A method and device for sensing operation status of electric power communication network

Abstract

The present invention provides an operation status perception method and device for an electric power communication network, which obtains a communication signal index value set running on a digital twin electric power communication network; inputs the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve; uses a bidirectional long short-term memory network to fit the time series information in the first fitting curve to obtain a set of index prediction values ​​at the next moment; inputs the set of index prediction values ​​at the next moment into a support vector machine classifier, and obtains the operation status category obtained by the support vector machine classifier according to the set of index prediction values ​​at the next moment. The present invention can predict the operation status category of the digital twin electric power communication network based on the data signal in the electric power communication network, improves the accuracy of network status prediction, and uses a rational fractal interpolation function to fit the data, thereby improving the stability of the model.

Description

Technical Field

[0001] The present invention relates to the field of smart grid technology, and in particular to a method and device for sensing the operating status of an electric power communication network. Background Art

[0002] The digital twin power communication network refers to a multi-disciplinary and multi-scale simulation process that makes full use of physical models, sensors, power communication network operation history and other data. It serves as a mirror of the physical product in the virtual space and reflects the entire life cycle of the corresponding physical product.

[0003] The current digital twin power communication network can extract the data features of the operation and maintenance data during the operation of the power communication network through deep learning technology, and perceive the operation status and operation situation of the network based on these data features. However, the existing neural network can only predict the network situation based on the power data of the distribution network. For the data signals in the power communication network with more data types and greater fluctuations, the network situation prediction results are not accurate enough. Summary of the invention

[0004] In view of the problems existing in the prior art, the embodiments of the present invention provide a method and device for sensing the operation status of a power communication network.

[0005] The present invention provides a method for sensing the operation status of a power communication network, comprising:

[0006] Obtain a set of communication signal indicator values ​​running on the digital twin power communication network;

[0007] Inputting the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve;

[0008] Using a bidirectional long short-term memory network to fit the time series information in the first fitting curve, to obtain a set of indicator prediction values ​​at the next moment;

[0009] The set of predicted index values ​​at the next moment is input into a support vector machine classifier, and the operating situation category obtained by the support vector machine classifier through classification according to the set of predicted index values ​​at the next moment is obtained.

[0010] According to a method for sensing the operation status of a power communication network provided by the present invention, the step of inputting the set of indicator prediction values ​​at the next moment into a support vector machine classifier, and obtaining the operation status category obtained by the support vector machine classifier by classifying the set of indicator prediction values ​​at the next moment, comprises:

[0011] Inputting the set of index prediction values ​​at the next moment into a support vector machine classifier;

[0012] Allocating different indicator weights to different types of indicator prediction values ​​at the next moment according to the types of indicator prediction values ​​in the indicator prediction value set at the next moment;

[0013] Based on the different indicator weights, a preset decision function is used to classify the indicator prediction value set at the next moment to obtain the operating status category.

[0014] According to a method for sensing the operation status of a power communication network provided by the present invention, the method uses a bidirectional long short-term memory network to fit the time series information in the first fitting curve to obtain a set of indicator prediction values ​​at the next moment, including:

[0015] Inputting the fitting value in the first fitting curve and the hidden state value at the previous moment into the forward long short-term memory network layer to obtain the current hidden state value;

[0016] Inputting the fitting value in the first fitting curve and the hidden state value at the next moment into the backward long short-term memory network layer to update the current hidden state value;

[0017] The updated hidden state value, the hidden state value at the previous moment, and the hidden state value at the next moment are taken as a combined output to obtain a set of indicator prediction values ​​at the next moment.

[0018] According to a method for sensing the operation status of a power communication network provided by the present invention, the step of inputting the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve comprises:

[0019] If the communication signal index value is within a preset interval range, the communication signal index values ​​located in different sub-intervals are fitted using different types of preset rational fractal interpolation functions; wherein the preset interval range includes the different sub-intervals;

[0020] The rational fractal curve function is:

[0021]

[0022]

[0023]

[0024]

[0025] Wherein, t is the normalized value of the communication signal indicator value; i is the interval number of the subinterval; α i and β i are shape parameters respectively.

