A digital twin model construction method and system of a smart grid and a medium

Abstract

This application relates to the field of digital twin technology, specifically to a method, system, and medium for constructing a digital twin model of a smart grid. The method includes: acquiring grid parameters for each unit within the grid at each time step; determining the distribution volatility of each grid parameter in each unit at each time step; determining the modal similarity of each grid parameter in each unit at each time step; determining the joint fluctuation characteristic value of each unit at each time step; determining the synergistic effect strength of each unit at each time step; determining the fault perception strength at each time step; constructing a fault prediction model; and constructing a physical virtual model and a knowledge model of the grid based on textual data of the physical structural relationships and inherent attributes of various devices in the grid, respectively, and fusing these models with the fault prediction model to obtain a digital twin model of the grid. This application can utilize the digital twin model to perform real-time grid monitoring and fault prediction of the smart grid, improving the fault prediction capability of the digital twin model of the grid.

Description

Technical Field

[0001] This application relates to the field of digital twin technology, specifically to a method, system, and medium for constructing a digital twin model of a smart grid. Background Technology

[0002] With the ever-increasing demand for electricity, traditional power grids face numerous challenges, such as high safety risks, unstable power supply, and significant environmental impact. To address these challenges, new and efficient digital smart grids have attracted widespread attention. Furthermore, digital twin technology plays a crucial role in digital smart grids.

[0003] The application of digital twins in power grids can cover all aspects of the grid, including fault diagnosis, operation management, and intelligent decision-making. By collecting real-time grid data and statistically analyzing the fault characteristics of grid operation, a digital twin model is established by fusing the grid's physical virtual model, knowledge model, and fault prediction model. This digital twin model can then be used for grid monitoring and fault prediction in smart grids. However, given the massive amount of grid data, existing technologies cannot accurately perceive all aspects of fault characteristics during grid operation, resulting in poor accuracy in fault prediction by digital twin models. Summary of the Invention

[0004] To address the aforementioned technical issues, this paper provides a method, system, and medium for constructing a digital twin model of a smart grid, thereby resolving existing problems.

[0005] The solution to the technical problem presented in this application is to provide a method, system, and medium for constructing a digital twin model of a smart grid, comprising the following steps:

[0006] In a first aspect, embodiments of this application provide a method for constructing a digital twin model of a smart grid, the method comprising the following steps:

[0007] The system acquires the grid parameters of each unit within the grid at each time point, including voltage, current, power, and frequency; and obtains each modal component of each grid parameter of each unit at each time point based on the data of each grid parameter of each unit at different times.

[0008] Based on the discreteness of different modal components of each grid parameter in each unit at each time, the distribution volatility of each grid parameter in each unit at each time is determined; based on the similarity of the changing trends of different modal components of each grid parameter in each unit at each time, the modal similarity of each grid parameter in each unit at each time is determined; based on the distribution volatility and the modal similarity, the joint fluctuation characteristic value of each unit at each time is determined.

[0009] Based on the similarity of the joint fluctuation characteristic values ​​of each unit with those of the other units at each time step, the cooperative effect strength of each unit at each time step is determined; based on the joint fluctuation characteristic values ​​and the cooperative effect strength, the fault perception strength at each time step is determined.

[0010] Based on all grid parameters of all units at all times within the grid, and combined with the fault perception intensity, a fault prediction model is constructed; based on the text data of the physical structure relationships and inherent attributes of each device in the grid, a physical virtual model and a knowledge model of the grid are constructed, and combined with the fault prediction model, the models are fused to obtain a digital twin model of the grid.

[0011] Determining the joint fluctuation characteristic value of each unit at each time step includes:

[0012] The product of the distribution volatility and the modal similarity is used as the abnormal superposition feature value of each grid parameter of each unit at each time.

[0013] The average of the superimposed abnormal characteristic values ​​of all grid parameters in each unit at each time moment is taken as the joint fluctuation characteristic value of each unit at each time moment.

