Security defense method for dc microgrid against false data injection attack

Abstract

The application discloses a DC micro-grid security defense method against false data injection attack, and comprises the following steps: establishing an integrated information physical DC micro-grid model, designing an attack observer based on unknown input observation, and designing a variable gain flexible quadratic controller. By fully utilizing physical information, the FDIA attack values at different positions are quickly and accurately estimated. The estimated attack can be used as the basis for positioning attack, and the estimated attack can also be used to make the DC micro-grid have robustness to FDIA through a feedback compensation method. By adjusting the control gain, the controller can reduce the influence of FDIA in different communication links. The controller can be applied to most DC micro-grid FDIA scenarios, including the cases that communication links, sensors and controllers are subjected to FDIA, and various collaborative attack scenarios.

Description

Technical Field

[0001] This invention relates to the technical field of DC microgrids and the safe operation of the system, and in particular to a DC microgrid security defense method against spoofed data injection attacks. Background Technology

[0002] With the large-scale integration of new energy sources, the power system is gradually transforming into a new type of power system dominated by new energy sources, and the operating characteristics of the power grid have changed significantly. Among them, DC microgrids are widely used due to their advantages such as high efficiency and easy integration of distributed energy resources. However, the distributed control system of DC microgrids is highly dependent on communication networks and faces cybersecurity threats such as False Data Injection Attacks (FDIA). Traditional defense methods are usually designed based only on controller parameters and cannot effectively deal with the combined effects of communication link attacks and local attacks (such as sensor and controller attacks), and lack deep coupling analysis of physical layer dynamics. In existing technologies, trust-based collaborative control schemes can isolate attacked nodes, but cannot completely eliminate the impact of attacks, and the system stability is difficult to guarantee under complex multi-target attacks.

[0003] For DC microgrids, control systems typically employ a hierarchical structure, including primary droop control to maintain voltage stability and power balance, secondary control responsible for voltage regulation, especially when voltage deviations occur after primary droop control. Third-level control handles economic dispatch. Distributed secondary cooperative control has been widely studied due to its ease of expansion, low communication burden, and support for islanded operation. However, distributed control requires information from neighboring nodes to generate the control volume through a consensus algorithm-based controller. This method is highly dependent on communication links and vulnerable to network attacks, particularly spoofing attacks, leading to operational instability. Based on the attack location, spoofing attacks can be categorized into two types: communication link attacks and local attacks. Local attacks include injecting spoofed data into sensors or hijacking the entire controller.

[0004] Currently, proactive defense technologies for network attack security include data-driven and model-driven approaches. Data-driven approaches use machine learning, reinforcement learning, and other techniques to analyze large amounts of abnormal and normal data, extract features, and restore abnormal data to values ​​under normal operating conditions. Model-driven approaches use various types of observers to estimate the state of DC microgrid models and feed back attack values ​​to mitigate the impact of attacks. However, current defense methods are not effective against network attacks on communication links of distributed DC microgrids, and the accuracy of the models is not high, leading to potential biases in the state estimation process. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings and deficiencies of the prior art and provide a security defense method for DC microgrids against spoofed data injection attacks. This method can effectively address FDIA problems in different locations and is applicable to DC microgrids.

[0006] To achieve the above objectives, the technical solution provided by this invention is: a security defense method for DC microgrids against spoofed data injection attacks, comprising the following steps:

[0007] S1: Improve the communication link of the power network by adding an auxiliary variable as a new transmission signal in the communication between different power network nodes to represent the difference between the control parameters in the system equation of the cyber-physical microgrid control layer and the impedance model parameters of the physical layer system.

[0008] S2: By using this new transmission signal, the control parameters and impedance model parameters are unified to construct an integrated distributed cyber-physical DC microgrid model, which solves the problem that existing DC microgrid modeling methods only consider the control layer system equations and do not consider the impedance model of the physical layer system.

[0009] S3: Based on the improved cyber-physical DC microgrid modeling method, considering the characteristic that the physical circuit topology is not affected by the attack when it occurs, and combining the disturbance estimation and compensation techniques, the possible false data injection attack is regarded as a disturbance, and an augmented state observer is established to detect and estimate the false data injection attack.

