Switching type safety control method and device for wind turbine generator set under replay attack

Abstract

The invention provides a switching type safety control method and device for a wind turbine generator set under replay attack. The method comprises the steps of establishing a wind turbine generator dynamical model based on a wind turbine generator operation mechanism and a physical structure; according to the replay attack characteristics, establishing a replay attack model aiming at the measured value of the wind turbine generator rotor speed sensor; setting a lower limit of a detection threshold value, designing a dynamic detector with an updated threshold value, and judging whether a replay attack exists or not; based on the Lyapunov stability theory, a rotor rotating speed state estimator and a switching type safety controller are designed, and normal operation and expected rotor rotating speed output performance of the wind turbine generator under the replay attack are achieved. According to the method, an effective solution is provided for the attack and defense problem of the wind turbine generator set under the nonlinear characteristic, normal operation of the wind turbine generator set can be ensured, malicious influences of replay attacks in the wind turbine generator set are relieved, and the power output performance of the wind turbine generator set is improved.

Description

Technical Field

[0001] The present invention relates to the field of information security of wind turbines, and particularly to a switching-type security control method and device for wind turbines under replay attacks. Background Art

[0002] With the continuous increase in the energy demand of modern society, the overconsumption of fossil fuels and environmental problems such as air pollution caused by them have become increasingly obvious. As the clean energy technology with the largest decarbonization potential per megawatt, wind power has been highly regarded by countries around the world, and using wind energy to generate electricity has become the mainstream method of new energy power generation in various countries.

[0003] With the continuous growth of the installed capacity of wind turbines, wind turbines are gradually developing towards the trend of large-scale and networked, and the communication vulnerability of its control system is gradually exposed and has gradually become the target of malicious hackers. In particular, it is worth noting that the Ethernet-based communication link commonly used in wind turbines is easily exploited by attackers to carry out network attacks such as replay attacks, which may lead to the incorrect transmission of measurement signals.

[0004] In the existing research on the defense of wind turbine information security, network attack scenarios such as false data injection attacks, denial-of-service attacks, and delay attacks are mainly considered. In addition, due to the lack of an anomaly detection mechanism in wind turbines, the implementation of replay attacks has typical characteristics of high impact and strong concealment, which will have an adverse impact on the information transmission and safe operation of wind turbines. Therefore, the research on the impact analysis of replay attacks in wind turbines and corresponding defense measures is of great significance. Summary of the Invention

[0005] The purpose of the present invention is to analyze the communication vulnerability of wind turbines, construct a replay attack model, and design a dynamic state estimation mechanism. This method combines the wind turbine dynamics model and operation data, and predicts sensor data by combining the correlation relationships of different variables to achieve real-time compensation for the abnormal sensor information caused by replay attacks, providing an information reference basis for the design of security control schemes. Furthermore, a flexible security control and defense method is proposed, which can weaken the impact of replay attacks and restore the operating performance of wind turbines without increasing the system's computing and communication burdens.

[0006] The purpose of the present invention is achieved by the following technical solutions:

[0007] According to the first aspect of this specification, a switching-type security control method for wind turbines under replay attacks is provided, and the method includes the following steps:

[0008] S1, based on the operating mechanism and physical structure of the wind turbine, establish a wind turbine dynamics model;

[0009] S2. Establish a replay attack model for the measured values of the rotor speed sensor of the wind turbine according to the replay attack characteristics;

[0010] S3. Set the lower limit of the detection threshold, design a dynamic detector with an updated threshold, and determine whether there is a replay attack;

[0011] S4. Based on the Lyapunov stability theory, design a rotor speed state estimator and a switching-type safety controller to achieve the normal operation of the wind turbine under replay attacks and the expected rotor speed output performance.

[0012] Furthermore, the expression of the wind turbine dynamics model is:

[0013]

[0014] where ω(t) is the function of rotor speed with respect to time t, and J t is the equivalent moment of inertia, and T a (ω(t)) is the function of aerodynamic torque with respect to rotor speed ω(t), K t is the damping coefficient, and T g (t) is the function of generator torque with respect to time t.

[0015] Furthermore, the expression of the replay attack model is:

[0016]

[0017] where y a (t) represents the sensor measurement value output by the replay attack model at time t, ω(t - r) represents the replayed sensor measurement value, and r represents the replayed time length.

[0018] Furthermore, when intercepting and replaying the sensor measurement value data, the attacker needs to ensure that the replayed data satisfies the physical constraints of the wind turbine.

