Fractional order power system large model construction and security prediction method under network attack

Abstract

The invention relates to the technical field of electric power system operation control, in particular to a fractional order electric power system large model construction and security prediction method under network attacks, which comprises the following steps: constructing an electric power system fractional order first model according to a feature sample data set of each subsystem in an electric power system under a first state; according to intermittent network attack modeling and the characteristic sample data set of each subsystem in a second state, determining a power system fractional order second model; according to the feature sample data set of each subsystem in the second state and the power system fractional order second model, constructing a power system fractional order multiple regression model in the second state; determining the sensitivity of each subsystem in the power system to parameter perturbation and intermittent network attack; and constructing a prediction model according to error indexes of the sensitivity, the intermittent network attack initial prediction model and the parameter perturbation initial prediction model. According to the method, the estimation precision of intermittent network attacks and parameter perturbation suffered by the fractional order power system is improved.

Description

Technical Field

[0001] This specification relates to the field of power system operation control technology, especially the construction of a large model of a fractional-order power system under network attacks and a security prediction method. Background Art

[0002] At present, as an emerging network communication architecture, deterministic networks guarantee service quality through technologies such as resource reservation, service guarantee, and explicit routing. Therefore, more stringent requirements are placed on data transmission in deterministic networks (for example, low packet loss rate between end-to-end and real-time accuracy of information). However, power systems are vulnerable to intermittent network attacks and parameter perturbations during operation, resulting in low estimation accuracy, which in turn affects their safe and reliable operation.

[0003] In the prior art, the Chinese invention patent with application number 202010801108.6 uses a deep learning architecture to establish a nonlinear mapping relationship between time series feature quantities and frequency security, realizes end-to-end frequency security assessment, and improves the assessment accuracy by optimizing the key parameters of the power system frequency security assessment model. However, deep learning models often require a large amount of training data and are prone to the risk of overfitting, especially when the data is insufficient or the data quality is poor, the generalization ability of the model may be poor, and it is impossible to fully and accurately describe the system characteristics, thereby affecting the accuracy of the assessment results. The Chinese invention patent with application number 202210272522.1 estimates the attack signal by using a decentralized sliding mode observer. Although this method is effective in some scenarios, its performance is highly dependent on parameter settings. Improper parameter settings will weaken the detection capability of the observer, resulting in false positives or false negatives of attack signals, and reduce the estimation accuracy, and will also aggravate the chattering phenomenon of the controller under network attacks. In addition, the above method does not consider the estimation of system parameter perturbations, ignores its impact on the system state, and limits its application and effectiveness in complex power systems. Summary of the invention

[0004] In order to solve the problem that the existing technology cannot fully and accurately describe the system characteristics, and has low estimation accuracy when facing intermittent network attacks and system parameter perturbations, this manual provides a large model construction and security prediction method for fractional-order power systems under network attacks.

[0005] The embodiment of the present specification provides a method for constructing a large model of a fractional-order power system under a network attack, the method comprising: constructing a first fractional-order model of the power system according to a feature sample data set of each subsystem in the power system in a first state; determining a second fractional-order model of the power system according to intermittent network attack modeling, a feature sample data set of each subsystem in a second state and the first fractional-order model of the power system, wherein the second state indicates that the power system is subjected to parameter perturbation and intermittent network attack; constructing a fractional-order multivariate regression model of the power system in the second state according to the feature sample data set of each subsystem in the second state and the second fractional-order model of the power system; determining the sensitivity of each subsystem in the power system to parameter perturbation and intermittent network attack according to the fractional-order multivariate regression model of the power system; constructing an intermittent network attack initial prediction model and a parameter perturbation initial prediction model respectively according to the sensitivity; iteratively updating weight parameters in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model according to error indicators of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model, constructing an intermittent network attack prediction model and a parameter perturbation prediction model, wherein the intermittent network attack prediction model and the parameter perturbation prediction model constitute a large model of the fractional-order power system under a network attack.

[0006] According to one aspect of an embodiment of this specification, constructing a first fractional-order model of a power system according to a feature sample data set of each subsystem in a power system in a first state includes:

[0007] The fractional-order model of the power system is constructed by the following formula:

[0008] Among them, D α represents the fractional differential operator, x(t) represents the power system state matrix, x(t)=[ΔP AE ,ΔP FC ,ΔP DEG ,ΔP BESS , ΔP c ,Δf] T , ΔP AE Indicates the change in power generation of the water electrolyzer, ΔP FC Indicates the change in fuel cell power generation, ΔP DEG Indicates the change in power generation of the diesel generator, ΔP BESS Indicates the change in power generation of the battery energy storage system, ΔP c represents the output change of the integral controller, Δf represents the frequency change of the power system, represents the power system state moment, T AE represents the time constant of the water electrolyzer, T FC represents the fuel cell time constant, T DEG Represents the diesel engine time constant; TBESS Represents the time constant of the battery energy storage system, K I represents the integral controller gain, K p Represents the frequency characteristic coefficient of the power system, T p represents the power system time constant, represents the power system control matrix; u(t) represents the control input of the power system, B u =[K AE / T AE ,K FC / T FC ,K DEG / T DEG ,K BESS / T BESS ,0,0],K AE represents the water electrolyzer gain, K FC represents the fuel cell gain, K DEG Indicates diesel engine gain, K BESS Represents the battery energy storage system gain, ΔA, ΔB u They represent the parameter perturbation matrices of the power system state matrix and control matrix, y(t) represents the power system output, and C represents the power system output matrix;

[0009] Where f(x) is a function, Γ(·) is a gamma function, x is a variable, α is a fractional order, n is an integer higher than α, and ξ is the integral variable.

[0010] According to one aspect of an embodiment of this specification, based on intermittent network attack modeling, a characteristic sample data set of each subsystem in the second state, and the first fractional-order model of the power system, determining the second fractional-order model of the power system includes: constructing the second fractional-order model of the power system by the following formula:

[0011] Among them, δ(x,t)=ΔAx(t)+ΔB u u(t) represents the fractional-order power system parameter perturbation, and η(t) represents the intermittent network attack related to the system state, which is determined by the intermittent network attack modeling.

[0012] According to one aspect of an embodiment of the present specification, based on the fractional-order multivariate regression model of the power system, determining the sensitivity of each subsystem in the power system to parameter perturbations and intermittent network attacks includes: using the least squares method to estimate the regression coefficient matrix of multiple subsystems of the fractional order of the power system; determining the sum of the second element and the fourth element in each regression coefficient matrix as the sensitivity of the subsystem to intermittent network attacks; and determining the sum of the third element and the fifth element in each regression coefficient matrix as the sensitivity of the subsystem to parameter perturbations.

