Research On Resilient Control Strategy For Cyber-Physical Microgrid Under Cyber Attacks

With the large-scale deployment and application of communication network,programmable controllers and power electronic devices,microgrid has experienced an evolution from a single electrical network to a typical cyber physical system(CPS).The isolated operation condition of microgrid has been changed,which shows the new characteristics of openness and interconnection.In the cyber physical environment,microgrid system may be affected by various types of disturbances,among which cyber attacks pose a serious threat to the safe and stable operation of microgrid due to their stealthiness and unpredictability.In order to reduce the adverse effects of cyber attacks and other insecurity factors on system performance,resilient control has become an important research interest of CPS integrated security control framework.At present,the research on the microgrid resilient control strategy under cyber attacks is still in the theoretical exploration stage,and the research results at all levels are not perfect and need to be further improved.Through the research on the typical abnormal scenarios caused by the cyber attacks on the measurement signal,control signal,control decision unit and communication network of microgrid control system,the proposed resilient control strategy is able to effectively guarantee the capability of microgrid to provide auxiliary services and support continuous power supply to the key infrastructure.The research has important theoretical research significance and practical engineering value in promoting the stable operation and the construction of CPS integrated security control system of microgrids.In order to ensure the constant power control of a grid-connected mode microgrid under abnormal measurement signals,the control strategy based on sliding mode observer with variable universe fuzzy control has been proposed,which can estimate the abnormal signals and reconstruct the real current states.The typical model of current transformer faults and cyber attacks is constructed.The sliding mode observer is designed according to the linear matrix inequality method,and variable universe fuzzy control is introduced to dynamically adjust the sliding mode gain,which can realize the reconstruction of the real inverter output current states.The proposed control strategy has the advantages of simple structure,accurate estimation and strong robustness,and eliminates the adverse effects of abnormal measurement signals on the constant control objectives.Simulation results show that the proposed control strategy can effectively estimate the abnormal measurement signals in various fault and cyber attack scenarios.Compared with the sliding mode observer with conventional fuzzy control method,the proposed strategy provides faster and more accurate estimation results.The fast unbiased tracking of distributed generation(DG)output power to the reference signal can be achieved,so as to meet the power transmission requirements of microgrid.In view to solve the frequency and active power control problem of the islanded microgrid under abnormal control signals,the adverse effects of abnormal control signals caused by actuator faults and cyber attacks on the conventional leader following consensus based secondary control strategy of microgrid are quantitatively analyzed.The distributed adaptive sliding mode control strategy considering time delay is proposed.Based on the Artstein transformation theory,the delayed system states are transformed into the delay-free states.According to the transformed system,a distributed sliding mode control method is proposed.This method neither needs to know the prior information of abnormal signals nor assumes that the abnormal signals conform to a certain distribution,and the theoretical analysis proves that the proposed control strategy is able to suppress any forms of abnormal control signal adaptively.Simulation results show that the proposed control strategy can ensure the effective compensation of non-uniform and time-varying delays,and can achieve better performance to resist various types of abnormal control signals compared with existing control methods,so as to ensure that the microgrid frequency can restore to the rated value and that the reasonable active power sharing can be realized according to the DGs' droop coefficients,which improves the frequency stability and delay robustness of the microgrid.Aiming at the typical attack scenario of injecting false data into the control decision unit,the adverse influence of the injected false data on the estimated average information based distributed voltage and reactive power control strategy of islanded microgrid are analyzed,and the necessary and sufficient conditions for the intruder to implement stealthy attack and probing attack are obtained.A trust based resilient control strategy is proposed,which can realize distributed detection of DGs' abnormal behavior,resist the influence of colluding attack on the trust evaluation process,and theoretical analysis proves that the isolation and compensation recovery mechanism can ensure the correct estimation of average voltage and reactive power.The simulation results verify the effectiveness of the proposed control strategy in the scenarios of transient disturbance,continuous attack and colluding attack.Compared with the control strategy without compensation and recovery mechanism.the proposed strategy can eliminate the adverse impact of the false data cumulative effect on the voltage and reactive power control objectives,so as to ensure that the average voltage of normal DG can restore to the rated value and that the reasonable reactive power sharing can be achieved according to the droop coefficients,so as to promote the security and voltage control accuracy of the microgrid system.On account of the problem of increased communication burden due to the traditional periodic communication method,a self-triggered mechanism based distributed voltage and reactive power control strategy is proposed,which is able to ensure the voltage stability of islanded microgrid and achieve reasonable reactive power sharing according to DGs'capacities.Aiming at the typical scenario of communication network interruption caused by the Denial of Service attacks,an ACK response based communication link monitoring mechanism is proposed and new trigger condition is set,which overcomes the drawbacks that the attack frequency should be less than the communication attempts in existing methods.The theoretical analysis demonstrates the convergence of the proposed control strategy under Denial of Service attacks.Simulation results show that the proposed control strategy can effectively reduce the communication burden and achieve need-based data exchange of microgrid control system.Compared with the conventional control strategy without ACK response mechanism,the proposed control strategy shows the robustness to resist high frequency Denial of Service attacks,and the control goal of reactive power sharing according to DGs' capacity in abnormal communication interruption situations can still be achieved.Thus,the reliance of the communication service integrity can be reduced by the proposed control strategy.In order to realize resilient control of microgrid under cyber attacks,the adverse effects of abnormal measurement signal,control signal,control decision unit and communication network on the performance of microgrid control system are systematically analyzed.The sliding mode observer based constant power control method,the distributed adaptive sliding mode based frequency and active power control method,the trust based distributed voltage and reactive power control method,and the self-triggered communication based distributed voltage and reactive power control method are proposed respectively,achieving the goal of safe and stable operation of microgrid under CPS anomalies and cyber attacks.The theoretical framework and implementation method of microgrid resilient control are presented,which provides technical support for the further practical applications.