Frequency response service attack detection method, device, storage medium and equipment

Abstract

This invention provides a method, apparatus, storage medium, and device for detecting frequency response service attacks, comprising: acquiring frequency measurement signals and regional control error measurement signals of a power grid control area; inputting the frequency measurement signals into a trained adaptive neural fuzzy inference model to obtain a regional control error prediction signal for the power grid control area; obtaining a root mean square error (RMSE) based on the regional control error measurement signal and the regional control error prediction signal; if the RMS error exceeds a preset threshold, an attack is detected in the power grid control area, and the regional control error is set as the regional control error prediction signal; repeating the above steps until the frequency response control system operation time reaches a preset time threshold. By comparing the regional control error prediction signal and the measurement signal of the power grid control area, it is determined whether an attack exists in the power grid control area, and a response is taken to maintain the frequency stability of the power grid control area.

Description

Technical Field

[0001] This invention relates to the field of power system network security, specifically to methods, apparatus, storage media, and devices for detecting frequency response service attacks. Background Technology

[0002] Under the architecture of the smart grid, the construction of a complete enterprise intranet for the power system is carried out by utilizing the Internet and information and communication technologies, thereby realizing the master-slave transmission, processing and control of various power information services, marking the mature application of the power wide-area information system.

[0003] The frequency of a power system is primarily determined by active power, and the stability control between active power and frequency is called frequency control. Frequency stability and tie-line power stability are crucial indicators of power quality in multi-regional power systems. Any sudden change in load can lead to deviations in the power exchanged between tie-lines and fluctuations in system frequency. Furthermore, potential cyberattacks can significantly impact system stability, potentially even causing system instability. Therefore, determining whether fluctuations in the system are caused by normal load fluctuations or cyberattacks and responding accordingly is key to ensuring the effectiveness of frequency response commands. Summary of the Invention

[0004] Based on this, the present invention provides a frequency response service attack detection method, apparatus, storage medium and device, which determines whether there is an attack in the power grid control area by comparing the area control error prediction signal and the measurement signal of the power grid control area, and responds accordingly, thereby maintaining the frequency stability of the power grid control area.

[0005] In a first aspect, the present invention provides a method for detecting frequency response service attacks, the method being applied to a frequency response control system, comprising:

[0006] Step S1: Obtain the frequency measurement signal and the area control error measurement signal of the power grid control area;

[0007] Step S2: Input the frequency measurement signal into the trained adaptive neural fuzzy inference model to obtain the regional control error prediction signal of the power grid control area;

[0008] Step S3: Obtain the root mean square error based on the measured regional control error signal and the predicted regional control error signal;

[0009] Step S4: If the root mean square error exceeds a preset threshold, then there is an attack in the power grid control area, and the area control error is set as the area control error prediction signal.

[0010] Step S5: Repeat steps S1-S4 until the frequency response control system runs for the time specified in the preset time threshold.

[0011] Furthermore, step S4 also includes:

[0012] If the root mean square error does not exceed the set threshold, then there is no attack behavior in the power grid control area, and the area control error is set as the area control error measurement signal.

[0013] Furthermore, the adaptive neural fuzzy reasoning model includes an input layer, a fuzzification layer, a rule layer, a normalization layer, and a defuzzification layer;

[0014] The specific expression for the input layer is:

[0015] ,

[0016] in, For input variables, For frequency measurement signals in the power grid control area;

[0017] The specific expression for the data blurring layer is:

[0018] ,

[0019] in, For fuzzy sets, The membership degree of the input variable to the fuzzy set. It is a Gaussian membership function;

[0020] The specific expression for the fuzzy rule in the rule layer is:

[0021] Rule i: If yes ,but ,

[0022] in, For output variables, The first parameter to be determined, The second parameter to be determined is the activation weight of each rule, which is expressed as follows: , For the The activation weight of each rule;

[0023] The specific expression for the normalization layer is:

[0024] ,

[0025] in, The normalized rule activation strength, The total number of rules;

[0026] The specific expression for the deblurring layer is:

[0027] .

[0028] Furthermore, the root mean square error obtained from the area control error measurement signal and the area control error prediction signal is specifically expressed as follows:

[0029] ,

[0030] in, The root mean square error, For regional control error prediction signals, For area control error measurement signals, This represents the number of nodes in the power grid control area.