[0013] According to one aspect of an embodiment of the present specification, constructing an intermittent network attack initial prediction model and a parameter perturbation initial prediction model respectively according to the sensitivity includes: constructing an intermittent network attack estimation weight matrix according to the sensitivity of the subsystem to intermittent network attacks; constructing a parameter perturbation estimation weight matrix according to the sensitivity of the subsystem to parameter perturbations; determining the output of the hidden layer neurons of the intermittent network attack prediction model according to the intermittent network attack estimation weight matrix; determining the output of the hidden layer neurons of the parameter perturbation prediction model according to the parameter perturbation estimation weight matrix; determining the intermittent network attack initial prediction model according to the output of the hidden layer neurons of the intermittent network attack prediction model and the output weight of the prediction model; determining the parameter perturbation initial prediction model according to the output of the hidden layer neurons of the parameter perturbation prediction model and the output weight of the prediction model.

[0014] According to one aspect of an embodiment of this specification, based on the sensitivity, constructing an intermittent network attack initial prediction model and a parameter perturbation initial prediction model respectively includes:

[0015] The characteristic data sets of each subsystem collected in the second state are input into the intermittent network attack prediction model to determine the output of the neurons in the hidden layer of the intermittent network attack prediction model:

[0016] h j,m is the output of the mth hidden layer neuron of the intermittent network attack prediction model of the super-helical fractional-order constrained RBF neural network, ρ j represents the intermittent network attack prediction weight matrix or parameter perturbation prediction weight matrix, c j,m represents the center point vector value of the mth hidden layer neuron in the intermittent network attack prediction model of the super-helical fractional-order constrained RBF neural network, b j,m represents the width of the Gaussian basis function of the mth hidden layer neuron in the intermittent network attack prediction model of the super-helical fractional-order constrained RBF neural network, and b j,m >0, subscript m=1, 2, …, M respectively represent the mth hidden layer neuron of the super helical fractional order constrained RBF neural network intermittent network attack prediction model, and M is the total number of hidden layer neurons of the super helical fractional order constrained RBF neural network intermittent network attack prediction model.

[0017] According to one aspect of an embodiment of this specification, the error index of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model is determined by the following formula:

[0018] Among them, E 1 (t) represents the intermittent network attack prediction error evaluation index, ν1 represents the prediction error weight of intermittent network attacks, η(t) represents the intermittent network attacks on the system, E 2 (t) represents the parameter perturbation prediction error evaluation index, ν 2 represents the parameter perturbation error weight; δ(x,t) represents the system parameter perturbation, They are the observed values ​​of η(t) and δ(x, t) of the initial prediction model of intermittent network attacks and the initial prediction model of parameter perturbation, respectively.

[0019] According to one aspect of an embodiment of the present specification, iteratively updating the weight parameters in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model includes: using a superhelix algorithm to construct a superhelix path to guide the weights in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model to converge along the optimal direction; using an exponentially dependent barrier function to constrain the outputs of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model.

[0020] The embodiments of this specification also provide a security prediction method, including: obtaining characteristic data of each subsystem in the power system in the second state; inputting the characteristic data of each subsystem in the power system in the second state into an intermittent network attack prediction model and a parameter perturbation prediction model to obtain prediction results for each subsystem.

[0021] The embodiment of the present specification also provides a device for constructing a large model of a fractional-order power system under a network attack, the device comprising: a first construction unit, for constructing a first fractional-order model of the power system according to a feature sample data set of each subsystem in the power system in a first state; a second construction unit, for determining a second fractional-order model of the power system according to intermittent network attack modeling, a feature sample data set of each subsystem in a second state and the first fractional-order model of the power system, wherein the second state indicates that the power system is subjected to parameter perturbation and intermittent network attack; a third construction unit, for constructing a fractional-order multivariate regression model of the power system in the second state according to the feature sample data set of each subsystem in the second state and the second fractional-order model of the power system; A determining unit is used to determine the sensitivity of each subsystem in the power system to parameter perturbations and intermittent network attacks according to the fractional-order multivariate regression model of the power system; a fourth constructing unit is used to construct an intermittent network attack initial prediction model and a parameter perturbation initial prediction model respectively according to the sensitivity; a fifth constructing unit is used to iteratively update the weight parameters in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model according to the error indicators of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model, and construct an intermittent network attack prediction model and a parameter perturbation prediction model, which constitute a large model of the fractional-order power system under network attacks.

[0022] The embodiments of this specification also provide a security prediction device, which includes: an acquisition unit, used to acquire characteristic data of each subsystem in the power system in the second state; a prediction unit, used to input the characteristic data of each subsystem in the power system in the second state into an intermittent network attack prediction model and a parameter perturbation prediction model to obtain prediction results for each subsystem.

[0023] An embodiment of the present specification also provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the method for constructing a large model of a fractional-order power system under network attacks and for security prediction.

[0024] The embodiments of the present specification also provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for constructing a large model of a fractional-order power system under a network attack and for predicting security is implemented.

[0025] The present invention improves the estimation accuracy of intermittent network attacks and parameter perturbations suffered by fractional-order power systems, effectively making up for the deficiency that traditional methods cannot fully and accurately describe system characteristics, especially the problem of low estimation accuracy when facing intermittent network attacks and system parameter perturbations, and significantly improves the accuracy and robustness of security assessment.

[0026] This manual uses the least squares method to estimate the regression coefficients of intermittent network attacks and parameter perturbations, analyzes the sensitivity of the fractional-order power system state to intermittent network attacks and parameter perturbations, designs the estimated weights of intermittent network attacks and parameter perturbations, and compensates the neural network prediction model; the neural network prediction model uses a superhelix algorithm to construct a superhelix path to guide the neural network weights to converge rapidly in the optimal direction, and uses an exponentially dependent barrier function to constrain the output of the neural network, effectively preventing the neural network from experiencing large fluctuations during the training process, thereby improving the prediction accuracy of intermittent network attacks and system parameter perturbations. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0028] Figure 1 It is a flowchart of a method for constructing a large model of a fractional-order power system under a network attack according to an embodiment of this specification;

[0029] Figure 2 Shown is a flow chart of a method for determining the sensitivity of each subsystem to parameter perturbations and intermittent network attacks according to an embodiment of this specification;

[0030] Figure 3 The flowchart of a method for constructing an intermittent network attack initial prediction model and a parameter perturbation initial prediction model according to an embodiment of the present specification is shown;

[0031] Figure 4 The flowchart of a method for iteratively updating an initial prediction model of intermittent network attacks and an initial prediction model of parameter perturbation according to an embodiment of the present specification is shown;

[0032] Figure 5 The flowchart of an embodiment of this specification showing intermittent network attacks and parameter perturbation predictions for each subsystem;

[0033] Figure 6 The figure shows a schematic diagram of the structure of a device for constructing a large model of a fractional-order power system under a network attack according to an embodiment of the present specification;

[0034] Figure 7 The figure shows a schematic diagram of the structure of a safety prediction device according to an embodiment of the present specification;

[0035] Figure 8 The figure shows a schematic diagram of the structure of a computer device according to an embodiment of the present specification.