[0014] The strength of the cooperative effect of each unit at each time step is determined by the following formula: ,in, For the first The unit in the first The strength of the synergistic effect at any given moment. The number of all units in the power grid. To calculate the correlation coefficient, For the first The unit in the first The joint feature sequence at time step, For the first The unit in the first A joint feature sequence at each time step, wherein the joint feature sequence is a sequence composed of the joint fluctuation feature values ​​of multiple time steps prior to each time step of each unit;

[0015] The fault perception intensity at each time point is the average of the product of the cooperative effect intensity of all units at each time point and the joint fluctuation characteristic value.

[0016] The construction of the fault prediction model includes:

[0017] The average value of each power grid parameter in all units at each time moment is used as the average data of each power grid parameter at each time moment; the average data of all power grid parameters at all times moment is used as the training set, and the fault perception intensity at all times moment is used as the training label.

[0018] Based on the training set and the training labels, the neural network model is trained, and the trained neural network model is denoted as the fault prediction model.

[0019] Preferably, obtaining each modal component of each grid parameter of each unit at each time step includes:

[0020] A signal decomposition algorithm is used to decompose the data of each grid parameter of each unit at multiple times before each time, and obtain each modal component of each grid parameter of each unit at each time.

[0021] Preferably, determining the distribution fluctuation of each grid parameter in each unit at each time step includes:

[0022] Calculate the degree of dispersion of each modal component of each grid parameter in each unit at each time step;

[0023] The mean of the dispersion of all modal components of each grid parameter in each unit at each time step is taken as the distribution fluctuation of each grid parameter in each unit at each time step.

[0024] Preferably, determining the modal similarity of each grid parameter in each unit at each time step includes:

[0025] Calculate the similarity between the first-order difference sequence of each modal component of each grid parameter in each unit at each time step and the first-order difference sequence of its previous modal component;

[0026] The mean of all similarities of each grid parameter in each unit at each time step is taken as the modal similarity of each grid parameter in each unit at each time step.

[0027] Secondly, embodiments of this application also provide a digital twin model construction system for smart grids, the system comprising:

[0028] The power grid data acquisition module acquires the power grid parameters of each unit within the power grid at each time point, including voltage, current, power, and frequency.

[0029] The fault prediction model construction module obtains each modal component of each grid parameter in each unit at each time step based on the data of each grid parameter in each unit at different times; determines the distribution volatility of each grid parameter in each unit at each time step based on the discreteness of different modal components in each unit at each time step; determines the modal similarity of each grid parameter in each unit at each time step based on the similarity of the changing trends of different modal components in each unit at each time step; determines the joint fluctuation characteristic value of each unit at each time step based on the distribution volatility and the modal similarity; determines the cooperative effect strength of each unit at each time step based on the similarity of the joint fluctuation characteristic value of each unit with the other units at each time step; determines the fault perception strength at each time step based on the joint fluctuation characteristic value and the cooperative effect strength; and constructs a fault prediction model based on the grid parameters of all units in the grid at all times step, combined with the fault perception strength.

[0030] The digital twin model construction module constructs a physical virtual model and a knowledge model of the power grid based on textual data of the physical structural relationships and inherent attributes of various devices in the power grid, and then fuses these models with the fault prediction model to obtain a digital twin model of the power grid.

[0031] Thirdly, embodiments of this application also provide a digital twin model construction medium for a smart grid, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described methods for constructing a digital twin model of a smart grid.

[0032] This application has at least the following beneficial effects:

[0033] This application determines the distribution volatility of each grid parameter in each unit at each time step based on the discreteness of different modal components of each grid parameter in each unit at each time step; it determines the modal similarity of each grid parameter in each unit at each time step based on the similarity of the changing trends of different modal components of each grid parameter in each unit at each time step; and it determines the joint fluctuation characteristic value of each unit at each time step. The beneficial effect of this is that it considers the fluctuation of each grid parameter when a grid fault occurs, reflecting the phenomenon of joint fluctuation in grid data when multiple devices in the grid fail simultaneously, enabling a more accurate and comprehensive perception of fault characteristics during grid operation. Based on the similarity of the joint fluctuation characteristic values ​​of each unit with those of other units at each time step, it determines the synergistic effect strength of each unit at each time step; and it determines the fault perception strength at each time step. Its beneficial effects lie in considering the synergistic effects between different units after a fault occurs, thereby reflecting the degree of impact of the fault on the power grid operation, so as to more accurately perceive the fault characteristics of the power grid operation in a comprehensive manner; based on the power grid parameters of all units at all times in the power grid, combined with the fault perception intensity, a fault prediction model is constructed; based on the text data of the physical structure relationship and inherent attributes of each device in the power grid, a physical virtual model and a knowledge model of the power grid are constructed, and combined with the fault prediction model, the models are fused to obtain a digital twin model of the power grid. Its beneficial effects are that it solves the problem that existing technologies cannot accurately perceive the fault characteristics of the power grid operation in a comprehensive manner, and uses the digital twin model to monitor the smart grid and predict faults in real time, thereby improving the fault prediction capability of the digital twin model of the power grid. Attached Figure Description

[0034] The following section provides a more detailed description of a digital twin model construction method for a smart grid according to the present application, with reference to the accompanying drawings.

[0035] Figure 1 A flowchart illustrating the steps of a method for constructing a digital twin model of a smart grid, as provided in this application embodiment;

[0036] Figure 2 and Figure 3 This is a schematic diagram illustrating the change in current before and after filtering, provided in an embodiment of this application.

[0037] Figure 4 A schematic diagram of the modal components of voltage at various times provided in the embodiments of this application;

[0038] Figure 5 A flowchart illustrating the steps of the method for obtaining the joint fluctuation characteristic value of each unit at each time step in the embodiments of this application;

[0039] Figure 6 A flowchart illustrating the steps of the method for obtaining fault perception intensity at various times according to embodiments of this application;

[0040] Figure 7 A block diagram of a digital twin model construction system for a smart grid provided in this application embodiment. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description, in conjunction with the accompanying drawings and implementation examples, provides a method, system, and medium for constructing a digital twin model of a smart grid proposed in this application. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0042] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0043] Please see Figure 1 The diagram illustrates a flowchart of a method for constructing a digital twin model of a smart grid according to an embodiment of this application. The method includes the following steps:

[0044] Step 1: Obtain the grid parameters of each unit in the grid at each time. The grid parameters include voltage, current, power and frequency.

[0045] A power grid is a complex operating system composed of multiple different units, including substations, transmission lines, and distribution lines. Therefore, when a fault occurs in the power system, the operating states of different units fluctuate differently. Thus, by installing voltage, current, power, and frequency sensors in each unit of the power grid, and setting the sampling frequency of the sensors to f, the system collects voltage, current, power, and frequency data from all sensors in each unit when a power grid fault occurs. All collected data is then filtered and processed, and the data is ordered to obtain the power grid parameters at each sampling time for each unit. These parameters include voltage, current, power, and frequency.

[0046] Preferably, the sampling frequency of the sensor is set to 30Hz. As another implementation method, the implementer can set it according to the actual situation. Secondly, the Kalman filter algorithm is used for filtering. The Kalman filter algorithm is a well-known technology and will not be described in detail here. As another implementation method, the implementer can use other methods of the prior art, such as mean filtering, etc. This embodiment does not impose any special restrictions on this.

[0047] Furthermore, the schematic diagram of the current change before and after filtering provided in the embodiments of this application is as follows: Figure 2 and Figure 3 As shown, where Figure 2 and Figure 3 The horizontal axis represents time, and the vertical axis represents current. Figure 2 This shows the current change before filtering. Figure 3 This shows the current change after filtering, compared to Figure 2 , Figure 3 After noise reduction, the current change is smoother, thus reducing the impact of noise on the current.

[0048] At this point, the power grid parameters for each unit within the power grid at each time point are obtained.

[0049] Step 2: Based on the discreteness of different modal components of each grid parameter in each unit at each time, determine the distribution volatility of each grid parameter in each unit at each time; based on the similarity of the changing trends of different modal components of each grid parameter in each unit at each time, determine the modal similarity of each grid parameter in each unit at each time; based on the distribution volatility and the modal similarity, determine the joint fluctuation characteristic value of each unit at each time.