[0010] S4: Based on the augmented state observer's estimate of the false data injection attack, the estimate is fed back into the control layer system equations to calibrate and correct the received information, thereby completing the security defense against false data injection attacks and achieving stable grid operation. Considering the possible false data injection attacks on different communication links, the secondary controller of the DC microgrid is improved by adding elastic control where the gain changes with the corresponding impedance to fully enhance the security performance of the control and achieve comprehensive security protection for the DC microgrid. This secondary controller is applicable to most DC microgrid FDIA scenarios, including situations where communication links, sensors, and controllers are subjected to FDIA, as well as various collaborative attack scenarios.

[0011] Furthermore, in step S1, the newly added auxiliary variable θ i The differential equation for (t) is expressed as:

[0012]

[0013] In the formula, N i Let j be the set of neighboring nodes of node i in the microgrid; let j be the set of neighboring nodes of node i in the microgrid. i Node number in; a ijThe corresponding elements in the adjacency matrix of a DC microgrid represent the communication connectivity between different nodes. These are the average voltage estimation information for nodes i and j of the DC microgrid used for secondary control, respectively. For the newly added auxiliary variable θ i The differential of (t);

[0014] By using auxiliary variables in the establishment of cyber-physical DC microgrid models, the information redundancy of communication links between different nodes can be increased, reducing the risk of critical information being attacked.

[0015] Furthermore, the specific steps of step S2 are as follows:

[0016] S21: Based on the received voltage and current signals and auxiliary variable signals, the secondary controller equation for the voltage of node i in the DC microgrid is expressed as:

[0017]

[0018] In the formula, δ Vi (t) represents the voltage correction value of the secondary controller at microgrid node i; N i Let j be the set of neighboring nodes of node i in the microgrid; let j be the set of neighboring nodes of node i in the microgrid. i Node number in V; ref (t) represents the reference voltage of the microgrid DC bus; a ij These are the corresponding elements in the adjacency matrix of the DC microgrid; These are the average voltage estimation information for nodes i and j of the DC microgrid used in secondary control, respectively; θ i (t) is a newly added auxiliary variable; b i The leader node gain indicates whether the node receives the reference value; if it does, then b. i =1;

[0019] S22: Considering the transmission and reception of newly added auxiliary variables, the differential equation of the physical circuit model of node i in the DC microgrid is expressed as:

[0020]

[0021] In the formula, C i Here are the capacitor parameters in the output filter circuit of microgrid node i; R Li R represents the load of node i; ij The line impedance between different nodes; I i (t), V i (t) represents the current and voltage at node i, respectively; θ represents the derivative of the voltage at node i; j(t) represents the auxiliary variable received from the neighboring node j via the communication link;

[0022] S23: DC microgrid modeling, in addition to secondary controller, physical circuit model, and differential equation modeling, also includes droop control, voltage and current dual closed loop, and inverter modeling. The DC microgrid droop control model can be established as follows:

[0023] V ni (t)=V nom (t)-R di I i (t)

[0024] In the formula, R di V is the droop coefficient of node i; ni (t) is the local reference value generated by the droop control; V nom (t) is the voltage control quantity determined by the secondary control, and its generated equation is:

[0025]

[0026] In the formula, Let be the unit current at nodes i and j, respectively. I i (t), I j (t) represents the currents at nodes i and j, respectively; Let be the rated currents of nodes i and j, respectively; ∫(·)dt is the integral symbol;

[0027] The voltage and current dual closed-loop modeling of a DC microgrid is as follows:

[0028]

[0029] In the formula, w i (t), u i (t) represent the output variables of the voltage loop and the current loop in the voltage-current dual closed loop, respectively; These are the voltage loop and current loop gains, respectively.

[0030] S24: Based on the differential equations of the transmission signal, secondary controller, and physical circuit model, combined with the inverter, droop control, and voltage-current dual closed-loop model, the general form of the differential equations for the cyber-physical DC microgrid is obtained:

[0031]

[0032] y i (t)=C i x i (t)

[0033] In the formula, x i (t)=[V i (t),I i (t),u i (t),w i (t),V nom (t),θ i (t)] T For state variables; Represents the state variable x i The differential of (t); y i (t)=[V i (t),I i (t)] T For output variables; For input variables; A i B i C i This is the system matrix.