[0036] Description of the accompanying symbols:

[0037] 601. A first building block;

[0038] 602, second building unit;

[0039] 603, third building block;

[0040] 604. Determine the unit;

[0041] 605, fourth building block;

[0042] 606, fifth building block;

[0043] 701, a first acquiring unit;

[0044] 702. prediction result determination unit;

[0045] 802. Computer equipment;

[0046] 804, processor;

[0047] 806. Memory;

[0048] 808, driving mechanism;

[0049] 810, input / output module;

[0050] 812. Input device;

[0051] 814. Output device;

[0052] 816. Presentation equipment;

[0053] 818. Graphical user interface;

[0054] 820, network interface;

[0055] 822, communication link;

[0056] 824. Communication bus. DETAILED DESCRIPTION

[0057] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this specification.

[0058] It should be noted that the terms "first", "second", etc. in the description and claims of this specification and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of this specification described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product or equipment that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or equipment.

[0059] This specification provides method operation steps as described in the embodiments or flow charts, but more or fewer operation steps may be included based on routine or non-creative work. The order of steps listed in the embodiments is only one way of executing the order of many steps and does not represent the only execution order. When the system or device product is executed in practice, it can be executed in the order of the method shown in the embodiments or the drawings or in parallel.

[0060] It should be noted that the methods, devices and computer equipment of this specification can be used in the field of power system operation control technology. This specification does not limit the application field of the construction of large-scale models of fractional-order power systems and security prediction methods under network attacks.

[0061] Figure 1 The flowchart of a method for constructing a large model of a fractional-order power system under a network attack according to an embodiment of the present specification is shown, which specifically includes the following steps:

[0062] Step 101: construct a first fractional-order model of the power system according to a characteristic sample data set of each subsystem in the power system in a first state.

[0063] In this step, the first state represents the state of the power system when it is operating normally. In the first state, characteristic sample data of each subsystem in the power system is collected to construct a fractional-order first model of the power system. In the embodiment of this specification, the power system includes multiple types of subsystems, specifically including: water electrolyzer system, fuel cell system, diesel generator system, battery energy storage system, etc. Among them, each subsystem will generate different characteristic data when operating in the first state. For example, the characteristic data in the water electrolyzer system is the change in power generation of the water electrolyzer; the characteristic data in the fuel cell system is the change in power generation of the fuel cell; the characteristic data in the diesel generator system is the change in power generation of the diesel generator.

[0064] In order to construct the first fractional-order model of the power system, this step collects the characteristic data of each subsystem in the power system when it is running in the first state as sample data to form a characteristic sample data set. The characteristic sample data set is further used to construct the first fractional-order model of the power system, specifically including:

[0065] According to the characteristic sample data set of each subsystem in the power system in the first state, the first state matrix of the power system and the second state matrix of the power system used to describe the system response speed and the power system control matrix are constructed. Specifically, according to the time constant of each subsystem in the power system, the power system state matrix used to describe the system response speed is determined; according to the ratio of the gain of each subsystem to the time constant, the parameter perturbation matrix is ​​determined. Among them, the gain of each system is obtained through experimental testing.

[0066] Furthermore, based on the characteristic sample data set of each subsystem in the power system in the first state, a power system state matrix is ​​constructed.

[0067] The first fractional-order model of the power system is constructed by the following formula:

[0068] Among them, D α represents the fractional differential operator, x(t) represents the power system state matrix, represents the power system state matrix, represents the power system control matrix; u(t) represents the control input of the power system, ΔA, ΔB u denote the parameter perturbation matrices of the power system state matrix and control matrix respectively, y(t) denotes the power system output, C denotes the power system output matrix, x(t) = [ΔP AE ,ΔP FC ,ΔP DEG ,ΔP BESS , ΔP c ,Δf] T , ΔP AE Indicates the change in power generation of the water electrolyzer, ΔP FC Indicates the change in fuel cell power generation, ΔP DEG Indicates the change in power generation of the diesel generator, ΔP BESS Indicates the change in power generation of the battery energy storage system, ΔP c represents the output change of the integral controller, Δf represents the frequency change of the power system, B u =[K AE / T AE ,K FC / T FC ,K DEG / T DEG ,K BESS / T BESS ,0,0],K AE represents the water electrolyzer gain, K FC represents the fuel cell gain, K DEG Denotes diesel engine gain, K BESS Indicates the gain of the battery energy storage system. In this specification, various gains can be obtained through experimental testing.

[0069] in, T AE Represents the time constant of the water electrolyzer, which is used to describe the system response speed; T FC represents the fuel cell time constant, T DEG Represents the diesel engine time constant; T BESS Represents the time constant of the battery energy storage system, K I represents the integral controller gain, K p Represents the frequency characteristic coefficient of the power system, T p Represents the power system time constant.

[0070] Before constructing the first fractional-order model of the power system, a fractional-order differential operator is first constructed to obtain the fractional-order differential of the function: Where f(x) is a function, Γ(·) is a gamma function, x represents a variable, α represents a fractional order, nrepresents an integer higher than α, and ξ represents the integration variable.

[0071] Step 102, determining a second fractional-order model of the power system based on intermittent network attack modeling, a characteristic sample data set of each subsystem in the second state, and the first fractional-order model of the power system, wherein the second state indicates that the power system is subjected to parameter perturbation and intermittent network attacks.

[0072] In the embodiment of this specification, the second state indicates that the power system is subjected to parameter perturbation and indirect network attack. In the second state, it is necessary to construct a second fractional order model of the power system based on the characteristic sample data of each subsystem in the power system and the first fractional order model of the power system constructed in step 101.

[0073] Specifically, on the basis of the first fractional-order model of the power system, the parameters perturbation related factors of the power system and the intermittent network attacks related to the system status are added.

[0074] Specifically, the second fractional-order model of the power system is constructed by the following formula:

[0075] Among them, δ(x,t)=ΔAx(t)+ΔB u u(t) represents the perturbation of fractional-order power system parameters, and η(t) represents the intermittent network attack related to the system state, which is determined by the intermittent network attack modeling. Among them, δ(x,t)=ΔAx(t)+ΔB u u(t) is the fractional-order power system parameter perturbation; Is a positive number.

[0076] Step 103 , constructing a fractional-order multivariate regression model of the power system in the second state according to the characteristic sample data set of each subsystem in the second state and the second fractional-order model of the power system.

[0077] In this step, intermittent network attacks can cause system stability to decline and increase the risk of failure by interfering in irregular short periods of time. At the same time, attackers do not need to continuously occupy a large amount of resources. Therefore, this manual first models intermittent network attacks:

[0078] Where η(t) represents the intermittent network attack related to the system state, ρ(t) represents the gain coefficient of the intermittent network attack, and T DoS (t) represents the weight of intermittent network attack, t represents the current time t, n∈N represents the number of cycles, T represents the duration of a cycle of the power system, T on (t) represents the starting time of the intermittent network attack, T off (t) represents the attack duration of the intermittent network, T on(t) and T off (t) varies randomly and satisfies T on (t)+T off (t)≤T.

[0079] According to the intermittent network attack modeling and the second fractional-order model of the power system, when the power system described in step 102 is subjected to intermittent network attacks and parameter perturbations, the characteristic data and interference data set [η p (t),δ p (x,t)].