[0050] Under normal operating conditions, the power grid exhibits a stable operating state at different times. However, when a fault occurs in the power grid, the operating state of each unit fluctuates significantly. Furthermore, when one device in the power grid fails, isolated fluctuations in the grid data within that unit are likely to occur; conversely, when multiple devices fail simultaneously, joint fluctuations in the grid data within that unit are likely to occur, resulting in a high degree of data instability. Based on the above analysis, in order to comprehensively perceive the fault characteristics of the power grid during operation and avoid affecting the prediction accuracy of the fault prediction model, the operating state characteristics of the power grid data of each unit in the power grid are analyzed.

[0051] First, the fluctuations of various power grid parameters at multiple acquisition times prior to each acquisition time are analyzed to determine the distributed volatility, which reflects the degree of fluctuation in the power grid data. Specifically:

[0052] A signal decomposition algorithm is used to decompose the data of each grid parameter of each unit at multiple times before each time, and to obtain each modal component of each grid parameter of each unit at each time.

[0053] Preferably, in this embodiment, variational mode decomposition (VMD) is used to decompose the modal components. The number of modes is set to 5, the penalty coefficient is 2000, and the convergence tolerance is [value missing]. Variational mode decomposition algorithm is a well-known technology and will not be described in detail here. As other implementation methods, implementers may use other methods of existing technology, such as empirical mode decomposition (EMD) algorithm, etc. This embodiment does not impose any special restrictions on this.

[0054] Furthermore, embodiments of this application provide schematic diagrams of the modal components of voltage at various times, as shown below. Figure 4 As shown in the figure, the horizontal axis represents time and the vertical axis represents voltage. M1~M5 in the figure represent the modal components at different frequency components.

[0055] Furthermore, if high abnormal fluctuations occur simultaneously in all modal components of each power grid parameter, and the locations of abnormal fluctuations in different modal components are closer, the superposition effect of abnormal fluctuations on the corresponding power grid parameter is more obvious. That is, the joint fluctuation characteristics of the power grid data within the corresponding unit are stronger. At this time, the power grid is more likely to experience the phenomenon of multiple devices failing at the same time.

[0056] Therefore, the fluctuations within the modal components are analyzed to reflect the degree of fluctuation in the power grid parameters, specifically as follows:

[0057] Calculate the degree of dispersion of each modal component of each grid parameter in each unit at each time step;

[0058] The mean of the dispersion of all modal components of each grid parameter in each unit at each time step is taken as the distribution fluctuation of each grid parameter in each unit at each time step.

[0059] Preferably, in this embodiment, the information entropy of each modal component of each grid parameter of each unit at each time is calculated; wherein, the calculation of information entropy is a well-known technique and will not be described in detail here. As other implementation methods, implementers may use other methods of the prior art, such as variance, standard deviation, coefficient of variation, etc., and this embodiment does not impose any special restrictions on this.

[0060] It should be noted that the greater the degree of dispersion, the higher the disorder of the elements within the modal components, the more obvious the fluctuation characteristics of each power grid parameter, and the greater the distribution volatility.

[0061] Furthermore, if the fluctuations of modal components at different frequencies exhibit high similarity, and the fluctuation characteristics of each grid parameter are more pronounced, then the anomaly superposition characteristics of each grid parameter at different times are higher, meaning the fault characteristics during grid operation are more obvious. Therefore, analyzing the similarity between modal components corresponding to different frequency components and determining the modal similarity is specifically as follows:

[0062] Calculate the first-order difference sequence of each modal component of each grid parameter in each unit at each time step;

[0063] Calculate the similarity between the first-order difference sequence of each modal component of each grid parameter in each unit at each time step and its previous modal component;

[0064] Preferably, in this embodiment, the reciprocal of the DTW distance between each modal component of each grid parameter of each unit at each time and its previous modal component is calculated. As another implementation, the implementer may use other methods of the prior art, such as cosine similarity, etc. This embodiment does not impose any special restrictions on this.