[0019] Furthermore, the expression of the dynamic detector is as follows:

[0020]

[0021] where F = 1 indicates that no attack warning is reported, F = 0 indicates that an attack warning is reported, Δ min represents the lower limit of the detection threshold, the detection error e a (t) = y a (t) - ω d (t), ω d (t) represents the expected reference value of the rotor speed at time t, and the detection error e l (t - T) = y a (t - T) - ωd (t - T), where T is the time length corresponding to a sampling period, and e is the attack detection threshold at time t l (t) and its update rule logic are as follows: When |e a (t)| ≤ |e l (t - T)|, then e l (t) = e a (t), otherwise e l (t) = e l (t - T).

[0022] Furthermore, when the wind turbine communication is normal and no replay attack occurs, the rotor speed ω(t) is used as the output y(t) of the wind turbine control system, and the wind turbine dynamic model is expressed as:

[0023]

[0024] Design the following controller:

[0025]

[0026] where k1 > 0 represents the controller gain, and e(t) = ω(t) - ω d (t) represents the tracking error between the rotor speed ω(t) and its expected reference value ω d (t).

[0027] Furthermore, when the attacker intercepts the rotor speed measurement value and replays it, the information of the replayed rotor speed is received by the controller, and the wind turbine dynamic model is expressed as:

[0028]

[0029] Use a neural network to approximately express T a (ω(t)) as where represents the ideal weight of the neural network, δ(ω(t)) represents the activation function and satisfies δ T (ω(t)0δ(ω(t)) ≤ N, where N is the number of neural network nodes; σ(ω(t)) represents the approximation error of the neural network, and ε is a constant greater than zero;

[0030] Design a rotor speed state estimator based on the neural network, and the expression is:

[0031]

[0032] where represents the estimated value of ω(t), and the estimation error is represents the estimated value of T a (ω(t)), denote is the estimated value and the estimation error is L>0 denotes the state estimator gain, denotes the output of the state estimator.

[0033] The dynamic of the estimation error of the rotor speed is expressed as:

[0034]

[0035] Based on the Lyapunov stability theory, the following controller and adaptive law are designed:

[0036]

[0037] wherein, denotes the estimated value of the ideal weight of the neural network at time t, k2>0 and k3>0 respectively denote the controller gain and the adaptive law gain.

[0038] Furthermore, based on the S3 detection result, a switching Lyapunov function V(t)=(1 - P(t))V1(t)+P(t)V2(t) is designed, where V1(t) and V2(t) respectively denote the Lyapunov functions in the case of no replay attack and with replay attack; if the detector alarms, the controller switching signal is expressed as P(t)=1, and at this time, it is considered that there is a replay attack in the wind turbine; otherwise, P(t)=0, and at this time, it is considered that the communication of the wind turbine is normal and there is no replay attack; the controller is designed according to the two cases of with / without replay attack.

[0039] According to the second aspect of the present specification, a switching type security control device for a wind turbine under replay attack is provided, including a memory and one or more processors, and an executable code is stored in the memory, wherein when the processor executes the executable code, it is used to implement the switching type security control method for a wind turbine under replay attack as described in the first aspect.

[0040] According to the third aspect of the present specification, a computer-readable storage medium is provided, on which a program is stored, wherein when the program is executed by a processor, it implements the switching type security control method for a wind turbine under replay attack as described in the first aspect.

[0041] The present invention has the following beneficial effects: The present invention proposes a method for solving the problems of replay attack modeling and defense for wind turbines, and provides an effective solution for the attack and defense problems of wind turbines under non-linear characteristics. The present invention can ensure the normal operation of wind turbines, alleviate the malicious influence of replay attacks in wind turbines, and improve the power output performance of wind turbines. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0043] Figure 1 is the flowchart of the switching type security control method for a wind turbine under replay attack provided by an embodiment of the present invention;

[0044] Figure 2 is the operating state diagram of a wind turbine without attack provided by an embodiment of the present invention;

[0045] Figure 3 is the operating state diagram of a wind turbine obtained by comparing the security control method provided by the present invention with the traditional feedback control method under the first replay attack (attack replay step size is 100, attack time interval is [3s, 10s]) provided by an embodiment of the present invention;

[0046] Figure 4 is the operating state diagram of a wind turbine obtained by comparing the security control method provided by the present invention with the traditional feedback control method under the second replay attack (attack replay step size is 200, attack time interval is [4s, 16s]) provided by an embodiment of the present invention;

[0047] Figure 5 is the structural diagram of the switching type security control device for a wind turbine under replay attack provided by an embodiment of the present invention. Detailed implementation manners

[0048] To make the above objects, features, and advantages of the present invention more clearly understandable, the following will give a detailed description of the specific implementation manners of the present invention in conjunction with the drawings.