[0034] Furthermore, the specific steps of step S3 are as follows:

[0035] S31: Analyze and model spoofing attacks occurring at different locations, analyze the impact mechanism of attacks at different locations on DC microgrids, and implement detection and defense measures for attacks at different locations. Spoofing attacks may occur in the communication links between different nodes, and the secondary controller is affected in this case, with the secondary control voltage correction δ... Vi (t) becomes:

[0036]

[0037] In the formula, N i Let j be the set of neighboring nodes of node i in the microgrid; let j be the set of neighboring nodes of node i in the microgrid. i Node number in; a ij For each element in the adjacency matrix of a DC microgrid, V represents the communication connectivity between different nodes; ref (t) represents the reference voltage of the microgrid DC bus; The average voltage estimation information for DC microgrid node i used to participate in secondary control; This refers to the average voltage estimate of neighboring nodes received by microgrid node i after it has been attacked. δ represents the average voltage estimation information used by node j of the DC microgrid for secondary control. ij (t) represents the value of the spoofed data injection attack received by the communication link; b i The leader node gain indicates whether the node receives the reference value; if it does, then b. i =1;

[0038] When a spoofed data injection attack occurs in the sensor or actuator of node i, the secondary control voltage correction amount δ Vi (t) is represented as:

[0039]

[0040] In the formula, This refers to the local average voltage estimate received by sensor i after it has been attacked; δ si (t), δ ai (t) represents the values ​​of the spoofed data injection attack received by the sensor and actuator, respectively;

[0041] S32: Based on step S31, first consider the communication links between different nodes being attacked by spoofed data injection. The controller topology remains unchanged, and the input signal... Becomes an input variable after being attacked Since the physical link topology remains unchanged, but the input signal changes, the physical link model can be represented as follows:

[0042]

[0043] In the formula, C i R represents the capacitor parameters in the output filter circuit of node i; Li R represents the load of node i; ij The line impedance between different nodes; I i (t), V i (t) represents the current and voltage at node i, respectively; θ represents the derivative of the voltage at node i; j (t) is an auxiliary variable;

[0044] S33: Combining the constructed cyber-physical DC microgrid model, the attack is treated as a state variable of the cyber-physical DC microgrid model, and the augmented cyber-physical DC microgrid differential equation is constructed:

[0045]

[0046] In the formula, O i (t)=[V i (t),I i (t),u i (t),w i (t),V nom (t),θ i (t),δ ij (t)] T To augment the system state variables; To augment the differentials of the system state variables; w i (t), u i(t) represent the output variables of the voltage loop and current loop in the voltage-current dual closed loop, respectively; V nom (t) represents the voltage control quantity determined by the secondary control; θ i (t) is a newly added auxiliary variable; For unknown disturbances; The differential of the attack value is injected into the false data received by the communication link; E i , U is the coefficient matrix of the augmented system; i (t) is the input variable; y i (t) is the output variable;

[0047] An attack observer is constructed based on an augmented system model, utilizing an unknown input observer:

[0048]

[0049] In the formula, z i (t) represents an intermediate variable for the attack observer; The derivative of the intermediate variable of the attack observer; To augment the estimates of the system state variables; F i T i H i K i Let the coefficient matrix of the attack observer satisfy the following conditions:

[0050]

[0051] In the formula, K i1 K i2 I is an intermediate variable; considering that the sensor is subjected to a false data injection attack, the physical link model is changed to use V. ai (t) represents:

[0052]

[0053] In the formula, V ai (t) represents the local voltage measurement value after the attack; δ si (t) represents the value of the local attack received; due to the attack on the executor, δ ai It will not affect the value of the received input variable. It can be estimated using conventional observation methods without additional processing. Combining the above formula, we can obtain the observer that estimates the attack on the sensor and actuator. Compared with the case of the communication link being attacked, the observer design is the same, but the system matrix parameters are different.

[0054] Furthermore, in step S3, through disturbance estimation and feedback technology, the estimated value is fed back to the control layer system equations to ensure that each node of the DC microgrid receives accurate data, thereby achieving stable operation. Moreover, only one communication link of node i needs to be configured with an observer. When an attack occurs, this observer can output an attack estimate. However, the estimated value is not the same when different communication channels are attacked; it changes with the corresponding line impedance. The corresponding controller gain is modified accordingly, and adjustments are made based on the impedance to achieve comprehensive defense. The modified secondary controller equations are as follows:

[0055]