[0080] Among them, the characteristic data set of the pth subsystem is as follows: [ΔP AE,p ,ΔP FC,p ,ΔP DEG,p ,ΔP BESS,p , ΔP c,p ,Δf p ] T , subscript p = 1, 2, ..., n. Among them, ΔP AE Indicates the change in power generation of the water electrolyzer, ΔP FC Indicates the change in fuel cell power generation, ΔP DEG Indicates the change in power generation of the diesel generator, ΔP BESS Indicates the change in power generation of the battery energy storage system, ΔP c Represents the output change of the integral controller. Among them, η(t) in the interference data set is the intermittent network attack related to the system state, δ p (x, t) represents the fractional-order power system parameter perturbation.

[0081] Furthermore, based on the characteristic data of each subsystem when it is subjected to parameter perturbations and intermittent network attacks, as well as the regression coefficient matrix of the characteristic data of the subsystem, the variation error matrix of the characteristic data of each subsystem and the polynomial characteristic matrix, a fractional-order multivariate regression model of the power system in the second state is constructed. For a detailed description of the fractional-order multivariate regression model of the power system, see Figure 2 describe.

[0082] Step 104 , determining the sensitivity of each subsystem in the power system to parameter perturbations and intermittent network attacks according to the power system fractional-order multivariate regression model.

[0083] After constructing the fractional-order multivariate regression model of the power system, the least squares method is used to estimate the regression coefficients of intermittent network attacks and parameter perturbations, so as to analyze the sensitivity of the fractional-order power system state to intermittent network attacks and parameter perturbations.

[0084] Step 105: constructing an intermittent network attack initial prediction model and a parameter perturbation initial prediction model respectively according to the sensitivity.

[0085] In this step, after determining the sensitivity of each subsystem in the power system to the parameter perturbations and intermittent network attacks suffered by the power system, the sensitivity of all subsystems to the intermittent network attacks suffered by the power system is combined to form an intermittent network attack estimation weight matrix of the initial prediction model of intermittent network attacks; the sensitivity of all subsystems to the parameter perturbations suffered by the power system is combined to form a parameter perturbation estimation weight matrix of the initial prediction model of parameter perturbations. By introducing the intermittent network attack estimation weight matrix and the parameter perturbation estimation weight matrix into the initial prediction model, the initial prediction model pays more attention to the system state variables with greater sensitivity to intermittent network attacks or parameter perturbations.

[0086] Step 106, according to the error indicators of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model, iteratively update the weight parameters in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model, and construct an intermittent network attack prediction model and a parameter perturbation prediction model, which constitute a large model of the fractional-order power system under network attack.

[0087] This step uses the superhelix algorithm to construct a superhelix path, guides the weights of each layer of the neural network to converge quickly, and uses an exponentially dependent barrier function to constrain the output of the neural network to prevent large fluctuations in the neural network during the training process, thereby improving the estimation accuracy of intermittent network attacks and parameter perturbations suffered by the fractional-order power system, and finally constructs a large model of the fractional-order power system under network attacks. For a detailed description of training and building intermittent network attack prediction models and parameter perturbation prediction models, see Figure 4 describe.

[0088] Figure 2 The flowchart of a method for determining the sensitivity of each subsystem to parameter perturbations and intermittent network attacks according to an embodiment of the present specification is shown, which specifically includes the following steps:

[0089] Step 201 : using the least square method to estimate the regression coefficient matrix of multiple fractional-order subsystems of the power system.

[0090] In this step, before determining the regression coefficient matrix of each subsystem, the characteristic sample data of each subsystem in the power system in the second state is first collected to construct a fractional-order multivariate regression model of the power system. The formula of the multivariate regression model is as follows:

[0091] Among them, ΔP AE The data set represents the change in power generation of the water electrolyzer in the second state, ΔP AE =[ΔP AE,1 ΔPAE,2 … ΔP AE,n ] T , ΔP FC The data set representing the change in power generation of the fuel cell in the second state, ΔP FC =[ΔP FC,1 ΔP FC,2 … ΔP FC,n ] T , ΔP DEG The data set of diesel engine power generation change in the second state, ΔP DEG =[ΔP DEG,1 ΔP DEG,2 … ΔP DEG,n ] T , ΔP BESS Represents the data set of the power generation change of the battery energy storage system in the second state, ΔP BESS =[ΔP BESS,1 ΔP BESS,2 … ΔP BESS,n ] T , ΔP represents the output change data set of the integral controller, ΔP=[ΔP 1 ΔP 2 … ΔP n ] T , Δf represents the power system frequency variation data set Δf=[Δf 1 Δf 2 … Δf n ] T , β 1 Represents the regression coefficient matrix of the change in power generation of the water electrolyzer in the second state: β 1 =[β 1,1 ,β 2,1 ,β 3,1 ,β 4,1 ,β 5,1 ,β 6,1 ] T , β 2 Represents the regression coefficient matrix of the fuel cell power generation change in the second state: β 2 =[β 1,2 ,β 2,2 ,β 3,2 ,β 4,2 ,β 5,2 ,β 6,2 ] T , β 3 Represents the regression coefficient matrix of the diesel engine power generation change in the second state: β 3 =[β 1,3 ,β 2,3 ,β 3,3 ,β 4,3 ,β 5,3,β 6,3 ] T , β 4 Represents the regression coefficient matrix of the battery energy storage system power generation change in the second state: β 4 =[β 1,4 ,β 2,4 ,β 3,4 ,β 4,4 ,β 5,4 ,β 6,4 ] T , β 5 Represents the regression coefficient matrix of the integral controller output change in the second state: β 5 =[β 1,5 ,β 2,5 ,β 3,5 ,β 4,5 ,β 5,5 ,β 6,5 ] T , β 6 Represents the regression coefficient matrix of the power system frequency change in the second state: β 6 =[β 1,6 ,β 2,6 ,β 3,6 ,β 4,6 ,β 5,6 ,β 6,6 ] T , ε 1 The error matrix of the change in power generation of the water electrolyzer in the second state is: 1 =[ε 1,1 ,ε 2,1 ,ε 3,1 ,ε 4,1 ,ε 5,1 ,ε 6,1 ] T , ε 2 The error matrix representing the change in fuel cell power generation in the second state is: 2 =[ε 1,2 ,ε 2,2 ,ε 3,2 ,ε 4,2 ,ε 5,2 ,ε 6,2 ] T , ε 3 The error matrix of the diesel engine power generation change in the second state is: 3 =[ε 1,3 ,ε 2,3 ,ε 3,3 ,ε 4,3 ,ε 5,3 ,ε 6,3 ] T , ε 4Represents the error matrix of the battery energy storage system power generation change in the second state: ε 4 =[ε 1,4 ,ε 2,4 ,ε 3,4 ,ε 4,4 ,ε 5,4 ,ε 6,4 ] T , ε 5 Represents the error matrix of the integral controller output change in the second state: ε 5 =[ε 1,5 ,ε 2,5 ,ε 3,5 ,ε 4,5 ,ε 5,5 ,ε 6,5 ] T , ε 6 Represents the frequency variation error matrix of the power system in the second state: ε 6 =[ε 1,6 ,ε 2,6 ,ε 3,6 ,ε 4,6 ,ε 5,6 ,ε 6,6 ] T , X poly Represents the polynomial characteristic matrix

[0092] After constructing the power system fractional-order multivariate regression model, the power system fractional-order regression coefficient matrix of each subsystem in the power system fractional-order multivariate regression model is estimated by the least squares method. The form of the power system fractional-order regression coefficient matrix of each subsystem is as follows:

[0093] in, represents the regression coefficient matrix of the water electrolyzer power generation system, represents the regression coefficient matrix of the fuel cell power generation system, represents the regression coefficient matrix of the diesel generator system, represents the regression coefficient matrix of the battery energy storage system, represents the regression coefficient matrix of the integral controller system, represents the regression coefficient matrix of the power system frequency system, X polt is the polynomial feature matrix X in the multiple regression model poly .

