Research On Security Analysis Of Power Systems Under Malicious Attacks And Defense Method

The safe and stable operation of power systems is an important foundation for human life.It is also an important guarantee for the national economy and people’s well-being.However,with the increasing penetration of renewable energy sources in power systems such as wind power,its inherent uncertainty and nonlinear variations,and other characteristics pose serious risks to the safe and stable operation of the power system.In addition,the deep coupling of the cyber and physical layers of power systems makes the power system inevitably subject to cyberattacks,such as malicious data attacks.As a new type of cyberattack,the malicious data attack can modify the results of state estimation by tampering with state estimation measurement data,and it can bypass anomaly data detection.Malicious data attacks can induce the control center to misjudge the operating status of the power system and thus make incorrect decisions to cause the power system to disconnect or even cascading failure.Moreover,with the continuous development of human society,the frequency of malicious physical attacks,including extreme natural disasters,war,and man-made attacks,is increasing.Due to the complex structure of the power network,the power transmission lines cover a wide area and are directly exposed to the outside world,and the frequency and degree of damage to power transmission lines are becoming more and more serious.Therefore,there is an urgent need to consider the impact of renewable energy uncertainty and information physical network attacks,and other potential risks on the power system.In this thesis,an in-depth study is conducted on the security analysis and defense methods of power systems under malicious attacks,of which the main work is as follows:(1)A malicious data attack risk assessment model for power systems is proposed for malicious data attack risk.The KKT algorithm is used to solve this model to obtain the optimal global solution,which effectively evaluates the impact of malicious data attack on power system transmission line flows and provides guidance for power system security personnel to develop defense strategies.Then,a high stealth malicious data attack mechanism is proposed for the anomalous characteristic information of malicious attack data.Based on the high stealth malicious data attack mechanism,the attacker can construct an attack vector to hide the tampered measurement data in the normal data to avoid detection.At the same time,the attacker can optimally select multiple target lines for line overload attacks,effectively quantifying the impact of high stealth malicious data attacks on the secure and stable operation of power systems.(2)An intra-interval cybersecurity preventive dispatch strategy model is proposed for the risk of intra-interval malicious data attacks.First,the risk of line overload in the dispatch interval of power systems caused by nonlinear variations in wind power output is evaluated for the problem of short-term nonlinear variations in wind power output in the dispatch interval;then,considering the risk of malicious data attacks on power systems,the risk of attackers using nonlinear fluctuations in wind power output during the intrainterval to aggravate line overload over the dispatch interval is evaluated;finally,the Benders-like algorithm is used to solve the intra-interval cybersecurity preventive dispatch strategy model to obtain the best preventive dispatch strategy.The preventive dispatch strategy based on the s intrainterval cybersecurity preventive dispatch strategy model can effectively eliminate the risk of line overload in the dispatch interval and ensure the safe and stable operation of power systems in the dispatch interval.(3)A two-stage distributionally robust mitigation strategy model considering the probabilistic uncertainty of discrete attack scenarios is proposed for high stealth malicious data attack risk.First,the set of highstealth attack scenarios is obtained based on the high-stealth malicious data attack mechanism;then,the uncertainty set of the probability distribution of the high-stealth malicious data attack scenarios is constructed by taking the 1parameter and infinity paradigms as constraints;finally,based on the uncertainty set of the probability distribution of the attack scenarios,the robust mitigation strategy model of the high-stealth malicious data attack is established.An effective solution algorithm is proposed to obtain the global optimal solution.The strategy determined based on this model can not only eliminate the line overload issue caused by the high stealth malicious data attack but also prevent the attacker from constructing a feasible high stealth attack scenario again.(4)A novel malicious physical attack risk assessment model for power systems is proposed for the risk of malicious physical attacks.This model can effectively assess the combined impact of malicious physical attacks and wind power output uncertainty on the power system.To enhance the resilience of the power system to mitigate the combined impact of malicious physical attacks and wind power output uncertainty,an adaptive robust optimization model considering malicious physical attacks and oriented to the uncertainty of wind power output is proposed;and,the proposed model is solved by C&CG(Column and Constraint Generation)algorithm to obtain the global optimal solution.The line protection strategy based on this model can effectively reduce the operating cost of the power system and is of great significance to improve ability of the power system to deal with malicious physical attacks.(5)A defense model based on uncertainty of malicious physical attack resources is proposed to improve the defense capability of power system for the problem of malicious physical attack.However,attackers launch malicious physical attacks with continuity,stage,and periodicity.Therefore,the linehardening strategy formulated by a single time-based defense model may not be able to hedge against malicious physical attacks to the extent that the secure and stable operation of the power system is threatened.For the multiperiod problem of malicious physical attacks,a power system defense model of multi-period uncertain malicious physical attacks is proposed.In order to develop a more practical defense strategy,the model also considers the optimal allocation of attack resources in different attack stages.Finally,this proposed model is solved iteratively using the C&CG(Column and Constraint Generation)algorithm to determine the best line hardening strategy to ensure the secure and stable operation of the power system.