[0031] Furthermore, the training process of the adaptive neural fuzzy reasoning model includes:

[0032] Step S301: Obtain multiple historical frequency measurement signals and multiple historical area control error measurement signals of the power grid control area, and merge each historical frequency measurement signal with the corresponding historical area control error measurement signal into one historical data.

[0033] Step S302: Divide multiple historical data sets into training and testing sets according to a preset ratio;

[0034] Step S303: Input the training set into the adaptive neurofuzzy inference model to adjust the parameters and obtain the preliminary adaptive neurofuzzy inference model;

[0035] Step S304: Input the validation set into the preliminary adaptive neural fuzzy inference model to obtain the region control error prediction signal, and obtain the training error based on the historical region control error measurement signal and the region control error prediction signal;

[0036] Step S305: If the training error satisfies the output condition, the preliminary adaptive neurofuzzy inference model is output as the trained adaptive neurofuzzy inference model.

[0037] Step S306: If the training error does not meet the output condition, return to steps S303-S304 until the training error meets the output condition.

[0038] Furthermore, the expression for the training error is:

[0039] ,

[0040] ,

[0041] in, For training error, For the area control error prediction signal, This is a historical area control error measurement signal. The parameters of the adjusted adaptive neurofuzzy inference model, The parameters of the adaptive neurofuzzy inference model before adjustment. This is the learning rate.

[0042] Furthermore, attacks on the power grid control area include frequency measurement data attacks, regional tie-line power measurement data attacks, and regional ACE signal attacks;

[0043] The specific expression for the frequency measurement data attack is as follows:

[0044] ,

[0045] in, For the Each power grid control area in Frequency measurement data at a given time is used to attack the signal. For the Frequency deviation coefficient of each power grid control area For the Frequency measurement signals for each power grid control area This is the first attack signal. For the Regional capacity coefficient of each power grid control area For the The first power grid control area and the first Power deviation of tie lines between power grid control areas;

[0046] The specific expression for the regional tie-line power measurement data attack is as follows:

[0047] ,

[0048] in, For the The first power grid control area and the first Each power grid control area in Attack signal on regional tie-line power measurement data at a given time. This is the second attack signal. For the Regional capacity coefficient of each power grid control area For the The first power grid control area and the first Power deviation of tie lines between power grid control areas;

[0049] The specific expression for the regional ACE signal attack is as follows:

[0050] ,

[0051] in, For the first Each power grid control area in ACE signal attack signal at any time For the first Each power grid control area in Normal ACE signal at any time This is the third attack signal.

[0052] Secondly, the present invention also provides a frequency response service attack detection device, which is applied to a frequency response control system and includes:

[0053] The signal acquisition module is used to acquire frequency measurement signals and area control error measurement signals of the power grid control area;

[0054] The signal prediction module is used to input the frequency measurement signal into the trained adaptive neural fuzzy inference model to obtain the area control error prediction signal of the power grid control area;

[0055] The root mean square error calculation module is used to obtain the root mean square error based on the area control error measurement signal and the area control error prediction signal.

[0056] An attack detection module is used to set the regional control error as a regional control error prediction signal if the root mean square error exceeds a preset threshold.

[0057] The running time setting module is used to repeat all the above modules until the running time of the frequency response control system reaches the preset time threshold.

[0058] Thirdly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any one of the frequency response service attack detection methods in the first aspect.

[0059] Fourthly, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform any one of the frequency response service attack detection methods in the first aspect.

[0060] The beneficial effects of adopting the above technical solution are as follows: This embodiment obtains the regional control error prediction signal of the power grid control area by using an adaptive neural fuzzy inference model, and calculates the root mean square error of the regional control error prediction signal and the actual measurement signal of the power grid control area. It judges whether there is an attack behavior in the power grid control area by whether the root mean square error exceeds a preset threshold, thereby making a response operation to maintain the frequency stability in the power grid control area, ensuring the accuracy of frequency response services, and improving the frequency stability of the power system. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0062] Figure 1 This is a schematic diagram of a frequency response service attack detection method in one embodiment of this application;

[0063] Figure 2 This is a schematic diagram of a frequency response service attack detection method in one embodiment of this application;