[0094] Step 202: The sum of the second element and the fourth element in each regression coefficient matrix is ​​determined as the sensitivity of the subsystem to the intermittent network attack.

[0095] In the embodiment of this specification, the regression coefficient matrix β of the water electrolyzer power generation system 1 , the regression coefficient matrix β of the fuel cell power generation system 2 , the regression coefficient matrix β of the diesel generator system 3 , the regression coefficient matrix β of the battery energy storage system power generation system 4 , the regression coefficient matrix β of the integral controller system 5 and the regression coefficient matrix β of the power system frequency system 6 Each of the regression coefficient matrices of the six subsystems includes 6 elements, which are respectively called the first element, the second element, the third element, the fourth element, the fifth element and the sixth element.

[0096] In this specification, according to the formula of the fractional-order multivariate regression model of the power system, it can be determined that the characteristic data set of each subsystem in the power system can be expressed as the multiplication of the polynomial characteristic matrix and the regression coefficient matrix of each subsystem. For example, the water electrolyzer power generation change data set ΔP in the second state AE Expand it and get:

[0097] Among them, η 1 ×β 2 Item and These two items are related to the intermittent network attacks suffered by the water electrolyzer power generation system. Therefore, the coefficients β before these two items are 2 and β 4 It is determined as the coefficient related to the sensitivity of fractional-order power systems to intermittent cyber attacks.

[0098] Specifically, the second element and the fourth element in the regression coefficient matrix of each subsystem are added to determine the sensitivity of the fractional-order power system to intermittent network attacks. It is regarded as the sensitivity of the fractional-order power system to intermittent network attacks η(t). Thus, the sensitivities of the water electrolyzer power generation system, fuel cell power generation system, diesel engine power generation system, battery energy storage system, integral controller system and power system frequency system to intermittent network attacks are determined respectively.

[0099] After determining the sensitivity of each subsystem to intermittent network attacks, the regression coefficient matrix of each subsystem is further sorted according to the sensitivity to intermittent network attacks. The greater the sensitivity, the greater the impact of intermittent network attacks on the state of the fractional-order power system; the smaller the sensitivity, the smaller the impact of intermittent network attacks on the state of the fractional-order power system.

[0100] Step 203: The sum of the third element and the fifth element in each regression coefficient matrix is ​​determined as the sensitivity of the subsystem to the parameter perturbation.

[0101] In the embodiment of the present specification, similar to the determination of the term related to intermittent network attack in step 202, the determination of β in this step is 3 With β 5 Determined as a coefficient related to the sensitivity of the fractional order power system to parameter perturbations.

[0102] The third and fifth elements in the regression coefficient matrix of each subsystem are added together to determine the sensitivity of the fractional-order power system to intermittent network attacks. It is regarded as the sensitivity of the fractional-order power system to the parameter perturbation δ(x,t). Thus, the sensitivity of the water electrolyzer power generation system, fuel cell power generation system, diesel engine power generation system, battery energy storage system, integral controller system and power system frequency system to the parameter perturbation is determined respectively. And the sensitivity of the regression coefficient matrix of each subsystem to the parameter perturbation is sorted according to the size. Among them, the greater the sensitivity, the greater the impact of the parameter perturbation on the state of the fractional-order power system; the smaller the sensitivity, the smaller the impact of the parameter perturbation on the state of the fractional-order power system.

[0103] Figure 3 The flowchart of a method for constructing an intermittent network attack initial prediction model and a parameter perturbation initial prediction model according to an embodiment of the present specification is shown, which specifically includes the following steps:

[0104] Step 301, constructing an intermittent network attack prediction weight matrix according to the sensitivity of the subsystem to the intermittent network attack.

[0105] This step is based on Figure 2 The sensitivity calculated in is used to calculate the intermittent network attack estimation weight matrix and the parameter perturbation estimation weight matrix, as shown in the following formula:

[0106] Among them, ρ 1 It represents the intermittent network attack estimation weight matrix. The six elements in the intermittent network attack estimation weight matrix are the sum of the sensitivity of each of the six subsystems to intermittent network attacks and the sensitivity of all subsystems to intermittent network attacks.

[0107] Step 302: construct a parameter perturbation prediction weight matrix according to the sensitivity of the subsystem to the parameter perturbation.

[0108] Among them, ρ 2 Represents the parameter perturbation estimation weight matrix. The six elements in the parameter perturbation estimation weight matrix are the sum of the sensitivity of each of the six subsystems to intermittent network attacks and the sensitivity of all subsystems to intermittent network attacks.

[0109] Step 303: Determine the output of the hidden layer neurons of the intermittent network attack initial prediction model according to the intermittent network attack estimation weight matrix.

[0110] In this step, the intermittent network attack prediction model is constructed using the super-helical fractional order algorithm. In order to improve the accuracy of the model prediction precision, the intermittent network attack prediction weight matrix established in step 302 is introduced to make the intermittent network attack prediction model pay more attention to the system state variables with greater sensitivity to intermittent network attacks and parameter perturbations.

[0111] In this step, the characteristic data set of each subsystem collected in the second state: [ΔP AE,p ,ΔP FC,p ,ΔP DEG,p ,ΔP BESS,p , ΔP c,p ,Δf p ] T Input into the intermittent network attack prediction model to determine the output of the hidden layer neurons of the intermittent network attack prediction model. The hidden layer neurons are as follows: Among them, h j,m represents the output of the mth hidden layer neuron in the intermittent network attack prediction model, ρ j represents the intermittent network attack prediction weight matrix or parameter perturbation prediction weight matrix, c j,m =[c j,m,1 ,c j,m,2 ,c j,m,3 ,c j,m,4 ,c j,m,5 ,c j,m,6 ] T , c j,m represents the center point vector value of the mth hidden layer neuron in the intermittent network attack prediction model, b j,m represents the width of the Gaussian basis function of the mth hidden layer neuron in the intermittent network attack prediction model, and b j,m >0, subscript m=1, 2, …, M respectively represent the mth hidden layer neuron of the intermittent network attack prediction model, and M represents the total number of hidden layer neurons of the intermittent network attack prediction model.

