

With the large-scale application of distributed renewable energy and the rapid development of communication intelligence,modern power systems have developed into a type of cyber-physical power system(CPPS)that is deeply integrated by information networks(communication information systems)and power networks(power systems).CPPS has the ability to observe the states of large power grids,adapt to changing operating conditions and intelligently respond to emergencies,all of which depend on a reliable and secure cyber environment.However,due to the openness and fragility of communication networks,malicious attackers can attack the weaknesses of CPPS through the network,resulting in increasing security risks of power systems.Further,in various attack methods,attacking CPPS by injecting false data has strong concealment and destructiveness.For this reason,this paper studies false data injection attacks(FDIA)toward power systems and power market to lay the foundation for the monitoring center to find its own weaknesses and establish corresponding defense measures.The paper provides the following contributions:Aiming at the dynamic model of power systems,a FDIA scheme toward power systems is proposed from the perspective of the attacker.On account of generators dynamics and static power flows,a descriptor system model for power systems is firstly established.Then,a necessary and sufficient condition for perfect attack is given in terms of unstable eigenvector by the eigen-decomposition of system matrices.Based on this,a FDIA scheme,which can pass not only the observer detection but also the measurement limit detection,is presented to destabilize the power systems.The proposed scheme has the advantages of concealment and implementation simplicity when compared to the existing static attack counterpart.In addition,the research results in this paper find that the sensors in the direction of the above-mentioned unstable eigenvectors are weakness of power systems,and corresponding security protection measures are proposed based on the research results.Aiming at the power market,a FDIA scheme is proposed in the framework of nonlinear state estimation.The definition of a stealthy attack is given by comparing the residuals before and after the attack,and the profit formula is described through the virtual bidding,based on which a profitable attack process is proposed.Then,the Monte Carlo method is used to analyze the conditions of the profit attack,and the attack strategy is described as a kind of nonlinear programming problem.In addition,the research results of this paper find that the node price with the state estimation result is the weakness of the power market and develop the corresponding defense strategy.Furthermore,a FDIA scheme for power market based on topology tampering is proposed.By analyzing the consistency of the sensor measurement and the topology calculated by the network topology processor(NTP)after the attack and comparing the residuals before and after the attack,the definitions of stealthy attack are given which means the attack can avoid not only the topology error processing(TEP)and also residual detection.Then based on the definitions of stealthy attack and the column space of attack vector,an attack scheme is developed.By solving a convex programming problem,the generator can obtain the maximum revenue while ensuring the concealment.Compared with the existing FDIA attack scheme,the attack scheme combining topology tampering and FDIA proposed in this paper is more concealed and more profit.Finally,the validity and superiority of the above three schemes are verified through standard IEEE 9-bus benchmark.