[0064] Figure 3 This is a schematic diagram of a three-region interconnected power grid system divided by the IEEE 14-node system in one embodiment of this application;

[0065] Figure 4 This is a schematic diagram illustrating the frequency control effect of each test method in one embodiment of this application;

[0066] Figure 5 This is a schematic diagram illustrating the tie-line power control effect of various test methods in one embodiment of this application;

[0067] Figure 6 This is a schematic diagram of a frequency response service attack detection device in one embodiment of this application. Detailed Implementation

[0068] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention. To describe the present invention in more detail, the frequency response service attack detection method, apparatus, storage medium, and device provided by the present invention will be specifically described below with reference to the accompanying drawings.

[0069] Unless otherwise defined, the technical or scientific terms used in this application shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms "an," "a," or "the" do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms "comprising" or "including" mean that the preceding element or object encompasses the elements or objects listed after the word and their equivalents, without excluding other elements or objects. The terms "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up," "down," "left," "right," etc., are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0070] The frequency of a power system is primarily determined by active power, and the stability control between active power and frequency is called frequency control. Frequency stability and tie-line power stability are important indicators of power quality in multi-regional power systems. Any sudden change in load can lead to deviations in the power exchanged between tie-lines and fluctuations in system frequency. Furthermore, potential network attacks can significantly impact system stability, and in severe cases, even cause system instability.

[0071] To address this issue, the present invention provides a frequency response service attack detection method. By comparing the predicted and measured regional control error signals of the power grid control area, it determines whether an attack exists in the power grid control area and responds accordingly, thereby maintaining frequency stability in the power grid control area. The method is illustrated using an application to a terminal device as an example, in conjunction with the appendix. Figure 1 The diagram shows a frequency response service attack detection method.

[0072] This application provides an application scenario for the frequency response service attack detection method. This application scenario includes the terminal devices provided in the embodiments. The terminal devices include, but are not limited to, smartphones and computer devices, wherein the computer device can be at least one of desktop computers, portable computers, laptop computers, mainframe computers, tablet computers, etc. Users operate the terminal devices to obtain identification results regarding whether charging data has been subjected to intrusion or tampering. For details, please refer to the embodiments of the charging data intrusion feature detection method.

[0073] The frequency response service attack detection method in this embodiment is applied to a frequency response control system, which includes multiple grid control areas. Each grid control area is equipped with a controller, which controls and adjusts the generator output power within the grid control area by sending control signals. When the grid control area is not under attack, the controller in the Automatic Generation Control (AGC) system sends a normal Area Control Error (ACE) signal to maintain the overall balance of the grid control area. The AGC system can balance the tie-line power deviations between different grid control areas. And cause frequency deviation within the power grid control area The value remains at 0. The automatic generation control system receives tie-line power deviation and frequency deviation of the grid control area from distributed sensors, and then uses the tie-line power deviation as a basis for... Frequency deviation within the power grid control area Calculate the area control error (ACE) and send it to the controllers of each power grid control area.

[0074] The specific expression for the area control error when the power grid control area is not under attack is:

[0075] ,

[0076] in, For the Each power grid control area in Time-based regional control error, For the Frequency deviation coefficient of each power grid control area For the Regional capacity coefficient of each power grid control area For the Frequency measurement signals for each power grid control area This represents the total number of areas controlled by the power grid.

[0077] Based on the received ACE signal, it needs to be converted into a control signal by the controller to adjust the generator power output, thereby reducing the tie-line power deviation in different areas. and frequency deviation within the power grid control area To maintain a safe range, the specific expression for the above control signal is as follows:

[0078] ,

[0079] in, This is a control signal used to adjust the power setpoint of the generator set; The proportional gain coefficient determines the direct impact of the ACE signal on the control output. This is the integral gain coefficient, ensuring that the automatic power generation control system maintains zero steady-state error over a long period of time; The differential gain coefficient helps reduce overshoot and oscillation.

[0080] The frequency response service attack detection method in this embodiment specifically includes the following steps:

[0081] Step S1: Obtain the frequency measurement signal and the area control error measurement signal of the power grid control area.

[0082] Step S2: Input the frequency measurement signal into the trained adaptive neural fuzzy inference model to obtain the regional control error prediction signal of the power grid control area.