[0049] Many specific details are set forth in the following description to facilitate a thorough understanding of the present invention. However, the present invention may be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0050] The present invention provides a switching type security control method for a wind turbine under replay attack, as Figure 1 shown, including the following steps:

[0051] Step 1: Based on the operating mechanism and physical structure of the wind turbine, establish a dynamic model of the wind turbine:

[0052]

[0053] where ω(t) is a function of the rotor speed with respect to time t, and J t is the equivalent moment of inertia, and T a 9ω(t)) is the aerodynamic torque as a function of the rotor speed ω(t), and K t is the damping coefficient, and T g (t) is the generator torque as a function of time t.

[0054] Step 2: According to the communication framework between different control cabinets of the wind turbine generator set, the communication link between the tower base control cabinet and the nacelle control cabinet is used for the transmission of sensor information. However, the link uses a typical Ethernet communication protocol, which is vulnerable to being parsed and exploited by attackers, thus launching a destructive network attack.

[0055] Based on the dynamics and control logic of the wind turbine generator set, the unit operating in the low-speed section achieves maximum wind energy capture by adjusting the generator torque, and the rotor speed is the key variable for adjusting the generator torque. Therefore, attackers can intercept and replay sensor information through the Ethernet communication link, resulting in the controller receiving incorrect sensor signals and issuing incorrect control commands.

[0056] According to the operating characteristics of the wind turbine generator set, the considered replay attack may increase the load of the unit's transmission chain, reduce the power generation efficiency of the unit, and even affect the operating safety of the wind turbine generator set.

[0057] Through the above analysis, according to the characteristics of the replay attack, a replay attack model for the measured value of the rotor speed sensor of the wind turbine generator set is established, and the expression is:

[0058]

[0059] where y a (t) represents the measured value of the sensor output by the replay attack model at time t, ω(t - r) represents the replayed measured value of the sensor, and r represents the replayed time length. It should be noted that in order to ensure the operating safety of the wind turbine generator set, the value ranges of its various parameter variables have certain physical constraints. For example, the output power generally cannot exceed x% of the rated power, and x is selected according to different models, usually x = 5 or x = 20. Therefore, these constraints will cause certain limitations for attackers when designing attack parameters. To ensure the concealment of the replay attack, attackers usually need to intercept and replay sensor data while ensuring that the data does not exceed the physical constraints of the wind turbine generator set.

[0060] Step 3: In order to effectively determine whether there is a replay attack situation in the system, set the lower limit Δ of the detection threshold min , and design a dynamic detector with an updated threshold:

[0061]

[0062] Among them, F = 1 indicates that no attack warning is reported, F = 0 indicates that an attack warning is reported, and the detection error e at time t a (t) = y a (t) - ω d (t), ω d (t) represents the expected reference value of the rotor speed at time t, and the detection error e at time t - T l (t - T) = y a (t - T) - ω d (t - T), T is the time length corresponding to a sampling period, and the attack detection threshold e at time t l (t) and its update rule logic are: when |e a (t)| ≤ |e l (t - T)|, e l (t) = e a (t), otherwise e l (t) = e l (t - T).

[0063] This dynamic attack detection mechanism based on the updated threshold can achieve real-time dynamic detection of attack variables. The set lower limit of the detection threshold can effectively reduce the false detection rate and missed detection rate, making up for the deficiency that the traditional detection mechanism based on a fixed threshold is difficult to be applied to non-linear wind turbines, and providing important preliminary technical support for switched security control.

[0064] Step 4: Based on the detection results, design the following switched Lyapunov function, using the detection results as the switching index, that is: if the detector alarms, the controller switching signal is expressed as P(t) = 1, and it is considered that there is a replay attack in the wind turbine at this time; otherwise, P(t) = 0, and the wind turbine communication is normal and there is no replay attack at this time.

[0065] In order to achieve the stability analysis of the wind turbine, design the Lyapunov function V(t):

[0066] V(t) = (1 - P(t))V1(t) + P(t)V2(t)

[0067] Among them, V1(t) and V2(t) respectively represent the Lyapunov functions in the case of no replay attack and with replay attack.