[0026] A method for sensing the operation status of a power communication network provided by the present invention further includes:

[0027] If the volatility index of the communication signal index value set is greater than a preset threshold, the shape parameter is:

[0028] α i (x) = 0.3γ i e -x +0.06 (6)

[0029] β i (x) = 0.1 (γ i ln(x+1)+0.3) (7)

[0030] Among them, α i (x) and β i (x) are the shape parameters of the function shape; x represents the input communication signal index value, γ i is the adjustment factor.

[0031] According to a method for sensing the operation status of a power communication network provided by the present invention, the step of inputting the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve comprises:

[0032] If the communication signal index value in the communication signal index value set is less than 0, a first linear fitting function is used for fitting to obtain the first fitting curve; the first linear fitting function is:

[0033] y1=αx (8)

[0034] Wherein, y1 is the first linear fitting function; α is a constant; x is the communication signal index value;

[0035] as well as,

[0036] If the communication signal index value in the communication signal index value set is greater than 1, a sixth linear fitting function is used for fitting to obtain the first fitting curve; the sixth linear fitting function is:

[0037] y6=α(x-1)+1 (9)

[0038] Wherein, y6 is the first linear fitting function.

[0039] The present invention also provides an operation status awareness device for a power communication network, comprising:

[0040] A data acquisition module, used to obtain a set of communication signal indicator values ​​running on the digital twin power communication network;

[0041] A data fitting module, used for inputting the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve;

[0042] A prediction value calculation module, used to fit the time series information in the first fitting curve using a bidirectional long short-term memory network to obtain a set of indicator prediction values ​​at the next moment;

[0043] The operating situation category determination module is used to input the indicator prediction value set of the next moment into the support vector machine classifier, and obtain the operating situation category obtained by the support vector machine classifier according to the indicator prediction value set of the next moment.

[0044] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, an operation status awareness method of an electric power communication network as described in any one of the above-mentioned methods is implemented.

[0045] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for sensing the operating status of the power communication network as described in any one of the above is implemented.

[0046] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for sensing the operating status of the power communication network as described in any one of the above is implemented.

[0047] The method and device for sensing the operating status of the electric power communication network provided by the present invention improves the fitting ability of the non-smooth and irregular communication signal index value sets by fitting a plurality of different types of communication signal index value sets using a rational fractal neural network, and fits the timing information of the first fitting curve obtained after fitting using a bidirectional long short-term memory network to obtain the index prediction value at the next moment, which can retain both past information and future information, and enable the network to better understand the information before and after a certain moment, and uses a support vector machine to classify the above-mentioned index prediction value set at the next moment to obtain the operating status category of the digital twin electric power communication network. The present invention can predict the operating status category of the digital twin electric power communication network based on the data signal in the electric power communication network, and improves the accuracy of network status prediction. Furthermore, the present invention also uses a rational fractal interpolation function to fit the data, which improves the stability of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 It is one of the flow charts of the method for sensing the operation status of the power communication network provided by the present invention;

[0050] Figure 2 This is the second flow chart of the method for sensing the operation status of the power communication network provided by the present invention;

[0051] Figure 3 It is a structural schematic diagram of the network model provided by the present invention;

[0052] Figure 4 It is a structural schematic diagram of the operation status awareness device of the power communication network provided by the present invention;

[0053] Figure 5 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] It should be noted that in the description of the embodiments of the present invention, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "include one..." do not exclude the existence of other identical elements in the process, method, article or device including the elements. The orientation or position relationship indicated by the terms "upper", "lower" and the like is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. Unless otherwise clearly specified and limited, the terms "installed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected, or indirectly connected through an intermediate medium, or it can be a connection between two elements. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0056] The terms "first", "second", etc. in this application are used to distinguish similar objects, and are not 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 the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0057] Combine the following Figure 1-Figure 5 Specific embodiments of the present invention are described.

