Detection And Defense Strategy Of False Data Injection Attack In Isolated Hybrid Microgrids Based On Data-driven

In recent years,hybrid AC/DC microgrids have received more attention because of high operational efficiency and wide application scenarios.Compared with the traditional centralized secondary control,the distributed secondary control only needs to exchange information of neighboring nodes,and the centerless controller design reduces the impact of single-point communication failure,which has a wide application scenario.The AC-DC hybrid microgrid based on distributed control is a typical cyber-physical system,and attackers deviate the microgrid from operation target by false data injection attack at the information layer,thus affecting the power supply quality.This thesis focuses on the problem of fast detection and defense of false data in hybrid AC-DC microgrid under distributed control,based on the analysis ideas and research status of AC microgrid scenarios,with special attention to DC microgrid and AC-DC interconnection unified control scenarios.The corresponding false data action paths are first analyzed and the adverse effects of false data are modeled,and then data-driven fast false data detection and defense strategies are proposed and carried out for experimental validation.The specific innovative work is as follows.(1)Aiming at the sub-microgrid scenario,a fast false data detection and defense strategy based on data-driven is proposed.Through off-line learning and on-line judgment of the transient disturbance process of the microgrid,the algorithm distinguish whether the disturbance comes from load changes or false data,and then reduce the adverse impact of cyber attack by reducing the weight of false data and resilient control,which has the characteristics of low-computation,independent of model,rich types of defense against attacks,and without affecting the dynamics of the original system.The simulation results verify the effectiveness of the proposed strategy.(2)Aiming at the interconnection device of AC/DC hybrid microgrid,a fast detection and defense strategy of false data based on data-driven of interconnection control is proposed based on the local measurement which is difficult to be tampered with.The abnormal influence of false data is detected by the logical regression method of offline learning and online judgment.Then a defense mechanism based on first-order holder is then proposed.Finally,simulation results are given to verify the correctness of the proposed strategy.(3)A cyber-physical platform of AC/DC hybrid microgrid is built.Considering the real cyber attacks,the AC/DC hybrid microgrid is attacked from the point of view of the attackers,and then the detection and defense strategies are further deployed.The experimental results verify the correctness of the derivation of the false data model,and also show that the proposed strategies are able to eliminate the adverse effects of false data,which is of guiding significance to further improve the security of the microgrid network.