[0068] Furthermore, carry out controller design and stability analysis according to the two cases of with / without replay attack:

[0069] ① When the wind turbine communication is normal and there is no replay attack, the controller switching signal P(t) = 0, and the Lyapunov function is Taking the rotor speed ω(t) as the output y(t) of the wind turbine control system, the dynamic model of the wind turbine can be expressed as:

[0070]

[0071] Design the following controller:

[0072]

[0073] where k1 > 0 represents the controller gain, e(t) = ω(t) - ω d (t) represents the tracking error between the rotor speed ω(t) and its expected reference value ω d (t). By taking the derivative of the Lyapunov function and substituting the designed controller into the derivative of the Lyapunov function, we can obtain: where

[0074] ② When the attacker intercepts the rotor speed measurement value and replays this information, the controller receives the replayed rotor speed information. At this time, the dynamic model of the wind turbine can be expressed as:

[0075]

[0076] Considering T a (ω(t)) as a non - linear smooth function, it can be approximately expressed by a neural network with a three - layer architecture of "input - hidden - output" as where, represents the ideal weight of the neural network, (·) T represents the transpose operation, δ(ω(t)) represents the activation function (selected as the hyperbolic tangent function in this embodiment) and satisfies δ T (ω(t))δ(ω(t)) ≤ N, where N is the number of neural network nodes; σ(ω(t)) represents the approximation error of the neural network, and ε is a constant greater than zero.

[0077] To compensate for the replayed rotor speed information of the replay attack, design the following neural - network - based rotor speed state estimator:

[0078]

[0079] where, represents the estimated value of ω(t), and its estimation error is represents the estimated value of T a (ω(t)0, represents the estimated value of and the corresponding estimation error is L > 0 represents the state estimator gain; represents the output of the state estimator.

[0080] Note that, in the case where the system is affected by a replay attack, the above state estimator developed in the present invention, y a (t) = ω(t - r) represents the attacked rotor speed variable. Different from traditional closed-loop estimators, the state estimator proposed in the present invention is based on the difference between the attacked sensor information and the output information of the state estimator for iterative and calibration of the estimator, realizing state estimation in the case where the true output value of the system is unknown.

[0081] Furthermore, the dynamic of the estimation error of the rotor speed can be expressed as:

[0082]

[0083] At this time, the controller switching signal P(t) = 1, and the Lyapunov function Based on the Lyapunov stability theory, the following controller and adaptation law are designed:

[0084]

[0085] where represents the estimated value of the ideal weight of the neural network at time t, and k2 > 0 and k3 > 0 represent the controller gain and the adaptation law gain respectively.

[0086] By taking the derivative of the Lyapunov function and substituting the designed controller into the derivative of the Lyapunov function, we can obtain: where

[0087]

[0088] where represents the upper limit of ω d (t), and ω N represents the upper limit of ω(t).

[0089] Based on the above analysis, for the two cases of the wind turbine with / without replay attack, it can be obtained that the derivative of the Lyapunov function V(t) satisfies

[0090]

[0091] where

[0092]

[0093] Furthermore, the stability analysis result of applying the switched safety controller can be obtained:

[0094]

[0095] where, Am = min{A1, A2}, B M = B2, where t0 represents the start time of the system operation.

[0096] Based on the Lyapunov stability criterion and the above analysis results, it shows that the proposed switched safety controller of the present invention can achieve the stability of the system and the boundedness of all signals.

[0097] The following uses a specific example to illustrate the above-mentioned switched safety control method for wind turbine generators against replay attacks provided by the embodiments of the present invention, specifically as follows:

[0098] Step 1, establish the dynamic model of the wind turbine generator:

[0099] Specifically, Figure 2 Show the simulation operation of the wind turbine generator in the non - attack state. The wind turbine generator can achieve the expected rotor speed output tracking performance during the operation cycle, and the tracking error remains within a very small range.

[0100] Step 2, establish the replay attack model in the wind turbine generator:

[0101] Example: The attacker intercepts the measured data of the rotor speed sensor and launches two replay attacks against the wind turbine generator in the time intervals of [3s, 10s] and [4s, 16s] respectively, and the replay step sizes of the attacks are 100 and 200 respectively.

[0102] Step 3, set the lower limit of the detection threshold, develop a dynamic attack detection mechanism based on the updated threshold, and provide guidance for the design and application of the switched safety controller.