[0083] Step S3: Obtain the root mean square error based on the measured regional control error signal and the predicted regional control error signal.

[0084] In step S3 above, the root mean square error is obtained based on the measured signal and the predicted signal of the regional control error. The specific expression is as follows:

[0085] ,

[0086] in, The root mean square error, For regional control error prediction signals, This is the area control error measurement signal. This represents the number of nodes in the power grid control area.

[0087] Step S4: If the root mean square error exceeds a preset threshold, then there is an attack in the power grid control area, and the area control error is set as the area control error prediction signal.

[0088] It should be noted that step S4 also includes:

[0089] If the root mean square error does not exceed the set threshold, then there is no attack behavior in the power grid control area, and the area control error is set as the area control error measurement signal.

[0090] In this embodiment, the threshold used to determine whether an attack has occurred can be set to... .

[0091] Step S5: Repeat steps S1-S4 until the frequency response control system runs for the time specified in the preset time threshold.

[0092] The preset time threshold can be the rated operating time of the frequency response control system.

[0093] The above steps enable the determination of whether the power grid control area is under attack and to take corresponding actions to maintain the frequency stability of the power grid control area.

[0094] The Adaptive Neural Fuzzy Inference Model (ANFIS) used in step S2 above includes an input layer, a fuzzification layer, a rule layer, a normalization layer, and a defuzzification layer. The ANFIS combines the training capabilities of fuzzy logic (FL) with artificial neural networks (ANN), using the "if-then" rules of fuzzy logic and membership functions (MFs) to construct input-output pairs, providing a suitable input-output mapping relationship. In this embodiment, the ANFIS has one input, namely the frequency measurement signal of the power grid control area, and one output, namely the area control error prediction signal of the power grid control area. The specific modules of the ANFIS are as follows:

[0095] (1) The specific expression for the input layer is:

[0096] ,

[0097] in, For input variables, This is a frequency measurement signal for the power grid control area.

[0098] (2) Input variables in the fuzzification layer The fuzzification layer maps input variables to fuzzy sets using membership functions. The membership function maps the input variable to the membership degree of its corresponding fuzzy set. The specific expression for the fuzzification layer is:

[0099] ,

[0100] in, For fuzzy sets, The membership degree of the input variable to the fuzzy set. It is a Gaussian membership function.

[0101] (3) In the rule layer, fuzzy rules are generated based on the membership degree of the input. For a variable with only one input and one output, the specific expression of the fuzzy rule in the rule layer is as follows:

[0102] Rule i: If yes ,but ,

[0103] in, For output variables, The first parameter to be determined, The second parameter to be determined is the activation weight of each rule, which is expressed as follows: , For the The activation weight of each rule.

[0104] (4) In the normalization layer, the activation strength of all rules is normalized so that the sum of the activation strengths of all rules is 1; the specific expression of the normalization layer is:

[0105] ,

[0106] in, The normalized rule activation strength, This represents the total number of rules.

[0107] (5) In the deblurring layer, the output of each rule is weighted and averaged according to the normalized activation intensity to obtain the final output; the specific expression of the deblurring layer is:

[0108] .

[0109] The training process for the above adaptive neural fuzzy reasoning model includes:

[0110] Step S301: Obtain multiple historical frequency measurement signals and multiple historical area control error measurement signals of the power grid control area, and merge each historical frequency measurement signal with the corresponding historical area control error measurement signal into one historical data.

[0111] Step S302: Divide multiple historical data sets into a training set and a test set according to a preset ratio. In this embodiment, the preset ratio can be set to training set:test set = 80%:20%.

[0112] Step S303: Input the training set into the adaptive neurofuzzy inference model to adjust the parameters and obtain the preliminary adaptive neurofuzzy inference model.

[0113] In this embodiment, the historical frequency measurement signal of the training set is used as the input signal, and the regional control error prediction signal is used as the output signal.

[0114] Step S304: Input the validation set into the preliminary adaptive neural fuzzy inference model to obtain the region control error prediction signal, and obtain the training error based on the historical region control error measurement signal and the region control error prediction signal.

[0115] Step S305: If the training error satisfies the output condition, the preliminary adaptive neurofuzzy inference model is output as the trained adaptive neurofuzzy inference model.

[0116] Step S306: If the training error does not meet the output condition, return to steps S303-S304 until the training error meets the output condition.