[0056] In the formula, N i Let N be the set of neighboring nodes of node i in the microgrid; j and k are sets N. i Node number in; a ij a ik For each element in the adjacency matrix of a DC microgrid, V represents the communication connectivity between different nodes; ref (t) represents the reference voltage of the microgrid DC bus; δ Vi (t) represents the secondary control voltage correction amount; The average voltage estimate for node k in a microgrid without an observer; For the compensated voltage signal; These are the average voltage estimation information for nodes i and j of the DC microgrid used for secondary control, respectively. These are the estimated values ​​for communication link attacks, sensor attacks, and actuator attacks, respectively; K ij K ik The gain values ​​of the controller for different nodes satisfy K ij / K ik =R ij / R ik ;

[0057] With this controller, the output of each node of the DC microgrid will not be affected and can still maintain stable operation in the absence of an attack. However, after an attack, the attack observer will play a role in generating corresponding estimates and feeding them back into the control layer system equations to ensure that the DC microgrid is protected from damage by the attack.

[0058] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0059] 1. This invention introduces a new transmission signal into the distributed communication system of cyber-physical DC microgrids, unifies the control parameters in the control layer system equations with the impedance model parameters of the physical layer system, constructs a distributed cyber-physical DC microgrid model, better utilizes physical layer information, and enables the model to more accurately represent the system operating state.

[0060] 2. This invention designs a cyber-physical distributed attack observer capable of estimating communication link attacks and local attacks, enabling rapid and accurate attack estimation. This invention can accurately estimate spoofing attacks by leveraging the physical topology's immunity to network attacks, even when input variables in the communication link are under attack.

[0061] 3. This invention defends against FDIA by feeding back the attack signal estimated by the designed observer into the control layer system equations. By adjusting the gain of the secondary controller, even when only one communication link is configured with an attack observer, the effects of attacks on different communication links can be eliminated, reducing design costs and achieving comprehensive defense. This controller is suitable for most DC microgrid FDIA scenarios, including situations where communication links, sensors, and controllers are subjected to FDIA, as well as various coordinated attack scenarios. Attached Figure Description

[0062] Figure 1 This is a design illustration of a DC microgrid security defense method against false data injection attacks in an embodiment of the present invention; in the figure, m and n are the node numbers of the set of adjacent nodes belonging to microgrid node i; Here are the average voltage estimates for nodes i, m, and n in the microgrid; These are the estimated values ​​for communication link attacks, sensor attacks, and actuator attacks, respectively; K im K in ∫ represents the gain value of the controller for different nodes; ∫ represents the integral sign; PWM represents pulse width modulation; DC / DC represents a DC-DC converter; FDIA stands for False Data Injection Attack. Detailed Implementation

[0063] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0064] like Figure 1 As shown in the figure, this embodiment discloses a security defense method for DC microgrids against spoofed data injection attacks, the specific details of which are as follows:

[0065] S1: Improve the communication link of the power network by adding an auxiliary variable as a new transmission signal in the communication between different power network nodes to represent the difference between the control parameters in the system equation of the cyber-physical microgrid control layer and the impedance model parameters of the physical layer system.

[0066] The newly added auxiliary variable θ i The differential equation for (t) is expressed as:

[0067]

[0068] In the formula, N i Let j be the set of neighboring nodes of node i in the microgrid; let j be the set of neighboring nodes of node i in the microgrid. i Node number in; a ij The corresponding elements in the adjacency matrix of a DC microgrid represent the communication connectivity between different nodes. These are the average voltage estimation information for nodes i and j of the DC microgrid used for secondary control, respectively. For the newly added auxiliary variable θ i The differential of (t);

[0069] By using auxiliary variables in the establishment of cyber-physical DC microgrid models, the information redundancy of communication links between different nodes can be increased, reducing the risk of critical information being attacked.

[0070] S2: Utilizing this new transmission signal, the control parameters and impedance model parameters are unified to construct an integrated distributed cyber-physical DC microgrid model. This solves the problem that existing DC microgrid modeling methods only consider the control layer system equations and do not consider the impedance model of the physical layer system. The specific operation steps are as follows:

[0071] S21: Based on the received voltage and current signals and auxiliary variable signals, the secondary controller equation for the voltage of node i in the DC microgrid is expressed as:

[0072]

[0073] In the formula, δ Vi (t) represents the voltage correction value of the secondary controller at microgrid node i; N i Let j be the set of neighboring nodes of node i in the microgrid; let j be the set of neighboring nodes of node i in the microgrid. i Node number in V; ref (t) represents the reference voltage of the microgrid DC bus; a ij These are the corresponding elements in the adjacency matrix of the DC microgrid; These are the average voltage estimation information for nodes i and j of the DC microgrid used in secondary control, respectively; θ i (t) is a newly added auxiliary variable; b i The leader node gain indicates whether the node receives the reference value; if it does, then b. i =1;