[0058] In one embodiment, a method for sensing the operation status of a power communication network is provided, such as Figure 1 As shown, including:

[0059] Step 101, obtaining a set of communication signal indicator values ​​running on the digital twin power communication network;

[0060] The digital twin power communication network refers to the use of digital twin technology based on the digital twin model of physical assets to display data information on the power communication network by building a visualized three-dimensional model. The communication signal index value set refers to the communication signal performance indicators running on the power communication network, such as packet loss rate, signal-to-noise ratio, etc.

[0061] Specifically, a set of communication signal indicator values ​​running on the digital twin power communication network is input into the calculation model.

[0062] Step 102: input the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve.

[0063] Among them, the rational fractal neural network module refers to a module that uses rational fractal functions to preprocess and preliminarily fit the input data.

[0064] Specifically, the communication signal index value set is input into the rational fractal neural network module, and the rational fractal neural network module uses the rational fractal function to perform interpolation fitting on the above data points to obtain a first fitting curve.

[0065] Step 103: Use a bidirectional long short-term memory network to fit the time series information in the first fitting curve to obtain a set of indicator prediction values ​​at the next moment.

[0066] The bidirectional long short-term memory network is composed of two LSTM (long short-term memory) layers. One LSTM unit processes the forward input and the other LSTM unit processes the reverse input. It can retain both past and future information, allowing the network to better understand the information before and after a certain moment.

[0067] Specifically, the data in the first fitting curve processed by the fractal neural network module is input into the bidirectional long short-term memory network. For a certain time t, its forward LSTM layer considers the input data sequence X t and the past hidden state values Then calculate the current hidden state value Similarly, its backward LSTM layer considers the future hidden state values and the current input data sequence X t , update the hidden state value at the current moment The final output of the hidden layer combines the hidden states and activation functions of the network's forward and backward layers, and combines the output results of the previous and next input sequences to obtain the final output Y, which is the predicted value of the next moment for the current data. Through the bidirectional long short-term memory network, both past and future data can affect the output of the network prediction information.

[0068] Step 104 , input the set of predicted indicators at the next moment into a support vector machine, and obtain the operating situation category obtained by the support vector machine through classification based on the set of predicted indicators at the next moment.

[0069] Specifically, the predicted values ​​of each indicator at the next moment in the operation of the digital twin power communication network are constructed into an indicator numerical vector y t , y t Input into the support vector machine classifier, and make a decision through the decision function of the support vector machine classifier. According to the network operation status, the health status of the network operation is divided into four categories: 0, 1, 2, and 3. The larger the value of the category, the healthier the network operation status. If the input data is in the 0-class space determined by the optimal classification hyperplane function (that is, the above-mentioned decision function), the output result of the classifier is marked as 0, and it is determined that the network corresponding to the data is faulty and interrupted; if the classification of the input data belongs to the 1-class space determined by the hyperplane function, the corresponding network is congested during operation; if the classification of the input data belongs to the 2-class space, the corresponding network is under normal load during operation; if the classification of the input data belongs to the 3-class space, the corresponding network is idle during operation.

[0070] The above embodiment improves the fitting ability of non-smooth and irregular communication signal index value sets by fitting a plurality of different types of communication signal index value sets using a rational fractal neural network module, and fits the timing information of the first fitting curve obtained after fitting using a bidirectional long short-term memory network to obtain the index prediction value at the next moment, which can retain both past information and future information, and enable the network to better understand the information before and after a certain moment, and uses a support vector machine to classify the above-mentioned index prediction value set at the next moment to obtain the operating status category of the digital twin power communication network. The present invention can predict the operating status category of the digital twin power communication network based on the data signal in the power communication network, and improves the accuracy of network status prediction.

[0071] In one embodiment, if Figure 2 As shown, Figure 2 A flow chart of a classification method using a support vector machine with a weighted function is provided, including:

[0072] Step 201, inputting the set of index prediction values ​​at the next moment into a support vector machine classifier;

[0073] Among them, Support Vector Machine (SVM) is a type of generalized linear classifier that performs binary classification of data in a supervised learning manner, and its decision boundary is the maximum margin hyperplane solved for learning samples; SVM can perform nonlinear classification through kernel functions.