[0112] When j is equal to 1, the output of the hidden layer neurons of the intermittent network attack prediction model can be obtained as:

[0113] Among them, h 1,m represents the output of the mth hidden layer neuron of the intermittent network attack prediction model. The intermittent network attack prediction model is constructed by the intermittent network attack estimation weight matrix. 1,m Represents the center point vector value of the mth hidden layer neuron in the intermittent network attack prediction model: c1,m =[c 1,m,1 ,c 1,m,2 ,c 1,m,3 ,c 1,m,4 ,c 1,m,5 ,c 1,m,6 ] T , b 1,m represents the width of the Gaussian basis function of the mth hidden layer neuron in the intermittent network attack prediction model, and b 1,m >0, subscript m=1, 2, …, M respectively represent the mth hidden layer neuron of the intermittent network attack prediction model, and M represents the total number of hidden layer neurons of the intermittent network attack prediction model.

[0114] Step 304: Determine the output of the hidden layer neuron of the parameter perturbation initial prediction model according to the parameter perturbation prediction weight matrix. This step is similar to step 303. When j is equal to 2, h 2,m represents the output of the mth hidden layer neuron of the parameter perturbation prediction model. The parameter perturbation prediction model is constructed by the parameter perturbation estimation weight matrix. 2,m Represents the center point vector value of the mth hidden layer neuron in the parameter perturbation prediction model: c 2,m =[c 2,m,1 ,c 2,m,2 ,c 2,m,3 ,c 2,m,4 ,c 2,m,5 ,c 2,m,6 ] T , b 2,m represents the width of the Gaussian basis function of the mth hidden layer neuron in the parameter perturbation prediction model, and b 2,m >0, the subscript m=1, 2, …, M respectively represent the mth hidden layer neuron of the parameter perturbation prediction model, and M represents the total number of hidden layer neurons of the parameter perturbation prediction model.

[0115] Step 305, determining the intermittent network attack initial prediction model according to the output of the hidden layer neurons of the intermittent network attack initial prediction model and the output weight of the prediction model; determining the parameter perturbation initial prediction model according to the output of the hidden layer neurons of the parameter perturbation initial prediction model and the output weight of the prediction model.

[0116] In this step, the output of the neural network prediction model is:

[0117] Among them, y 1 ,y 2 They represent the observed values ​​of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model for the system intermittent network attack η(t) and the system parameter perturbation δ(x, t), respectively. The observed values ​​are expressed as: wi,m represents the output weight of the i-th initial prediction model at the m-th layer, and the subscripts i=1 and 2 represent the prediction output weight of the intermittent network attack prediction model and the prediction output weight of the parameter perturbation prediction model, respectively. i Represents the output of the hidden layer neurons of the i-th initial prediction model.

[0118] The outputs of the hidden layer neurons of the intermittent network attack initial prediction model and the outputs of the hidden layer neurons of the parameter perturbation initial prediction model determined in step 303 and step 304 are respectively substituted into the neural network prediction model formula to construct an intermittent network attack initial prediction model and a parameter perturbation initial prediction model respectively.

[0119] After constructing the intermittent network attack initial prediction model and the parameter perturbation initial prediction model, this specification further needs to calculate the model performance of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model, so as to further fine-tune the two models. Specifically, the error indicators of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model are calculated using the following formulas:

[0120] Among them, E 1 (t), ν 1 They represent the error evaluation index and prediction error weight of the initial prediction model of intermittent network attacks, E 2 (t), ν 2 Represents the error evaluation index and prediction error weight of the initial prediction model of parameter perturbation. It represents the observed value of the intermittent network attack prediction model for the system intermittent network attack η(t), which can also be called the predicted value; It represents the observed value of the parameter perturbation prediction model for the system parameter perturbation δ(x,t), which can also be called the predicted value.

[0121] Figure 4 The flowchart of a method for iteratively updating an intermittent network attack initial prediction model and a parameter perturbation initial prediction model according to an embodiment of the present specification specifically includes the following steps:

[0122] Step 401, using a superhelical algorithm to construct a superhelical path, guiding the weights in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model to converge along the optimal direction.

[0123] Specifically, the specific formula of the super spiral algorithm is as follows:

[0124] χ i (t)=-λ|D α x(t)| ζ tanh(D α-1x(t))+χ 1,i (t);

[0125]

[0126] Among them, χ i (t) represents the superhelical global error adjustment factor, χ 1,i (t) represents the superhelical local error suppression factor, and λ and ζ are constants.

[0127] In this step, the super spiral algorithm is used to guide the weights of each layer in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model to converge rapidly along the optimal direction, thereby improving the training efficiency of the two prediction models.

[0128] Step 402: Use an exponentially dependent barrier function to constrain the outputs of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model.

[0129] In this step, the superhelical algorithm calculation formula determined in step 401 is used to construct a superhelical fractional-order constraint optimization formula, and then an exponentially dependent barrier function is constructed according to the following formula:

[0130]

[0131] k(t)=k a e -rt +k b ;

[0132] Among them, Λ(t) represents the superhelical fractional order constraint optimization function, k(t) represents the output error constraint function, and E i (t) represents the error evaluation index of the i-th prediction model, h i represents the output of the hidden layer neuron of the i-th prediction model; k a , k b , r, Ψ 1 ,Ψ 2 are constants respectively. In this specification, the exponentially dependent barrier function is used to limit the error evaluation index E i (t) Large fluctuations occur during the training process, which improves the estimation accuracy of intermittent network attacks and parameter perturbations suffered by fractional-order power systems.

[0133] Specifically, the formula for constructing a superhelical path using the superhelical algorithm is as follows:

[0134]

[0135] The superhelical algorithm is used to construct a superhelical path to guide the neural network weights w iConverge quickly along the optimal direction and use an exponentially dependent barrier function to the output of the neural network (Formula meaning previous Figure 3 As explained in the text, the neural network is constrained to prevent large fluctuations in the training process, thereby improving the estimation accuracy of intermittent network attacks and parameter perturbations suffered by the fractional-order power system. When the error evaluation indicators of the intermittent network attack prediction model and the parameter perturbation prediction model converge to the preset thresholds (e.g., 0) during the training process, it is determined that the training of the intermittent network attack prediction model and the parameter perturbation prediction model is completed, and the intermittent network attack prediction model and the parameter perturbation prediction model are constructed, that is, the large model of the fractional-order power system under network attacks described in this specification.

[0136] Figure 5 The flowchart of a method for performing intermittent network attacks and parameter perturbation prediction on each subsystem in an embodiment of this specification is shown, which specifically includes the following steps:

[0137] Step 501: Acquire characteristic data of each subsystem in the power system in a second state.

[0138] In this step, characteristic data of each subsystem in the power system that is suffering from intermittent network attacks and parameter perturbations are collected, including: power generation data of the water electrolyzer system, power generation data of the fuel cell system, power generation data of the diesel generator system, power generation data of the battery energy storage system, changes in the integral controller, and changes in the frequency of the power system.

[0139] Step 502 , input the characteristic data of each subsystem in the power system in the second state into the intermittent network attack prediction model and the parameter perturbation prediction model to obtain the prediction results for each subsystem.