[0074] S22: Considering the transmission and reception of newly added auxiliary variables, the differential equation of the physical circuit model of node i in the DC microgrid is expressed as:

[0075]

[0076] In the formula, C i Here are the capacitor parameters in the output filter circuit of microgrid node i; R Li R represents the load of node i; ij The line impedance between different nodes; I i (t), V i (t) represents the voltage and current at node i, respectively; θ represents the derivative of the voltage at node i; j (t) represents the auxiliary variable received from the neighboring node j via the communication link;

[0077] S23: DC microgrid modeling, in addition to secondary controller, physical circuit model, and differential equation modeling, also includes droop control, voltage and current dual closed loop, and inverter modeling. The DC microgrid droop control model can be established as follows:

[0078] V ni (t)=V nom (t)-R di I i (t)

[0079] In the formula, R di V is the droop coefficient of node i; ni (t) is the local reference value generated by the droop control; V nom (t) is the voltage control quantity determined by the secondary control, and its generated equation is:

[0080]

[0081] In the formula, Let be the unit current at nodes i and j, respectively. I i (t), I j (t) represents the currents at nodes i and j, respectively; Let be the rated currents of nodes i and j, respectively; ∫(·)dt is the integral symbol;

[0082] The voltage and current dual closed-loop modeling of a DC microgrid is as follows:

[0083]

[0084] In the formula, w i (t), u i (t) represent the output variables of the voltage loop and the current loop in the voltage-current dual closed loop, respectively; These are the voltage loop and current loop gains, respectively.

[0085] S24: Based on the differential equations of the transmission signal, secondary controller, and physical circuit model, combined with the inverter, droop control, and voltage-current dual closed-loop model, the general form of the differential equations for the cyber-physical DC microgrid is obtained:

[0086]

[0087] y i (t)=C i x i (t)

[0088] In the formula, x i (t)=[V i (t),I i (t),u i (t),w i (t),V nom (t),θ i (t)] T For state variables; Represents the state variable x i The differential of (t); y i (t)=[V i (t),I i (t)] T For output variables; For input variables; A i B i C i This is the system matrix.

[0089] S3: Based on the improved cyber-physical DC microgrid modeling method, considering the characteristic that the physical circuit topology is unaffected by the attack, and combining disturbance estimation and compensation techniques, potential spoofing attacks are treated as disturbances. An augmented state observer is established to detect and estimate spoofing attacks. The specific operation steps are as follows:

[0090] S31: Analyze and model spoofing attacks occurring at different locations, analyze the impact mechanism of attacks at different locations on DC microgrids, and implement detection and defense measures for attacks at different locations. Spoofing attacks may occur in the communication links between different nodes, and the secondary controller is affected in this case, with the secondary control voltage correction δ... Vi (t) becomes:

[0091]

[0092] In the formula, N i Let j be the set of neighboring nodes of node i in the microgrid; let j be the set of neighboring nodes of node i in the microgrid. i Node number in; a ijFor each element in the adjacency matrix of a DC microgrid, V represents the communication connectivity between different nodes; ref (t) represents the reference voltage of the microgrid DC bus; The average voltage estimation information for DC microgrid node i used to participate in secondary control; This refers to the average voltage estimate of neighboring nodes received by microgrid node i after it has been attacked. δ represents the average voltage estimation information used by node j of the DC microgrid for secondary control. ij (t) represents the value of the spoofed data injection attack received by the communication link; b i The leader node gain indicates whether the node receives the reference value; if it does, then b. i =1;

[0093] When a spoofed data injection attack occurs in the sensor or actuator of node i, the secondary control voltage correction amount δ Vi (t) is represented as:

[0094]

[0095] In the formula, This refers to the local average voltage estimate received by sensor i after it has been attacked; δ si (t), δ ai (t) represents the values ​​of the spoofed data injection attack received by the sensor and actuator, respectively;

[0096] S32: Based on step S31, first consider the communication links between different nodes being attacked by spoofed data injection. The controller topology remains unchanged, and the input signal... Becomes an input variable after being attacked Since the physical link topology remains unchanged, but the input signal changes, the physical link model can be represented as follows:

[0097]

[0098] In the formula, C i R represents the capacitor parameters in the output filter circuit of node i; Li R represents the load of node i; ij The line impedance between different nodes; I i (t), V i (t) represents the voltage and current signals at node i; δ represents the derivative of the voltage at node i; ij (t) represents the value of the spoofed data injection attack received by the communication link; θ j (t) is an auxiliary variable;

[0099] S33: Combining the constructed cyber-physical DC microgrid model, the attack is treated as a state variable of the cyber-physical DC microgrid model, and the augmented cyber-physical DC microgrid differential equation is constructed:

[0100]

[0101] In the formula, O i (t)=[V i (t),I i (t),u i (t),w i (t),V nom (t),θ i (t),δ ij (t)] T To augment the system state variables; To augment the differentials of the system state variables; w i (t), u i (t) represent the output variables of the voltage loop and current loop in the voltage-current dual closed loop, respectively; V nom (t) represents the voltage control quantity determined by the secondary control; δ ij (t) represents the value of the spoofed data injection attack received by the communication link; θ i (t) is a newly added auxiliary variable; For unknown disturbances; The differential of the attack value is injected into the false data received by the communication link; E i , U is the coefficient matrix of the augmented system; i (t) is the input variable; y i (t) is the output variable;

[0102] An attack observer is constructed based on an augmented system model, utilizing an unknown input observer:

[0103]

[0104] In the formula, z i (t) represents an intermediate variable for the attack observer; The derivative of the intermediate variable of the attack observer; To augment the estimates of the system state variables; F i T i H i K i Let the coefficient matrix of the attack observer satisfy the following conditions:

[0105]

[0106] In the formula, K i1 Ki2 I is an intermediate variable; considering that the sensor is subjected to a false data injection attack, the physical link model is changed to use V. ai (t) represents:

[0107]

[0108] In the formula, V ai (t) represents the local voltage measurement value after the attack; δ si (t) represents the value of the local attack received; due to the attack on the executor, δ ai It will not affect the value of the received input variable. It can be estimated using conventional observation methods without additional processing. Combining the above formula, we can obtain the observer that estimates the attack on the sensor and actuator. Compared with the case of the communication link being attacked, the observer design is the same, but the system matrix parameters are different.

[0109] By employing disturbance estimation and feedback techniques, the estimated values ​​are fed back into the control layer system equations to ensure that each node of the DC microgrid receives accurate data, thereby achieving stable operation. Furthermore, only one communication link of node i needs to be configured with an observer. When an attack occurs, this observer can output an attack estimate. However, the estimated values ​​differ depending on the communication channel being attacked, varying with the corresponding line impedance. This allows for modification of the corresponding controller gain, adjusting it according to the impedance to achieve comprehensive defense. The modified secondary controller equations are as follows:

[0110]

[0111] In the formula, N i Let N be the set of neighboring nodes of node i in the microgrid; j and k are sets N. i Node number in; a ij a ik For each element in the adjacency matrix of a DC microgrid, V represents the communication connectivity between different nodes; ref (t) represents the reference voltage of the microgrid DC bus; δ Vi (t) represents the secondary control voltage correction amount; The average voltage estimate for node k in a microgrid without an observer; For the compensated voltage signal; These are the average voltage estimation information for nodes i and j of the DC microgrid used for secondary control, respectively. These are the estimated values ​​for communication link attacks, sensor attacks, and actuator attacks, respectively; K ij K ik The gain values ​​of the controller for different nodes satisfy K ij / Kik =R ij / R ik ;

[0112] With this controller, the output of each node of the DC microgrid will not be affected and can still maintain stable operation in the absence of an attack. However, after an attack, the attack observer will play a role in generating corresponding estimates and feeding them back into the control layer system equations to ensure that the DC microgrid is protected from damage by the attack.

[0113] S4: Based on the augmented state observer's estimate of the false data injection attack, the estimate is fed back into the control layer system equations to calibrate and correct the received information, thereby completing the security defense against false data injection attacks and achieving stable grid operation. Considering the possible false data injection attacks on different communication links, the secondary controller of the DC microgrid is improved by adding elastic control where the gain changes with the corresponding impedance to fully enhance the security performance of the control and achieve comprehensive security protection for the DC microgrid. This secondary controller is applicable to most DC microgrid FDIA scenarios, including situations where communication links, sensors, and controllers are subjected to FDIA, as well as various collaborative attack scenarios.