[0065] The mean of all similarities of each grid parameter in each unit at each time step is taken as the modal similarity of each grid parameter in each unit at each time step.

[0066] Preferably, in this embodiment, the modal similarity of each grid parameter in each unit at each time point is calculated as follows: ,in, For the first Unit 1 The first type of power grid parameter Modal similarity at time step, For the first Unit 1 The first type of power grid parameter The first moment First-order difference sequence of modal components For the first Unit 1 The first type of power grid parameter The first moment First-order difference sequence of modal components For the first Unit 1 The first type of power grid parameter The number of all modal components at time 10:00. To ensure that the value is greater than 0 and to avoid a denominator of 0, in this embodiment, The value is 0.01. As for other implementation methods, the implementer can set it according to the actual situation.

[0067] It should be noted that the greater the similarity, the more similar the fluctuations between different modal components are, and thus the greater the modal similarity.

[0068] Furthermore, based on the distribution volatility and the modal similarity, anomaly superposition feature values ​​are determined to reflect the characteristics of joint fluctuations in power grid data, specifically:

[0069] The product of the distribution volatility and the modal similarity is used as the abnormal superposition feature value of each grid parameter of each unit at each time.

[0070] It should be noted that the greater the modal similarity and the greater the distribution fluctuation, the better it reflects the abnormal superposition characteristics of power grid data when multiple devices in the power grid fail simultaneously. The more significant the joint fluctuation characteristics of the power grid data in the corresponding unit, the larger the abnormal superposition characteristic value.

[0071] Furthermore, the abnormal superposition feature value reflects the phenomenon of joint fluctuations in power grid data when multiple devices in the power grid fail simultaneously. In order to reflect the phenomenon of isolated fluctuations or joint fluctuations in power grid data within each unit, the joint fluctuation feature value is determined as follows:

[0072] The average of the superimposed abnormal characteristic values ​​of all grid parameters in each unit at each time moment is taken as the joint fluctuation characteristic value of each unit at each time moment.

[0073] It should be noted that the larger the combined fluctuation characteristic value, the more obvious the combined fluctuation characteristics are when multiple devices in the power grid fail simultaneously; the smaller the combined fluctuation characteristic value, the more obvious the isolated fluctuation characteristics are when one device in the power grid fails.

[0074] Furthermore, the flowchart of the method for obtaining the joint fluctuation characteristic value of each unit at each time step provided in the embodiments of this application is as follows: Figure 5 As shown.

[0075] Thus, the joint fluctuation characteristic values ​​of each unit at each time step are obtained.

[0076] Step 3: Based on the similarity of the joint fluctuation characteristic values ​​of each unit with the other units at each time, determine the synergistic effect strength of each unit at each time; based on the joint fluctuation characteristic values ​​and the synergistic effect strength, determine the fault perception strength at each time.

[0077] Typically, when multiple devices in a power grid fail simultaneously, synergistic effects can easily occur between different units. For example, the failure of multiple devices within a substation unit can lead to instability in the voltage rise and fall of the power supply within the substation. This instability can then affect the stability of power grid data in transmission and distribution units. In other words, the failure of multiple devices within a substation unit can generate strong synergistic effects, thereby impacting the stability of power supply in other units. Therefore, it is necessary to analyze the synergistic effects of different units at different times when a power grid operational failure occurs, and to determine the strength of these synergistic effects. Specifically:

[0078] The joint fluctuation characteristic values ​​of each unit at multiple times prior to each time are used to form a joint characteristic sequence for each unit at each time.

[0079] Calculate the correlation coefficient of the joint feature sequence of each unit at each time step with the same time steps of the other units, and take the mean of the absolute values ​​of the correlation coefficients of each unit with all other units at each time step as the synergistic strength of each unit at each time step.

[0080] Preferably, in this embodiment, the Pearson correlation coefficient of the joint feature sequence of each unit at each time point and the same time point of the other units is calculated. The calculation of the Pearson correlation coefficient is a well-known technique and will not be described in detail here. As other implementation methods, implementers can use other methods of the prior art, such as Spearman correlation coefficient, cosine similarity, etc. This embodiment does not impose any special restrictions on this.

