Resilience Enhancement Methods For Cyber-Physical Power Systems Against Intentional Attacks

Traditional N-1 or N-2 security criterion cannot assist power systems withstand the high impact and low probability(HILP)events such as intentional attacks and natural disasters.Efforts must be made to improve the resilience of power systems under such events.Moreover,the deep coupling between the primary system and the information system makes the disruption in either side likely to impact the normal operation of the other side,and even cause faults to spread alternately between the two sides,thus aggravating the security situation of the cyber-physical power system(CPPS)under extreme events.Against this background,this research focuses on the resilience enhancement methods for CPPS in intentional attack scenarios.Firstly,the defense methods against physical attacks and false data injection attacks(FDIAs)are proposed from the perspective of pre-event hardening.Then,for the event response,a two-stage preventive-corrective dispatch method considering attack uncertainty is developed.Finally,a cyber-physical coordinated recovery method under extreme attack scenarios is proposed for the post-event restoration.The main work and results of this research are as follows:(1)Propose a CPPS defense method to counteract physical attacks.Considering attacks on the transmission network and its communication system,the CPPS defense resource allocation problem is constructed as a defender-attacker-dispatcher optimization model with the goal of minimizing load shedding.By analyzing the functional and topological coupling relationships between the power network and the communication network,we formulate the operational constraints of CPPS considering the dual cyber-physical coupling characteristics,and impose these constraints on the underlying dispatcher problem,thereby simulating the impact of physical attacks on CPPS.After decomposing the model into a main problem and a subproblem,the column and constraint generation(C&CG)algorithm is used to solve the optimal strategy.The proposed method improves CPPS resilience from the perspective of pre-event hardening,and overcomes the shortcoming of traditional defense models which ignore the cyberphysical interactions in attack scenarios.(2)Propose a CPPS defense method to mitigate false data injection attacks.Considering the mechanism of FDIA that induces overload-related line outages and load loss by misleading the dispatcher’s decision with tampered measurement data,a quadlevel defense optimization model against this specific type of FDIA is constructed.The four levels correspond to the defender’s decision making,attacker’s decision making,operator’s dispatching based on false data and its dispatching based on real data,respectively.The problem is transformed into a bi-level model using KKT(KarushKuhn-Tucker)conditions and strong duality theorem,and an improved implicit enumeration algorithm is developed to seek the optimal strategy.The proposed method provides an optimal defense resource allocation scheme against the above-mentioned FDIA from the perspective of pre-event hardening,and overcomes the problem that traditional defense models do not consider the attack mechanism of this specific type of FDIA.(3)Propose a two-stage preventive-corrective dispatch method for CPPS considering the uncertainty of intentional attacks.Multiple attack scenarios obeying a certain probability distribution are introduced into the two-stage dispatch to take into account the impact of attack resource uncertainty.The preventive stage optimizes the current operation strategy by coordinating pre-and post-attack response measures and reduce the dispatch pressure in the corrective stage.The corrective stage develops shortterm response and long-term response schemes based on the structural and functional status of the power network and communication network after the attack.The entire problem is formulated as a tri-level robust stochastic optimization model to minimize the total system cost,and is solved using the modified C&CG algorithm.The proposed method provides a CPPS dispatch scheme under uncertain attacks from the perspective of event response,overcomes the problem that deterministic dispatching and singlestage dispatching cannot balance the economy and security,and can be applied to scenarios where the power network and communication network are damaged at the same time.(4)Propose a cyber-physical coordinated recovery method for power systems against intentional attacks.Firstly,an intentional attack model considering the cyberphysical coupling characteristics is established to provide a systematic attack scenario generation method.For the whole process of system response and restoration after the attack,a cyber-physical coordinated recovery model is established.With the goal of minimizing the energy not supplied,the proposed model utilizes substation reconfiguration to mitigate attacks,and considers the impacts of communication failure on generation adjustment and transmission line switching,and establishes the coordinated recovery constraints for the power network and communication network.The proposed method overcomes the traditional recovery models’ weaknesses of insufficient coordination of system response and recovery schemes,and insufficient consideration of the coordinated repair of damaged power and communication components.It can effectively speed up the power supply recovery and reduce power losses during the recovery phase.