Research On False Data Injection Attack Detection Method For Load Frequency Control System

Load frequency control(LFC)is an important part of automatic generation control(AGC)and one of the important means of power system operation and control.The main goal is to maintain the regional system frequency and tie-line exchange power at the planned value.In recent years,with the continuous development of network communication technology,the degree of informatization of the load frequency control system has continued to increase,which has also brought many new challenges to the load frequency control system,among which network security problems have become more prominent.To solve these problems,under the sponsorship of National Natural Science Foundation of China and State Grid Corporation of China,this dissertation specifically investigates the detection method of false data injection attacks in the load frequency control system.The main work of the dissertation is as follows:1.Research on the vulnerability of the false data injection attack of the load frequency control system.Firstly,the frequency regulation characteristics of load frequency control system,the mathematical models of different frequency regulation units and the differences between the models are introduced.Then,from the information layer and physical layer of the load frequency control system,the possible network security vulnerabilities are analyzed,and the attack scenarios are designed for different vulnerabilities to investigate the vulnerability of different vulnerabilities.Finally,through the design of network security vulnerability evaluation index,the vulnerability of different weak points is quantified.This part of the work provides the basis for the design of the vulnerable variable detection scheme for the third part.2.Research on the identification of key parameters(frequency deviation coefficient)in the attack detection model.First,according to the historical fault data of the power grid,the approximate daily variation curve of the frequency response characteristic of the power system is fitted,and the frequency deviation coefficient model of the whole day and multiple periods is determined according to the curve variation trend.Then,in order to meet the requirements of the actual power grid for frequency recovery,the determination of the frequency deviation coefficient is connected with the frequency,and a strategy of correcting the frequency deviation coefficient according to the frequency deviation is proposed.And on this basis,the influence of different frequency deviation coefficient deviations on the frequency control effect of the system in different areas and the adjustment of AGC units is investigated.This part of the work provides a model basis for the third part.3.Research on false data injection attack detection method of load frequency control system.First,for the two attack scenarios of the false data injection attack of the exchange power of the area tie-line and the remote tampering attack of the load,the Unknown Input Observer(UIO)and the Extended Disturbance Observer(EDO)are designed respectively to obtain the estimated signal of the attacked variable.Then,according to the attack scenario of area tie-line exchange power,by analyzing the characteristics of the attack model,extracting the slope difference and the numerical difference to design the signal credibility function to realize the detection of the attack.When the attack is detected,the signal of area tie-line exchange power in the attacked area is compared with the estimated signal of UIO to obtain the specific attack type.Aiming at the remote tampering attack scenario of the load,the detection of the attack is realized by comparing the estimated signal of EDO with the historical load data.Finally,the defense strategies of the above two attack scenarios are preliminarily investigated.