[0081] In this embodiment, the method for calculating the cooperative effect strength of each unit at each time step is as follows: In the formula, For the first The unit in the first The strength of the synergistic effect at any given moment. The number of all units in the power grid. To calculate the correlation coefficient, For the first The unit in the first The joint feature sequence at time step, For the first The unit in the first The joint feature sequence at time step.

[0082] It should be noted that the larger the correlation coefficient, the greater the synergistic effect on other units in the power grid when multiple devices in the corresponding unit fail, and the stronger the synergistic effect.

[0083] Furthermore, in order to comprehensively perceive the fault characteristics at different times during the operation of the power grid fault, the fault perception intensity is determined based on the strength of the synergistic effect and the joint fluctuation characteristic value, so as to reflect the degree of impact of the fault on the operation of the power grid, specifically as follows:

[0084] The average of the product of the cooperative effect strength and the joint fluctuation characteristic value of all units at each time moment is taken as the fault perception strength at each time moment.

[0085] It should be noted that the intensity of synergistic effect reflects the synergistic effect generated when a fault occurs in the corresponding unit. The greater the intensity of synergistic effect, the greater the impact of the fault in the corresponding unit on the power grid. The larger the joint fluctuation characteristic value, the greater the impact of multiple devices in the power grid failing at the same time on the stable operation of the power grid. At this time, the fault characteristics of the power grid are more obvious, and the fault perception intensity is greater, that is, the fault characteristics of the power grid are stronger.

[0086] In this embodiment, the formula for calculating the fault perception intensity at each moment is: In the formula, For the first The intensity of fault perception at any given moment. For the first The unit in the first The strength of the synergistic effect at any given moment. For the first The unit in the first The joint fluctuation eigenvalue at time t. This represents the number of all units in the power grid.

[0087] Furthermore, the flowchart of the method for obtaining the fault perception intensity at various times provided in the embodiments of this application is as follows: Figure 6 As shown.

[0088] Thus, the fault perception intensity at each moment is obtained.

[0089] Step 4: Based on the grid parameters of all units at all times within the grid, and combined with the fault perception intensity, construct a fault prediction model; based on the text data of the physical structure relationships and inherent attributes of each device in the grid, construct a physical virtual model and a knowledge model of the grid, and combine them with the fault prediction model to perform model fusion to obtain a digital twin model of the grid.

[0090] Furthermore, based on the aforementioned fault perception intensity, a fault prediction model is constructed, specifically as follows:

[0091] The average value of each power grid parameter in all units at each time moment is taken as the average data of each power grid parameter at each time moment;

[0092] The average data of all power grid parameters at all times are used as the training set, and the fault perception intensity at all times is used as the training label.

[0093] Based on the training set and the training labels, the neural network model is trained, and the trained neural network model is denoted as the fault prediction model.

[0094] Preferably, in this embodiment, a Long Short Term Memory Network (LSTM) model is used for model training, wherein the activation function is the ReLU function, the optimizer is the Adam optimizer, and the loss function is the mean squared error function. The LSTM network model is a well-known technology and will not be described in detail here. As other implementation methods, implementers may use other methods of the prior art, such as recurrent neural network models, etc. This embodiment does not impose any special restrictions on this.

[0095] Furthermore, based on the aforementioned fault prediction model, a digital twin model is constructed, specifically as follows:

[0096] Based on the structural relationships between various devices in the power grid, a physical virtual model of the power grid is constructed;

[0097] A knowledge model of the power grid is constructed based on text data containing the inherent attributes of various devices in the power grid.

[0098] It should be noted that the construction of physical virtual models and knowledge models are well-known technologies, and will not be elaborated upon here.

[0099] A digital twin model is obtained by fusing the physical virtual model, knowledge model, and fault prediction model of the power grid.

[0100] It should be noted that model fusion in digital twin models is a well-known technology and will not be elaborated upon here.