[0140] In this step, the data collected in step 501 is input into the trained intermittent network attack prediction model and parameter perturbation prediction model to obtain the prediction results of the model on whether each subsystem is subject to intermittent network attack and parameter perturbation. This specification improves the estimation accuracy of intermittent network attack and parameter perturbation suffered by fractional-order power system.

[0141] like Figure 6 The figure shows a schematic diagram of the structure of a device for constructing a large model of a fractional-order power system under a network attack according to an embodiment of the present specification. In this figure, the basic structure of the device for constructing a large model of a fractional-order power system under a network attack is described. The functional units and modules therein can be implemented in software, or a general chip or a specific chip can be used to implement the construction of a large model of a fractional-order power system under a network attack. The device specifically includes:

[0142] A first construction unit 601 is used to construct a first fractional-order model of the power system according to a characteristic sample data set of each subsystem in the power system in a first state;

[0143] A second construction unit 602 is used to determine a second fractional-order model of the power system according to the intermittent network attack modeling, a characteristic sample data set of each subsystem in the second state, and the first fractional-order model of the power system, wherein the second state indicates that the power system is subjected to parameter perturbation and intermittent network attacks;

[0144] The third construction unit 603 is used to construct a fractional-order multivariate regression model of the power system in the second state according to the characteristic sample data set of each subsystem in the second state and the second fractional-order model of the power system;

[0145] A determination unit 604 is used to determine the sensitivity of each subsystem in the power system to parameter perturbations and intermittent network attacks according to the power system fractional-order multivariate regression model;

[0146] A fourth construction unit 605 is used to construct an intermittent network attack initial prediction model and a parameter perturbation initial prediction model according to the sensitivity;

[0147] The fifth construction unit 606 is used to iteratively update the weight parameters in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model according to the error evaluation index of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model, so as to construct an intermittent network attack prediction model and a parameter perturbation prediction model, which constitute a large model of the fractional-order power system under network attack.

[0148] Figure 7 The figure shows a schematic diagram of the structure of a security prediction device according to an embodiment of the present specification. The basic structure of the security prediction device is described in this figure. The functional units and modules therein can use software to implement security prediction based on a large model of a fractional-order power system under network attacks, or can use general chips or specific chips to implement security prediction. The device specifically includes:

[0149] A first acquisition unit 701 is used to acquire characteristic data of each subsystem in the power system in the second state;

[0150] The prediction result determination unit 702 is used to input the characteristic data of each subsystem in the power system in the second state into the intermittent network attack prediction model and the parameter perturbation prediction model to obtain the prediction results for each subsystem.

[0151] Figure 8The diagram shows a schematic diagram of a computer device provided by an embodiment of the present specification. The method for constructing a large model of a fractional-order power system under a network attack and for predicting security can be applied to the computer device. The computer device 802 may include one or more processors 804, such as one or more central processing units (CPUs), each of which may implement one or more hardware threads. The computer device 802 may also include any memory 806 for storing any kind of information such as code, settings, data, etc. Non-limitingly, for example, the memory 806 may include any one or more combinations of the following: any type of RAM, any type of ROM, flash memory device, hard disk, optical disk, etc. More generally, any memory may use any technology to store information. Further, any memory may provide volatile or non-volatile retention of information. Further, any memory may represent a fixed or removable component of the computer device 802. In one case, when the processor 804 executes an associated instruction stored in any memory or a combination of memories, the computer device 802 may perform any operation of the associated instruction. The computer device 802 also includes one or more drive mechanisms 808 for interacting with any storage, such as a hard disk drive mechanism, an optical disk drive mechanism, etc.

[0152] The computer device 802 may also include an input / output module 810 (I / O) for receiving various inputs (via input devices 812) and for providing various outputs (via output devices 814). A specific output mechanism may include a presentation device 816 and an associated graphical user interface (GUI) 818. In other embodiments, the input / output module 810 (I / O), the input device 812, and the output device 814 may not be included, and the computer device 802 may be used as a computer device in a network. The computer device 802 may also include one or more network interfaces 820 for exchanging data with other devices via one or more communication links 822. One or more communication buses 824 couple the components described above together.

[0153] The communication link 822 may be implemented in any manner, for example, through a local area network, a wide area network (e.g., the Internet), a point-to-point connection, etc., or any combination thereof. The communication link 822 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc. governed by any protocol or combination of protocols.

[0154] Corresponds to Figures 1 to 5 In addition to the method shown, an embodiment of the specification also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above method are executed.

[0155] The embodiment of the present specification also provides a computer-readable instruction, wherein when the processor executes the instruction, the program therein causes the processor to execute the following Figures 1 to 5 The method shown.

[0156] It should be understood that in the various embodiments of this specification, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this specification.

[0157] It should also be understood that in the embodiments of this specification, the term "and / or" is only a description of the association relationship of the associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this specification generally indicates that the associated objects before and after are in an "or" relationship.

[0158] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in this specification can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this specification.

[0159] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0160] In the several embodiments provided in this specification, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, or it can be an electrical, mechanical or other form of connection.

[0161] 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 on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of this specification.

[0162] In addition, each functional unit in each embodiment of this specification may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0163] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this specification is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of this specification. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0164] Specific embodiments are used in this specification to illustrate the principles and implementation methods of this specification. The description of the above embodiments is only used to help understand the methods and core ideas of this specification. At the same time, for those skilled in the art, according to the ideas of this specification, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on this specification.

Claims

1. A method for constructing a large-scale model of a fractional-order power system under network attacks, characterized in that: The method comprises: According to the characteristic sample data set of each subsystem in the power system in the first state, a first fractional-order model of the power system is constructed; Determine the second fractional-order model of the power system according to the intermittent network attack modeling, the characteristic sample data set of each subsystem in the second state and the first fractional-order model of the power system, wherein the second state indicates that the power system is subjected to parameter perturbation and intermittent network attacks; Constructing a fractional-order multivariate regression model of the power system in the second state according to the characteristic sample data set of each subsystem in the second state and the second fractional-order model of the power system; Determine the sensitivity of each subsystem in the power system to parameter perturbations and intermittent network attacks according to the power system fractional-order multivariate regression model; According to the sensitivity, an intermittent network attack initial prediction model and a parameter perturbation initial prediction model are respectively constructed; According to the error indicators of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model, the weight parameters in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model are iteratively updated to construct an intermittent network attack prediction model and a parameter perturbation prediction model, which constitute a large model of the fractional-order power system under network attack.