[0114] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A security defense method for DC microgrids against spoofed data injection attacks, characterized in that, Includes the following steps: S1: Improve the communication link of the power network by adding an auxiliary variable as a new transmission signal in the communication between different power network nodes to represent the difference between the control parameters in the system equation of the cyber-physical microgrid control layer and the impedance model parameters of the physical layer system. New auxiliary variables The differential equation is expressed as: ； In the formula, Let j be the set of neighboring nodes of node i in the microgrid; j is a set of... Node number in; The corresponding elements in the adjacency matrix of a DC microgrid represent the communication connectivity between different nodes. , These are the average voltage estimation information for nodes i and j of the DC microgrid used for secondary control, respectively. For newly added auxiliary variables The differential; S2: Using this new transmission signal, the control parameters and impedance model parameters are unified to construct an integrated distributed cyber-physical DC microgrid model; S3: Based on the improved cyber-physical DC microgrid modeling method, considering the characteristic that the physical circuit topology is not affected by the attack when it occurs, and combining disturbance estimation and compensation techniques, the possible false data injection attack is regarded as a disturbance, and an augmented state observer is established to detect and estimate the false data injection attack. S4: Based on the augmented state observer's estimate of the false data injection attack, the estimate is fed back into the control layer system equations to calibrate and correct the received information, thereby completing the security defense against the false data injection attack and achieving stable grid operation; and considering the possible false data injection attacks on different communication links, the secondary controller of the DC microgrid is improved by adding elastic control with gain varying with the corresponding impedance to fully improve the security performance of the control and achieve comprehensive security protection for the DC microgrid.

2. The DC microgrid security defense method against spoofed data injection attacks according to claim 1, characterized in that, The specific steps for step S2 are as follows: S21: Based on the received voltage and current signals and auxiliary variable signals, the secondary controller equation for the voltage of node i in the DC microgrid is expressed as: ； In the formula, This is the secondary control voltage correction value for microgrid node i; Let j be the set of neighboring nodes of node i in the microgrid; j is a set of... Node number in; This is the reference voltage for the DC bus of the microgrid. These are the corresponding elements in the adjacency matrix of the DC microgrid; , These are the average voltage estimation information for nodes i and j of the DC microgrid used for secondary control, respectively. To add an auxiliary variable; The leader node gain indicates whether the node receives the reference value; if it does, then... ; S22: Considering the transmission and reception of newly added auxiliary variables, the differential equation of the physical circuit model of node i in the DC microgrid is expressed as: ； In the formula, The capacitor parameters in the output filter circuit of microgrid node i; Let i be the load of node i; The line impedance between different nodes; , Let be the current and voltage at node i, respectively; This represents the derivative of the voltage at node i; This is an auxiliary variable received from the neighboring node j via the communication link; S23: DC microgrid modeling, in addition to secondary controller, physical circuit model, and differential equation modeling, also includes droop control, voltage and current dual closed loop, and inverter modeling. The DC microgrid droop control model can be established as follows: ； In the formula, Let be the droop coefficient of node i; A local reference value is generated for droop control; The voltage control quantity, determined by secondary control, is generated by the following equation: ； In the formula, , Let be the unit current at nodes i and j, respectively. , ; , Let be the currents at nodes i and j, respectively. , These are the rated currents of nodes i and j, respectively; The integral symbol is used. The voltage and current dual closed-loop modeling of a DC microgrid is as follows: ； In the formula, , These are the output variables of the voltage loop and the current loop in the voltage-current dual closed loop, respectively; , These are the voltage loop and current loop gains, respectively. S24: Based on the differential equations of the transmission signal, secondary controller, and physical circuit model, combined with the inverter, droop control, and voltage-current dual closed-loop model, the general form of the differential equations for the cyber-physical DC microgrid is obtained: ； In the formula, For state variables; Represents state variables The differential; For output variables; For input variables; , , This is the system matrix.