[0117] in,

[0118] ,

[0119] ,

[0120] in, For training error, For the area control error prediction signal, This is a historical area control error measurement signal. The parameters of the adjusted adaptive neurofuzzy inference model, The parameters of the adaptive neurofuzzy inference model before adjustment. The learning rate. The output condition for the training error in step S306 above is: The minimum value is reached, and the value is continuously updated using gradient descent.

[0121] It should be noted that, in this embodiment, the attack behaviors in the power grid control area include frequency measurement data attacks, regional tie-line power measurement data attacks, and regional ACE signal attacks.

[0122] The specific expression for the frequency measurement data attack is as follows:

[0123] ,

[0124] For the first Each power grid control area in Frequency measurement data at a given time is used to attack the signal. For the first Frequency deviation coefficient of each power grid control area For the first Frequency measurement signals for each power grid control area This is the first attack signal. For the first Regional capacity coefficient of each power grid control area For the first The first power grid control area and the first Power deviation of tie lines between power grid control areas This represents the total number of power grid control areas.

[0125] The specific expression for the area tie-line power measurement data attack is:

[0126] ,

[0127] For the first The first power grid control area and the first Each power grid control area in Attack signal on regional tie-line power measurement data at a given time. This is the second attack signal. For the first Regional capacity coefficient of each power grid control area For the first The first power grid control area and the first Power deviation of tie lines between power grid control areas;

[0128] The specific expression for a regional ACE signal attack is:

[0129] ,

[0130] For the first Each power grid control area in ACE signal attack signal at any time For the first Each power grid control area in Normal ACE signal at any time This is the third attack signal.

[0131] To better illustrate the frequency response service attack detection method of this embodiment, the IEEE 14-node system is used as an example, as shown in the attached diagram. Figure 3 As shown, the IEEE 14-node system is divided into a three-area interconnected power grid system, and the generator and tie-line parameters are shown in Table 1:

[0132]

[0133] Table 1. Parameter information of the three-area interconnected power system of the IEEE 14-bus system.

[0134] From the attacker's perspective, assuming that at t=0s, the frequency signal in region 1 is tampered with by a 10% offset signal, causing a frequency drop, and the attack duration is 3s, the steady-state frequency and tie-line power of all three regions will decrease to a certain extent.

[0135] Based on this, a distributed frequency response strategy based on PID is used, with the ACE error as the input to the controller. The transmission substations for the three regions are nodes 2, 6, and 9, respectively, with the control center located at node 1. Frequency and tie-line power information for the three regions are transmitted to the master station via nodes 2, 6, and 9, respectively.

[0136] Conventional frequency response strategies, due to the use of incorrect frequency measurement signals and ACE calculation results, result in the controlled variable in frequency response services failing to reflect system disturbances in a timely manner, leading to significant overshoot and decreased system stability. The detection method of this invention allows for compensatory control using the inferred ACE results; simulation results are attached. Figure 4 Appendix Figure 5 As shown in Table 2, it can be seen that under the effect of the compensation detection method, the frequency and tie-line power can be rapidly restored, and the overshoot and settling time are greatly reduced, which greatly improves the effect of the control system.

[0137]

[0138] Table 2 Comparison of control effects under different detection methods

[0139] After designing an ANFIS-based estimation and detection system, it is crucial to evaluate the model's effectiveness by comparing the dynamic behavior of frequency response services. For this purpose, a specific region's data is used... and The dataset was used to train the ANFIS unit. The actual data was compared with the training data. After testing ANFIS, the comparison between the actual data and the network data showed that the training process was performed correctly, with an average RMSE of [value missing]. It can maintain stable power frequency and power in most systems.

[0140] It should be understood that, although attached Figure 1 The steps in the flowchart are shown sequentially according to the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise expressly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the accompanying drawings may include multiple sub-steps or sub-stages, which are not necessarily completed at the same time, but may be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but may be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0141] The frequency response service attack detection method is described in detail in the embodiments disclosed above. Since the above-disclosed method can be implemented using various types of devices, this invention also discloses a frequency response service attack detection device. Figure 6 The following are specific embodiments for detailed explanation.