[0074] Specifically, the support vector machine classifier can represent data in an n-dimensional space and classify the time series information fitted by the bidirectional long short-term memory network. The category of the operation status of the digital twin power communication network is represented by one-hot encoding during training. Input the feature vector X of the network operation data i , i∈[1,n). Let X i , i∈[1,n) through the kernel function K(x i ,x) calculation and converted into high-dimensional data. In this method, the kernel function K(x i ,x) selects the radial basis kernel function for training. Then, the high-order data is passed through the decision function sgn(). Using the feature vector X i , i∈[1,n) for training, and obtain the two optimal parameters of the support vector machine, that is, determine the classification hyperplane, thus ending the model building process. The set of indicator prediction values ​​at the next moment is input into the support vector machine classifier.

[0075] Step 202, assigning different indicator weights to different types of indicator prediction values ​​at the next moment according to the types of indicator prediction values ​​in the indicator prediction value set at the next moment;

[0076] Specifically, the present invention improves the classic support vector machine model, uses a weighted function for classification, and assigns different weights to different network operation indicators, so that the model can construct a decision plane according to the relative importance of each indicator. The main idea of ​​the weighted support vector machine model is to transform the original optimization problem into a quadratic programming problem, construct an optimal decision plane, assign high weights to operation indicators with important information, and similarly, assign lower weights to indicators with less important information. The improved support vector machine uses a weighted function for calculation, which reduces the model's attention to non-important indicators.

[0077] The training data set uses T = (X1, y1), (X2, y2), ... (X N ,y N )∈R N ×{1,2,3} means that after weighted processing of these data, the weighted training data set obtained is as follows:

[0078] T=(X1,y1,v1),(X2,y2,v2),…(X N ,y N ,v N ) (1)

[0079] Among them, the weights of different operating data indicators are pre-set mainly by expert scoring, and the value range of the weights satisfies ∈ <v i <1,∈ is a small enough positive number, y i is a sample Xi The corresponding label.

[0080] Step 203: Based on different indicator weights, a preset decision function is used to classify the indicator prediction value set at the next moment to obtain an operation status category.

[0081] Specifically, the numerical vector y of each indicator at the next moment of the operation process of the digital twin power communication network is t Input into the support vector machine classifier, and make a decision through the decision function. According to the network operation status, the health of the network operation is divided into four categories: 0, 1, 2, and 3. The larger the value of the category, the healthier the network operation status. If the input data is in the 0-category space determined by the optimal classification hyperplane function, the output result of the classifier is marked as 0, which means that the network corresponding to the data is faulty and interrupted; if the classification of the input data belongs to the 1-category space determined by the hyperplane function, the corresponding network is congested during operation; if the classification of the input data belongs to the 2-category space, the corresponding network is under normal load during operation; if the classification of the input data belongs to the 3-category space, the corresponding network is idle during operation.

[0082] The above embodiment, by using a support vector machine improved by a weighted function to identify and classify network operation status, can increase the model's attention to important indicators, thereby improving the reliability and accuracy of the model classification.

[0083] In one embodiment, the above-mentioned step 103 includes: inputting the fitting value in the first fitting curve and the hidden state value at the previous moment into the forward LSTM layer to obtain the current hidden state value; inputting the fitting value in the first fitting curve and the hidden state value at the next moment into the backward LSTM layer to update the current hidden state value; and using the updated hidden state value, the hidden state value at the previous moment and the hidden state value at the next moment as a combined output to obtain the indicator prediction value at the next moment.

[0084] Specifically, the data X′ in the first fitting curve processed by the fractal neural network module is input into the bidirectional long short-term memory network. For a certain time t, its forward LSTM layer considers the input data sequence X t ′ and the past hidden state value Then calculate the current hidden state value Similarly, its backward LSTM layer considers the future hidden state values and the current input data sequence X t ′, update the hidden state value at the current moment The final output of the hidden layer combines the hidden states and activation functions of the network's forward and backward layers, and combines the output results of the previous and next input sequences to obtain the final output Y, as shown in Figure 3 As shown, it is the predicted value of the next moment for the current data.