[0103] Step 4, based on the Lyapunov stability theory, design a state estimator and a switched safety controller to ensure the stability of the wind turbine generator under the proposed control strategy, and achieve the normal operation of the wind turbine generator under replay attacks and the expected rotor speed output performance;

[0104] Specifically, the rotor speed state estimator designed by the present invention can estimate the rotor speed data affected by replay attacks, and the developed switched safety controller makes different control strategy adjustments when replay attacks occur / do not occur in the wind turbine generator to ensure the expected control objectives of the wind turbine generator under replay attacks.

[0105] Specifically, Figure 3The figure shows the simulation operation of a wind turbine under a replay attack state (attack parameters: attack time interval [3s, 10s], replay step size 100) when the control scheme proposed in the present invention is applied. For the wind turbine applying the traditional control method, its output performance deteriorates significantly and the tracking error increases significantly within the attack period [3s, 10s]. However, when the control scheme proposed in the present invention is applied, it can effectively mitigate the attack impact and restore the system output performance.

[0106] Figure 4 The figure shows the simulation operation of a wind turbine under a replay attack state (attack parameters: attack time interval [4s, 16s], replay step size 200) when the control scheme proposed in the present invention is applied. The simulation results show that compared with the traditional feedback control method, when the control method proposed in the present invention is applied, the wind turbine can maintain the expected rotor speed output performance during the entire operation period, and the output error remains within a corresponding small range without attack, verifying that the proposed scheme can effectively mitigate the attack impact and restore the system performance.

[0107] These simulation results verify the effectiveness of the proposed switching-type security control method for wind turbines under replay attacks.

[0108] Corresponding to the foregoing embodiments of the switching-type security control method for wind turbines under replay attacks, the present invention also provides embodiments of a switching-type security control device for wind turbines under replay attacks.

[0109] See Figure 5 , the switching-type security control device for wind turbines under replay attacks provided by the embodiments of the present invention includes a memory and one or more processors. Executable code is stored in the memory. When the processor executes the executable code, it is used to implement the switching-type security control method for wind turbines under replay attacks in the above embodiments.

[0110] The embodiments of the switching-type security control device for wind turbines under replay attacks of the present invention can be applied to any device with data processing capabilities. Such a device with data processing capabilities can be a device or apparatus such as a computer. The device embodiments can be implemented by software, or by hardware, or by a combination of software and hardware. Taking software implementation as an example, as a logically meaningful device, it is formed by the processor of any device with data processing capabilities reading the corresponding computer program instructions in the non-volatile memory into the memory for operation. From the hardware level, as Figure 5 shown, it is a hardware structure diagram of any device with data processing capabilities where the switching-type security control device for wind turbines under replay attacks of the present invention is located. Except for Figure 5In addition to the processor, memory, network interface, and non-volatile memory shown, any device with data processing capabilities where the device in the embodiment is located generally may further include other hardware according to the actual functions of the device with data processing capabilities, which will not be elaborated herein.

[0111] For the specific implementation processes of the functions and roles of each unit in the above device, please refer to the implementation processes of the corresponding steps in the above method, which will not be elaborated herein.

[0112] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can be referred to the partial description of the method embodiment. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the present invention. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0113] The embodiment of the present invention further provides a computer-readable storage medium, on which a program is stored. When the program is executed by a processor, it implements the switching type security control method for a wind turbine group facing replay attacks in the above embodiments.

[0114] The computer-readable storage medium may be an internal storage unit of any device with data processing capabilities described in any of the foregoing embodiments, such as a hard disk or memory. The computer-readable storage medium may also be an external storage device of any device with data processing capabilities, such as a plug-in hard disk, a smart media card (SMC), an SD card, a flash card, etc. equipped on the device. Further, the computer-readable storage medium may also include both the internal storage unit and the external storage device of any device with data processing capabilities. The computer-readable storage medium is used to store the computer program and other programs and data required by the device with data processing capabilities, and may also be used to temporarily store the data that has been output or will be output.

[0115] The above are only the preferred embodiments of one or more embodiments of this specification, and are not intended to limit one or more embodiments of this specification. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of one or more embodiments of this specification shall be included within the protection scope of one or more embodiments of this specification.

Claims

1. A switching-type security control method for a wind turbine facing replay attacks, characterized in that It includes the following steps: S1. Based on the operating mechanism and physical structure of the wind turbine, establish a dynamic model of the wind turbine; S2. According to the replay attack characteristics, establish a replay attack model for the measured values of the rotor speed sensor of the wind turbine; S3. Set the lower limit of the detection threshold, design a dynamic detector with an updated threshold, and determine whether there is a replay attack; S4. Based on the Lyapunov stability theory, design a rotor speed state estimator and a switching-type safety controller to achieve the normal operation of the wind turbine under replay attacks and the expected rotor speed output performance.