[0142] The signal acquisition module 401 is used to acquire the frequency measurement signal and the area control error measurement signal of the power grid control area;

[0143] Signal prediction module 402 is used to input the frequency measurement signal into the trained adaptive neural fuzzy inference model to obtain the area control error prediction signal of the power grid control area;

[0144] The root mean square error calculation module 403 is used to obtain the root mean square error based on the regional control error measurement signal and the regional control error prediction signal.

[0145] The attack detection module 404 is used to set the regional control error as a regional control error prediction signal if the root mean square error exceeds a preset threshold.

[0146] The running time setting module 405 is used to repeat all the above modules until the running time of the frequency response control system reaches the preset time threshold.

[0147] For details regarding the frequency response service attack detection device, please refer to the above description of the method's limitations; they will not be repeated here. Each module in the aforementioned device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the terminal device's processor in hardware form or independent of it, or stored in the terminal device's memory in software form, so that the processor can call and execute the corresponding operations of each module.

[0148] In one embodiment, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the frequency response service attack detection method described above.

[0149] The computer-readable storage medium may be an electronic storage device such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM (Erasable Programmable Read-Only Memory), hard disk, or ROM. Optionally, the computer-readable storage medium includes a non-transitory computer-readable storage medium. The computer-readable storage medium has storage space for program code that performs any of the method steps described above. This program code can be read from or written to one or more computer program products, and the program code may be compressed in an appropriate form.

[0150] In one embodiment, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the above-described frequency response service attack detection method when executing the computer program.

[0151] The computer device includes a memory, a processor, and one or more computer programs, wherein the one or more computer programs may be stored in the memory and configured to be executed by one or more processors, and one or more application programs are configured to perform the frequency response service attack detection method described above.

[0152] A processor may include one or more processing cores. The processor connects to various parts of the computer device using various interfaces and lines, and performs various functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in memory, and by calling data stored in memory. Optionally, the processor may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the displayed content; and the modem handles wireless communication. It is understood that the modem may also be implemented separately as a communication chip, without being integrated into the processor.

[0153] The memory may include random access memory (RAM) or read-only memory (ROM). The memory can be used to store instructions, programs, code, code sets, or instruction sets. The memory may include a program storage area and a data storage area. The program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as touch functionality, sound playback functionality, image playback functionality, etc.), and instructions for implementing the various method embodiments described above. The data storage area may also store data created by the terminal device during use.

[0154] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting frequency response service attacks, wherein the method is applied to a frequency response control system, characterized in that, include: Step S1: Obtain the frequency measurement signal and the area control error measurement signal of the power grid control area; Step S2: Input the frequency measurement signal into the trained adaptive neural fuzzy inference model to obtain the regional control error prediction signal of the power grid control area; Step S3: Obtain the root mean square error based on the measured regional control error signal and the predicted regional control error signal; Step S4: If the root mean square error exceeds a preset threshold, then there is an attack in the power grid control area, and the area control error is set as the area control error prediction signal. Step S5: Repeat steps S1-S4 until the frequency response control system runs for the time specified in the preset time threshold. The training process of the adaptive neural fuzzy reasoning model includes: Step S301: Obtain multiple historical frequency measurement signals and multiple historical area control error measurement signals of the power grid control area, and merge each historical frequency measurement signal with the corresponding historical area control error measurement signal into one historical data. Step S302: Divide multiple historical data sets into training and testing sets according to a preset ratio; Step S303: Input the training set into the adaptive neurofuzzy inference model to adjust the parameters and obtain the preliminary adaptive neurofuzzy inference model; Step S304: Input the validation set into the preliminary adaptive neural fuzzy inference model to obtain the region control error prediction signal, and obtain the training error based on the historical region control error measurement signal and the region control error prediction signal; Step S305: If the training error satisfies the output condition, the preliminary adaptive neurofuzzy inference model is output as the trained adaptive neurofuzzy inference model. Step S306: If the training error does not meet the output condition, return to steps S303-S304 until the training error meets the output condition.

2. The frequency response service attack detection method as described in claim 1, characterized in that, Step S4 also includes: If the root mean square error does not exceed the set threshold, then there is no attack behavior in the power grid control area, and the area control error is set as the area control error measurement signal.