2. The method according to claim 1, characterized in that According to the characteristic sample data set of each subsystem in the power system in the first state, constructing the first fractional-order model of the power system includes: The fractional-order model of the power system is constructed by the following formula: Among them, D α represents the fractional differential operator, x(t) represents the power system state matrix, x(t)=[ΔP AE ,ΔP FC ,ΔP DEG ,ΔP BESS , ΔP c ,Δf] T , ΔP AE Indicates the change in power generation of the water electrolyzer, ΔP FC Indicates the change in fuel cell power generation, ΔP DEG Indicates the change in power generation of the diesel generator, ΔP BESS Indicates the change in power generation of the battery energy storage system, ΔP c represents the output change of the integral controller, Δf represents the frequency change of the power system, represents the power system state moment, T AE represents the time constant of the water electrolyzer, T FC represents the fuel cell time constant, T DEG Represents the diesel engine time constant; T BESS Represents the time constant of the battery energy storage system, K I represents the integral controller gain, K p Represents the frequency characteristic coefficient of the power system, T p represents the power system time constant, represents the power system control matrix; u(t) represents the control input of the power system, B u =[K AE / T AE ,K FC / T FC ,K DEG / T DEG ,K BESS / T BESS ,0,0],K AE represents the water electrolyzer gain, K FC represents the fuel cell gain, K DEG Denotes diesel engine gain, K BESS Represents the battery energy storage system gain, ΔA, ΔB u They represent the parameter perturbation matrices of the power system state matrix and control matrix, y(t) represents the power system output, and C represents the power system output matrix; Where f(x) is a function, Γ(·) is a gamma function, x is a variable, α is a fractional order, n is an integer higher than α, and ξ is the integral variable.

3. The method according to claim 2, characterized in that According to the intermittent network attack modeling, the characteristic sample data set of each subsystem in the second state and the first fractional-order model of the power system, determining the second fractional-order model of the power system includes: The second fractional-order model of the power system is constructed by the following formula: Among them, δ(x,t)=ΔAx(t)+ΔB u u(t) represents the fractional-order power system parameter perturbation, and η(t) represents the intermittent network attack related to the system state, which is determined by the intermittent network attack modeling.

4. The method according to claim 1, characterized in that: According to the power system fractional-order multivariate regression model, the sensitivity of each subsystem in the power system to parameter perturbations and intermittent network attacks is determined as follows: The regression coefficient matrix of multiple subsystems of the fractional order of the power system is estimated using the least squares method; The sum of the second and fourth elements in each regression coefficient matrix is ​​determined as the sensitivity of the subsystem to intermittent network attacks; The sum of the third and fifth elements in each regression coefficient matrix is ​​determined as the sensitivity of the subsystem to parameter perturbations.

5. The method according to claim 4, characterized in that According to the sensitivity, the intermittent network attack initial prediction model and the parameter perturbation initial prediction model are respectively constructed including: According to the sensitivity of the subsystem to intermittent network attacks, an intermittent network attack prediction weight matrix is ​​constructed; According to the sensitivity of the subsystem to parameter perturbation, a parameter perturbation estimation weight matrix is ​​constructed; Determining outputs of neurons in a hidden layer of an initial prediction model for intermittent network attacks according to the intermittent network attack estimation weight matrix; Determining the output of the hidden layer neurons of the parameter perturbation initial prediction model according to the parameter perturbation prediction weight matrix; According to the output of the hidden layer neurons of the intermittent network attack initial prediction model and the output weight of the prediction model, the intermittent network attack initial prediction model is determined; according to the output of the hidden layer neurons of the parameter perturbation initial prediction model and the output weight of the prediction model, the parameter perturbation initial prediction model is determined.

6. The method according to claim 1, characterized in that According to the sensitivity, the intermittent network attack initial prediction model and the parameter perturbation initial prediction model are respectively constructed including: The characteristic sample data set of each subsystem collected in the second state is input into the initial prediction model of intermittent network attacks to determine the output of the neurons in the hidden layer of the intermittent network attack prediction model: h j,m is the output of the mth hidden layer neuron of the intermittent network attack prediction model of the super-helical fractional-order constrained RBF neural network, ρ j represents the intermittent network attack prediction weight matrix or parameter perturbation prediction weight matrix, c j,m represents the center point vector value of the mth hidden layer neuron in the intermittent network attack prediction model of the super-helical fractional-order constrained RBF neural network, b j,m represents the width of the Gaussian basis function of the mth hidden layer neuron in the intermittent network attack prediction model of the super-helical fractional-order constrained RBF neural network, and b j,m >0, subscript m=1, 2, …, M respectively represent the mth hidden layer neuron of the super helical fractional order constrained RBF neural network intermittent network attack prediction model, and M is the total number of hidden layer neurons of the super helical fractional order constrained RBF neural network intermittent network attack prediction model.

7. The method according to claim 6, characterized in that The error index of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model is determined by the following formula: Among them, E1(t) represents the intermittent network attack prediction error evaluation index, ν1 represents the intermittent network attack prediction error weight, η(t) represents the system intermittent network attack, E2(t) represents the parameter perturbation prediction error evaluation index, ν2 represents the parameter perturbation error weight; δ(x,t) represents the system parameter perturbation, They represent the observed values ​​of η(t) and δ(x, t) of the initial prediction model of intermittent network attacks and the initial prediction model of parameter perturbation respectively.

8. The method according to claim 1, characterized in that Iteratively updating the weight parameters in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model includes: The super-helical path is constructed by using the super-helical algorithm to guide the weights in the initial prediction model of intermittent network attacks and the initial prediction model of parameter perturbation to converge along the optimal direction; An exponentially dependent barrier function is used to constrain the outputs of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model.

9. A security prediction method, characterized in that: The method is applied to a large-scale fractional-order power system model under network attack according to any one of claims 1 to 8, comprising: Acquire characteristic data of each subsystem in the power system in the second state; The characteristic data of each subsystem in the power system in the second state is input into the intermittent network attack prediction model and the parameter perturbation prediction model to obtain the prediction results of each subsystem.

10. A device for constructing a large-scale model of a fractional-order power system under network attacks, characterized in that: The device comprises: A first construction unit is used to construct a first fractional-order model of the power system according to a characteristic sample data set of each subsystem in the power system in a first state; A second construction unit is used to determine a second fractional-order model of the power system according to the intermittent network attack modeling, a characteristic sample data set of each subsystem in the second state, and the first fractional-order model of the power system, wherein the second state indicates that the power system is subjected to parameter perturbation and intermittent network attacks; A third construction unit is used to construct a fractional-order multivariate regression model of the power system in the second state according to the characteristic sample data set of each subsystem in the second state and the second fractional-order model of the power system; A determination unit, used for determining the sensitivity of each subsystem in the power system to parameter perturbations and intermittent network attacks according to the power system fractional-order multivariate regression model; A fourth construction unit is used to construct an intermittent network attack initial prediction model and a parameter perturbation initial prediction model respectively according to the sensitivity; The fifth construction unit is used to iteratively update the weight parameters in the intermittent network attack initial prediction model and the parameter perturbation initial prediction model according to the error indicators of the intermittent network attack initial prediction model and the parameter perturbation initial prediction model, so as to construct an intermittent network attack prediction model and a parameter perturbation prediction model, which constitute a large model of the fractional-order power system under network attack.

11. A safety prediction device, characterized in that: The device comprises: A first acquisition unit, used to acquire characteristic data of each subsystem in the power system in the second state; The prediction result determination unit is used to input the characteristic data of each subsystem in the power system in the second state into the intermittent network attack prediction model and the parameter perturbation prediction model to obtain the prediction results for each subsystem.

12. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 9 is implemented.

13. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.

Images