2. The replay attack-resistant switching-type safety control method for a wind turbine according to claim 1, characterized in that, The expression of the dynamic model of the wind turbine is: where ω(t) is the function of the rotor speed with respect to time t, J t is the equivalent moment of inertia, t a (ω(t)) is the function of the aerodynamic torque with respect to the rotor speed ω(t), K t is the damping coefficient, T g (t) is the function of the generator torque with respect to time t.

3. The replay attack prevention type switching safety control method for a wind turbine according to claim 2, characterized in that The expression of the replay attack model is: where y a (t) represents the sensor measurement value output by the replay attack model at time t, ω(t - r) represents the replayed sensor measurement value, and r represents the length of the replayed time.

4. The replay attack prevention type safety control method for a wind turbine according to claim 1, characterized in that When intercepting and replaying the sensor measurement data, the attacker needs to ensure that the replayed data satisfies the physical constraints of the wind turbine.

5. The replay attack resistant switching type safety control method for a wind turbine according to claim 3, characterized in that The expression of the dynamic detector is as follows: Among them, F = 1 indicates that no attack warning is reported, F = 0 indicates that an attack warning is reported, Δ min represents the lower limit of the detection threshold, and the detection error e at time t a (t) = y a (t) - ω d (t), ω d (t) represents the expected reference value of the rotor speed at time t, and the detection error e at time t - T l (t - T) = y a (t - T) - ω d (t - T), where T is the time length corresponding to one sampling period, and the attack detection threshold e at time t l (t) and its update rule logic are: when |e a (t)| ≤ |e l (t - T)|, then e l (t) = e a (t), otherwise e l (t) = e l (t - T).

6. The replay attack prevention type safety control method for a wind turbine according to claim 3, characterized in that When the wind turbine communication is normal and no replay attack occurs, the rotor speed ω(t) is used as the output y(t) of the wind turbine control system, and the dynamic model of the wind turbine is expressed as: Design the following controller: where k1>0 represents the controller gain, and e(t) = ω(t) - ω d (t) represents the tracking error between the rotor speed ω(t) and its expected reference value ω d (t).

7. The replay attack prevention type safety control method for a wind turbine according to claim 3, characterized in that When the attacker intercepts and replays the rotor speed measurement value, the information of the replayed rotor speed is received by the controller, and the dynamic model of the wind turbine is expressed as: Using a neural network to approximate T a (ω(t)) as where denotes the ideal weight of the neural network, δ(ω(t)) denotes the activation function and satisfies δ T (ω(t)0δ(ω(t)) ≤ N, where N is the number of nodes in the neural network; σ(ω(t)) denotes the approximation error of the neural network, and ε is a constant greater than zero; Design a rotor speed state estimator based on a neural network, and the expression is: Among them, represents the estimated value of ω(t), and the estimation error is represents T a (the estimated value of ω(t)), represents the estimated value of and the estimation error is L > 0 represents the state estimator gain, represents the output of the state estimator. The dynamic representation of the estimation error of the rotor speed is: Based on the Lyapunov stability theory, design the following controller and adaptive law: wherein, represents the ideal weight estimation value of the neural network at time t, and k2>0 and k3>0 respectively represent the controller gain and the adaptive law gain.

8. The replay attack prevention type switching safety control method for a wind turbine according to claim 1, characterized in that Based on the detection result of S3, design a switching-type Lyapunov function V(t) = (1 - P(t))V1(t) + P(t)V2(t), where V1(t) and V2(t) respectively represent the Lyapunov functions in the case of no replay attack and with replay attack; if the detector alarms, the controller switching signal is expressed as P(t) = 1, and at this time, it is considered that there is a replay attack in the wind turbine; otherwise, P(t) = 0, and at this time, it is considered that the wind turbine communication is normal and no replay attack occurs; design the controller according to the two cases of with / without replay attack.

9. A switching-type safety control device for a wind turbine against replay attacks, comprising a memory and one or more processors, wherein executable code is stored in the memory, characterized in that, When the processor executes the executable code, it is used to implement the switching-type safety control method for the wind turbine under replay attack as described in any one of claims 1-8.

10. A computer-readable storage medium having a program stored thereon, characterized in that, When the program is executed by the processor, it implements the switching-type safety control method for the wind turbine under replay attack as described in any one of claims 1-8.

Images