3. The frequency response service attack detection method as described in claim 2, characterized in that, The adaptive neural fuzzy reasoning model includes an input layer, a fuzzification layer, a rule layer, a normalization layer, and a defuzzification layer; The specific expression for the input layer is: ， in, For input variables, For frequency measurement signals in the power grid control area; The specific expression for the blurring layer is: ， in, For fuzzy sets, The membership degree of the input variable to the fuzzy set. It is a Gaussian membership function; The specific expression for the fuzzy rule in the rule layer is: Rule i: If yes ,but , in, For output variables, The first parameter to be determined, The second parameter to be determined is the activation weight of each rule, which is expressed as follows: , For the The activation weight of each rule; The specific expression for the normalization layer is: ， in, The normalized rule activation strength, The total number of rules; The specific expression for the deblurring layer is: 。 4. The frequency response service attack detection method as described in claim 3, characterized in that, The root mean square error is obtained based on the measured signal and the predicted signal of the regional control error, and the specific expression is as follows: ， in, The root mean square error, For regional control error prediction signals, This is the area control error measurement signal. This represents the number of nodes in the power grid control area.

5. The frequency response service attack detection method as described in claim 1, characterized in that, The expression for the training error is: ， ， in, For training error, For the area control error prediction signal, This is a historical area control error measurement signal. For the sample size, The parameters of the adjusted adaptive neurofuzzy inference model, The parameters of the adaptive neurofuzzy inference model before adjustment. This is the learning rate.

6. The frequency response service attack detection method as described in claim 5, characterized in that, Attacks on the power grid control area include frequency measurement data attacks, regional tie-line power measurement data attacks, and regional ACE signal attacks. The specific expression for the frequency measurement data attack is as follows: ， in, For the first Each power grid control area in Frequency measurement data at a given time is used to attack the signal. For the first Frequency deviation coefficient of each power grid control area For the first Frequency measurement signals for each power grid control area This is the first attack signal. For the first Regional capacity coefficient of each power grid control area For the first The first power grid control area and the first Power deviation of tie lines between power grid control areas This represents the total number of power grid control areas. The specific expression for the regional tie-line power measurement data attack is as follows: ， in, For the The first power grid control area and the first Each power grid control area in Attack signal on regional tie-line power measurement data at a given time. This is the second attack signal. For the Regional capacity coefficient of each power grid control area For the The first power grid control area and the first Power deviation of tie lines between power grid control areas; The specific expression for the regional ACE signal attack is as follows: ， in, For the Each power grid control area in ACE signal attack signal at any time For the Each power grid control area in Normal ACE signal at any time This is the third attack signal.

7. A frequency response service attack detection device, wherein the frequency response service attack detection device is applied to a frequency response control system, characterized in that, include: The signal acquisition module is used to acquire frequency measurement signals and area control error measurement signals of the power grid control area; The signal prediction module is used to input the frequency measurement signal into the trained adaptive neural fuzzy inference model to obtain the area control error prediction signal of the power grid control area; The root mean square error calculation module is used to obtain the root mean square error based on the area control error measurement signal and the area control error prediction signal. An attack detection module is used to set the regional control error as a regional control error prediction signal if the root mean square error exceeds a preset threshold. The running time setting module is used to repeat all the above modules until the running time of the frequency response control system reaches the preset time threshold. The training process of the adaptive neural fuzzy reasoning model includes: Step S301: Obtain multiple historical frequency measurement signals and multiple historical area control error measurement signals of the power grid control area, and merge each historical frequency measurement signal with the corresponding historical area control error measurement signal into one historical data. Step S302: Divide multiple historical data sets into training and testing sets according to a preset ratio; Step S303: Input the training set into the adaptive neurofuzzy inference model to adjust the parameters and obtain the preliminary adaptive neurofuzzy inference model; Step S304: Input the validation set into the preliminary adaptive neural fuzzy inference model to obtain the region control error prediction signal, and obtain the training error based on the historical region control error measurement signal and the region control error prediction signal; Step S305: If the training error satisfies the output condition, the preliminary adaptive neurofuzzy inference model is output as the trained adaptive neurofuzzy inference model. Step S306: If the training error does not meet the output condition, return to steps S303-S304 until the training error meets the output condition.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of any one of the frequency response service attack detection methods in claims 1-6.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it performs any one of the frequency response service attack detection methods according to claims 1-6.

Images