[0085] The above embodiment, through the bidirectional long short-term memory network, can make both past and future data affect the output of network prediction information, making the prediction result more accurate.

[0086] In one embodiment, the above step 102 includes: if the communication signal index value is within a preset interval, fitting the communication signal index values ​​in different sub-intervals using different types of preset rational fractal interpolation functions; wherein the preset interval range includes the different sub-intervals;

[0087] The rational fractal curve functions ω1(t,i), ω2(t,i), ω3(t,i), and ω4(t,i) are respectively:

[0088]

[0089]

[0090]

[0091]

[0092] Wherein, t is the normalized value of the communication signal indicator; α i and β i are shape parameters respectively.

[0093] Specifically, first, a rational fractal neural network module is built, and a one-dimensional vector x is input into the input layer of the rational fractal neural network module, where x satisfies x∈[0,1], and the value of t(x) is calculated based on x, and then it is connected to the preprocessing layer. The preprocessing layer contains 5 neurons, which can expand x into a five-dimensional vector [x, ω_1(θ(x)), ω_2(θ(x)), ω_3(θ(x)), ω_4(θ(x))]. The hidden layer is composed of five neurons using ReLU as the activation function, and the output layer will be obtained after processing by this layer. In the rational fractal neural network module, the neurons between the preprocessing layer and the hidden layer, and between the hidden layer and the output layer all adopt a fully connected structure. To facilitate subsequent expression, the overall fractal neural network module is represented by the function FN(〃) in the present invention.

[0094] For example, the input data is divided into six intervals with 0, 0.25, 0.5, 0.75, and 1 as the boundaries, and different operations are performed on the input data of these six intervals to obtain the response output, and these six parts are integrated into one module; the schematic diagram of the network model structure is shown in Figure 3As shown, for input values ​​in a sub-interval between the interval [0, 1], different types of rational fractal interpolation functions can be used for interpolation fitting. The rational fractal interpolation functions are as follows:

[0095] In the rational fractal neural network module used in the present invention, it is assumed that When {x0, x1, ..., x N} is a node that evenly divides the interval [0,1] and x0=0, x N =1, using the following rational fractal interpolation function as the main basis, the rational fractal curve function ω can be expressed as shown in the above formulas (1)-(4);

[0096] Among them, α i and β i They are shape parameters, which are generally constants and can control the curve shape of the rational fractal interpolation function.

[0097] In the above embodiment, different rational fractal interpolation functions are set to perform interpolation fitting on the input data, which is beneficial to the subsequent prediction of the indicator prediction value at the next moment.

[0098] In one embodiment, if the volatility index of the communication signal index value is greater than a preset threshold, the shape parameter is:

[0099] α i (x) = 0.3γ i e -x +0.06 (6)

[0100] β i (x) = 0.1 (γ i ln(x+1)+0.3) (7)

[0101] Among them, α i (x) and β i (x) are the shape parameters of the function shape; x represents the input communication signal index value, and the adjustment coefficient γ i The value of can be obtained by Calculated.

[0102] Specifically, in order to improve the fitting ability of the fractal neural network when facing large disturbances in the data, the shape parameter α i and β i Instead of using a given constant, a function shape parameter is used to dynamically control the shape of the fractal curve function. The shape parameter α i (x) and β i (x) is calculated as shown in the above formula.

[0103] In the above embodiment, when a communication signal indicator value set is greatly disturbed, the shape parameter of the function shape is used to dynamically control the shape of the fractal curve function, which is beneficial to improving the fitting ability of the fractal neural network.

[0104] In one embodiment, the step 102 further includes:

[0105] If the communication signal index value in the communication signal index value set is less than 0, a first linear fitting function is used for fitting to obtain the first fitting curve; the first linear fitting function is:

[0106] y1=αx (8)

[0107] Wherein, y1 is the first linear fitting function; α is a constant; x is the communication signal index value;

[0108] as well as,

[0109] If the communication signal index value in the communication signal index value set is greater than 1, a sixth linear fitting function is used for fitting to obtain the first fitting curve; the sixth linear fitting function is:

[0110] y6=α(x-1)+1 (9)

[0111] Wherein, y6 is the first linear fitting function.