3. The DC microgrid security defense method against spoofed data injection attacks according to claim 1, characterized in that, The specific steps for step S3 are as follows: S31: Analyze and model spoofing attacks occurring at different locations, analyze the impact mechanism of attacks at different locations on DC microgrids, and implement detection and defense measures for attacks at different locations. Spoofing attacks may occur in the communication links between different nodes, and the secondary controller is affected in this case, including the secondary control voltage correction. It becomes: ； In the formula, Let j be the set of neighboring nodes of node i in the microgrid; j is a set of... Node number in; The corresponding elements in the adjacency matrix of a DC microgrid represent the communication connectivity between different nodes. This is the reference voltage for the DC bus of the microgrid. The average voltage estimation information for DC microgrid node i used to participate in secondary control; This refers to the average voltage estimate of neighboring nodes received by microgrid node i after it has been attacked. This provides the average voltage estimation information for node j in the DC microgrid to participate in secondary control. Injecting attack values ​​into false data received by the communication link; The leader node gain indicates whether the node receives the reference value; if it does, then... ; When a spoofed data injection attack occurs in the sensor or actuator of node i, the secondary control voltage correction amount... Represented as: ； In the formula, This refers to the local average voltage estimate received by the sensor at node i after it has been attacked. , These are the spoofed data injection attack values ​​that were injected into the sensors and actuators, respectively. S32: Based on step S31, first consider the communication links between different nodes being attacked by spoofed data injection. The controller topology remains unchanged, and the input signal... Becomes an input variable after being attacked Since the physical link topology remains unchanged, but the input signal changes, the physical link model can be represented as follows: ； In the formula, The capacitor parameters in the output filter circuit for node i; Let i be the load of node i; The line impedance between different nodes; , Let be the current and voltage at node i, respectively; This represents the derivative of the voltage at node i; As an auxiliary variable; S33: Combining the constructed cyber-physical DC microgrid model, the attack is treated as a state variable of the cyber-physical DC microgrid model, and the augmented cyber-physical DC microgrid differential equation is constructed: ； In the formula, To augment the system state variables; To augment the differentials of the system state variables; , These are the output variables of the voltage loop and the current loop in the voltage-current dual closed loop, respectively; The voltage control quantity is determined by secondary control; To add an auxiliary variable; For unknown disturbances; The differential of the attack value is injected into the false data received by the communication link; , , , For the coefficient matrix of the augmented system; For input variables; For output variables; An attack observer is constructed based on an augmented system model, utilizing an unknown input observer: ； In the formula, For attack observers; The derivative of the intermediate variable of the attack observer; To augment the estimated values ​​of system state variables; , , , Let the coefficient matrix of the attack observer satisfy the following conditions: ； In the formula, , As an intermediate variable, The identity matrix is ​​used; considering the sensor is susceptible to spoofing attacks, the physical link model is changed to use... express: ； In the formula, The voltage is a local measurement value after the attack. This represents the value of the local attack received; due to the executor being attacked. It will not affect the value of the received input variable. It can be estimated using conventional observation methods without additional processing. Combining the above formula, we can obtain the observer that estimates the attack on the sensor and actuator. Compared with the case of the communication link being attacked, the observer design is the same, but the system matrix parameters are different.

4. The DC microgrid security defense method against spoofed data injection attacks according to claim 1, characterized in that, In step S3, through disturbance estimation and feedback technology, the estimated value is fed back to the control layer system equations to ensure that each node of the DC microgrid receives accurate data, thereby achieving stable operation. Furthermore, only one communication link of node i needs to be configured with an observer. When an attack occurs, this observer can output an attack estimate. However, the estimated value differs depending on the communication channel being attacked, changing with the corresponding line impedance. This allows for modification of the corresponding controller gain, adjusting according to the impedance to achieve comprehensive defense. The modified secondary controller equations are as follows: ； In the formula, Let j and k be the set of neighboring nodes of node i in the microgrid; Node number in; , The corresponding elements in the adjacency matrix of a DC microgrid represent the communication connectivity between different nodes. This is the reference voltage for the DC bus of the microgrid. This is the secondary control voltage correction value; The average voltage estimate for node k in a microgrid without an observer; , For the compensated voltage signal; , These are the average voltage estimation information for nodes i and j of the DC microgrid used for secondary control, respectively. , , These are estimates for communication link attacks, sensor attacks, and actuator attacks, respectively. , The gain values ​​for the controller corresponding to different nodes satisfy... ; With this controller, the output of each node of the DC microgrid will not be affected and can still maintain stable operation in the absence of an attack. However, after an attack, the attack observer will play a role in generating corresponding estimates and feeding them back into the control layer system equations to ensure that the DC microgrid is protected from damage by the attack.

Images