[0112] Specifically, the fractal neural network is modularized to speed up the calculation speed as a whole. First, the input data is divided into six intervals with 0, 0.25, 0.5, 0.75, and 1 as the boundaries. Different operations are performed on the input data in the six intervals to obtain the response output, and these six parts are integrated into one module as a rational fractal neural network module.

[0113] For the data whose values ​​are less than 0, linear operation is used for processing, and the process can be expressed by the above formula (8):

[0114] y1=αx (8)

[0115] Among them, y1 is the fitting function that operates on the data with communication signal index values ​​less than 0; α is a constant term with a small value, which is similar to the LeakyLU activation function and is taken as 0.05 here. The existence of α in the fractal module is to prevent the fractal module from saturating, thereby affecting the network gradient update before the module.

[0116] like Figure 3As shown, for the data in the four intervals of [0, 0.25], [0.25, 0.5], [0.5, 0.75], and [0.75, 1], a one-step linear mapping operation is first performed, and then the data is input into the fractal neural network for data fitting. The calculations of these four intervals are shown in the following formulas:

[0117] y 2-5 =FN(x) (10)

[0118] Where FN(x) is the function of the i-th fractal neural network in the fractal module; y 2-5 are fitting functions operating on the intervals [0, 0.25], [0.25, 0.5], [0.5, 0.75], and [0.75, 1] ​​respectively;

[0119] For the data with values ​​greater than 1, linear operations are used for processing. The process can be expressed by the following formula, and the value of α is also 0.05:

[0120] y6=α(x-1)+1 (9)

[0121] In the above embodiment, linear fitting is performed on the data that is less than 0 and greater than 1 through a linear fitting function, which is beneficial to improving the accuracy of subsequent prediction results.

[0122] The operation status perception device of the electric power communication network provided by the present invention is described below. The operation status perception device of the electric power communication network described below and the operation status perception method of the electric power communication network described above can be referred to each other.

[0123] The present invention also provides an operation status perception device 400 for a power communication network, comprising: a data acquisition module 401, a data fitting module 402, a prediction value calculation module 403, and an operation status category determination module 404, wherein:

[0124] The data acquisition module 401 is used to obtain a set of communication signal indicator values ​​running on the digital twin power communication network;

[0125] A data fitting module 402, configured to input the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve;

[0126] A prediction value calculation module 403 is used to fit the time series information in the first fitting curve using a bidirectional long short-term memory network to obtain a set of indicator prediction values ​​at the next moment;

[0127] The operation status category determination module 404 is used to input the indicator prediction value set at the next moment into the support vector machine classifier, and obtain the operation status category obtained by the support vector machine classifier according to the indicator prediction value set at the next moment.

[0128] In one embodiment, the operation status category determination module 404 is further used to:

[0129] Inputting the set of index prediction values ​​at the next moment into a support vector machine classifier;

[0130] Allocating different indicator weights to different types of indicator prediction values ​​at the next moment according to the types of indicator prediction values ​​in the indicator prediction value set at the next moment;

[0131] Based on the different indicator weights, a preset decision function is used to classify the indicator prediction value set at the next moment to obtain the operating status category.

[0132] In one embodiment, the prediction value calculation module 403 is further used to:

[0133] Input the fitting value in the first fitting curve and the hidden state value at the previous moment into the forward LSTM layer to obtain the current hidden state value;

[0134] Input the fitting value in the first fitting curve and the hidden state value at the next moment into the backward LSTM layer to update the current hidden state value;

[0135] The updated hidden state value, the hidden state value at the previous moment, and the hidden state value at the next moment are used as a combined output to obtain the indicator prediction value at the next moment.

[0136] In one embodiment, the data fitting module 402 is further configured to:

[0137] If the communication signal index value is within a preset interval range, the communication signal index values ​​located in different sub-intervals are fitted using different types of preset rational fractal interpolation functions; wherein the preset interval range includes the different sub-intervals;

[0138] The rational fractal curve function is shown in the above formulas (2)-(5), which will not be described again here.

[0139] In one embodiment, the data fitting module 402 is further used to:

[0140] If the volatility index of the communication signal index value set is greater than a preset threshold, the shape parameter is as described in the above formulas (6)-(7) and will not be repeated here.

[0141] In one embodiment, the data fitting module 402 is further used to:

[0142] If the communication signal index value in the communication signal index value set is less than 0, a first linear fitting function is used for fitting to obtain the first fitting curve; the first linear fitting function is:

[0143] y1=αx

[0144] Wherein, y1 is the first linear fitting function; α is a constant; x is the communication signal index value;

[0145] as well as,

[0146] If the communication signal index value in the communication signal index value set is greater than 1, a sixth linear fitting function is used for fitting to obtain the first fitting curve; the sixth linear fitting function is:

[0147] y6=α(x-1)+1

[0148] Wherein, y6 is the first linear fitting function.

[0149] Figure 5 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 5 As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530 and a communication bus 540, wherein the processor 510, the communication interface 520 and the memory 530 complete mutual communication through the communication bus 540. The processor 810 may call the logic instructions in the memory 830 to execute the operation situation awareness method of the power communication network, the method comprising: obtaining a communication signal index value set running on the digital twin power communication network; inputting the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve; fitting the time series information in the first fitting curve using a bidirectional long short-term memory network to obtain a set of index prediction values ​​at the next moment; inputting the set of index prediction values ​​at the next moment into a support vector machine classifier, and obtaining the operation situation category obtained by the support vector machine classifier according to the set of index prediction values ​​at the next moment.

[0150] In addition, the logic instructions in the above-mentioned memory 830 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the 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 for 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 the present invention. 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.

[0151] On the other hand, the present invention also provides a computer program product, which includes a computer program, and the computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the operation status perception method of the power communication network provided by the above methods, and the method includes: obtaining a set of communication signal index values ​​running on the digital twin power communication network; inputting the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve; using a bidirectional long short-term memory network to fit the timing information in the first fitting curve to obtain a set of index prediction values ​​at the next moment; inputting the set of index prediction values ​​at the next moment into a support vector machine classifier, and obtaining the operation status category obtained by the support vector machine classifier according to the set of index prediction values ​​at the next moment.

[0152] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the operation status perception method of the power communication network provided by the above-mentioned methods, the method comprising: obtaining a set of communication signal index values ​​running on the digital twin power communication network; inputting the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve; using a bidirectional long short-term memory network to fit the timing information in the first fitting curve to obtain a set of index prediction values ​​at the next moment; inputting the set of index prediction values ​​at the next moment into a support vector machine classifier to obtain an operation status category obtained by the support vector machine classifier according to the set of index prediction values ​​at the next moment.

[0153] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0154] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0155] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for sensing the operation status of a power communication network, characterized in that: include: Obtain a communication signal indicator value set running on the digital twin power communication network; wherein the communication signal indicator value set includes at least one of a packet loss rate and a signal-to-noise ratio; Inputting the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve; specifically comprising: If the communication signal index value in the communication signal index value set is less than 0, a first linear fitting function is used for fitting to obtain the first fitting curve; the first linear fitting function is: （8） in, is the first linear fitting function; is a constant; is the communication signal index value; If the communication signal index value in the communication signal index value set is greater than 1, a sixth linear fitting function is used for fitting to obtain the first fitting curve; the sixth linear fitting function is: （9） in, is the sixth linear fitting function; is a constant; is the communication signal index value; If the communication signal index value is within different sub-intervals within the preset interval range, fitting the communication signal index values ​​within the different sub-intervals using different types of preset rational fractal interpolation functions; The different types of preset rational fractal interpolation functions are: When t is above 0 and less than 0.25, ; When t is above 0.25 and less than 0.5, ; When t is above 0.5 and less than 0.75, ; When t is above 0.75 and less than 1, ; Wherein, t is the normalized value of the communication signal indicator value; i is the interval number of the sub-interval; and are shape parameters respectively; Using a bidirectional long short-term memory network to fit the time series information in the first fitting curve, to obtain a set of indicator prediction values ​​at the next moment; The set of predicted index values ​​at the next moment is input into a support vector machine classifier, and the operating situation category obtained by the support vector machine classifier through classification according to the set of predicted index values ​​at the next moment is obtained.

2. The method for sensing the operation status of the power communication network according to claim 1, characterized in that: The step of inputting the set of predicted indicators at the next moment into a support vector machine classifier, and obtaining the operating situation category obtained by the support vector machine classifier through classification according to the set of predicted indicators at the next moment, comprises: Inputting the set of index prediction values ​​at the next moment into a support vector machine classifier; Allocating different indicator weights to different types of indicator prediction values ​​at the next moment according to the types of indicator prediction values ​​in the indicator prediction value set at the next moment; Based on the different indicator weights, a preset decision function is used to classify the indicator prediction value set at the next moment to obtain the operating status category.

3. The method for sensing the operation status of the power communication network according to claim 2, characterized in that: The step of fitting the time series information in the first fitting curve using a bidirectional long short-term memory network to obtain a set of indicator prediction values ​​at the next moment includes: Inputting the fitting value in the first fitting curve and the hidden state value at the previous moment into the forward long short-term memory network layer to obtain the current hidden state value; Inputting the fitting value in the first fitting curve and the hidden state value at the next moment into the backward long short-term memory network layer to update the current hidden state value; The updated hidden state value, the hidden state value at the previous moment, and the hidden state value at the next moment are taken as a combined output to obtain a set of indicator prediction values ​​at the next moment.

4. The method for sensing the operation status of the power communication network according to claim 1, characterized in that: When the volatility index of the communication signal index value set is less than or equal to a preset threshold, the shape parameter and are all constants.

5. The method for sensing the operation status of the electric power communication network according to claim 4, characterized in that: Also includes: If the volatility index of the communication signal index value set is greater than the preset threshold, the shape parameter is: （6） （7） in, and are the shape parameters of the function shape respectively; x represents the input communication signal index value, is the adjustment factor.

6. A device for sensing the operation status of a power communication network, characterized in that: The device comprises: A data acquisition module, used to obtain a communication signal indicator value set running on the digital twin power communication network; wherein the communication signal indicator value set includes at least one of a packet loss rate and a signal-to-noise ratio; A data fitting module is used to input the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve; the step of inputting the communication signal index value set into a rational fractal neural network module to obtain a first fitting curve specifically includes: if the communication signal index value in the communication signal index value set is less than 0, fitting is performed using a first linear fitting function to obtain the first fitting curve; the first linear fitting function is: （8） in, is the first linear fitting function; is a constant; is the communication signal index value; If the communication signal index value in the communication signal index value set is greater than 1, a sixth linear fitting function is used for fitting to obtain the first fitting curve; the sixth linear fitting function is: （9） in, is the sixth linear fitting function; is a constant; is the communication signal index value; If the communication signal index value is within different sub-intervals within the preset interval range, fitting the communication signal index values ​​within the different sub-intervals using different types of preset rational fractal interpolation functions; The different types of preset rational fractal interpolation functions are: When t is above 0 and less than 0.25, ; When t is above 0.25 and less than 0.5, ; When t is above 0.5 and less than 0.75, ; When t is above 0.75 and less than 1, ; Wherein, t is the normalized value of the communication signal indicator value; i is the interval number of the sub-interval; and are shape parameters respectively; A prediction value calculation module, used to fit the time series information in the first fitting curve using a bidirectional long short-term memory network to obtain a set of indicator prediction values ​​at the next moment; The operating situation category determination module is used to input the indicator prediction value set of the next moment into the support vector machine classifier, and obtain the operating situation category obtained by the support vector machine classifier according to the indicator prediction value set of the next moment.

7. An electronic 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 program, the method for sensing the operating status of the power communication network as described in any one of claims 1 to 5 is implemented.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for sensing the operating status of a power communication network as claimed in any one of claims 1 to 5 is implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for sensing the operating status of a power communication network as claimed in any one of claims 1 to 5 is implemented.

Images