[0101] Preferably, in this embodiment, a Bayesian network is used to fuse the physical virtual model, knowledge model, and fault prediction model of the power grid to obtain a joint probabilistic graphical model, which is then used as a digital twin model. Bayesian networks are a well-known technology and will not be described in detail here. It should be noted that Bayesian networks are used to represent the dependencies between different models.

[0102] Based on the output results of the physical virtual model, knowledge model and fault prediction model, as well as the actual power grid status, a joint dataset is created and divided into a joint training set and a joint test set.

[0103] The digital twin model is trained based on the joint training set, and the trained digital twin model is validated and evaluated based on the joint test set.

[0104] Therefore, by applying the digital twin model to the actual power grid, feedback data on the power grid status can be obtained in real time, and the model can be iteratively updated. In this way, the digital twin model can be used to monitor the smart grid and predict faults in real time, promptly detect abnormal situations in the power grid and issue early warnings, so as to take corresponding measures to control the stability of the power grid operation.

[0105] This application also provides a digital twin model construction system for smart grids, including:

[0106] The power grid data acquisition module acquires the power grid parameters of each unit within the power grid at each time point, including voltage, current, power, and frequency.

[0107] The fault prediction model construction module obtains each modal component of each grid parameter in each unit at each time step based on the data of each grid parameter in each unit at different times; determines the distribution volatility of each grid parameter in each unit at each time step based on the discreteness of different modal components in each unit at each time step; determines the modal similarity of each grid parameter in each unit at each time step based on the similarity of the changing trends of different modal components in each unit at each time step; determines the joint fluctuation characteristic value of each unit at each time step based on the distribution volatility and the modal similarity; determines the cooperative effect strength of each unit at each time step based on the similarity of the joint fluctuation characteristic value of each unit with the other units at each time step; determines the fault perception strength at each time step based on the joint fluctuation characteristic value and the cooperative effect strength; and constructs a fault prediction model based on the grid parameters of all units in the grid at all times step, combined with the fault perception strength.

[0108] The digital twin model construction module constructs a physical virtual model and a knowledge model of the power grid based on textual data of the physical structural relationships and inherent attributes of various devices in the power grid, and then fuses these models with the fault prediction model to obtain a digital twin model of the power grid.

[0109] Furthermore, a block diagram of a digital twin model construction system for a smart grid provided in this application embodiment is shown below. Figure 7 As shown.

[0110] Based on the same inventive concept as the above method, this application embodiment also provides a digital twin model construction medium for a smart grid, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described methods for constructing a digital twin model for a smart grid.

[0111] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0112] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0113] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this application, without departing from the content of the technical solution of this application, shall fall within the protection scope of the technical solution of this application.

Claims

1. A method for constructing a digital twin model of a smart grid, characterized in that, The method includes the following steps: The system acquires the grid parameters of each unit within the grid at each time point, including voltage, current, power, and frequency; and obtains each modal component of each grid parameter of each unit at each time point based on the data of each grid parameter of each unit at different times. Based on the discreteness of different modal components of each grid parameter in each unit at each time, the distribution volatility of each grid parameter in each unit at each time is determined; based on the similarity of the changing trends of different modal components of each grid parameter in each unit at each time, the modal similarity of each grid parameter in each unit at each time is determined; based on the distribution volatility and the modal similarity, the joint fluctuation characteristic value of each unit at each time is determined. Based on the similarity of the joint fluctuation characteristic values ​​of each unit with those of the other units at each time step, the cooperative effect strength of each unit at each time step is determined; based on the joint fluctuation characteristic values ​​and the cooperative effect strength, the fault perception strength at each time step is determined. Based on all grid parameters of all units at all times within the grid, and combined with the fault perception intensity, a fault prediction model is constructed; based on the text data of the physical structure relationships and inherent attributes of each device in the grid, a physical virtual model and a knowledge model of the grid are constructed, and combined with the fault prediction model, the models are fused to obtain a digital twin model of the grid. Determining the joint fluctuation characteristic value of each unit at each time step includes: The product of the distribution volatility and the modal similarity is used as the abnormal superposition feature value of each grid parameter of each unit at each time. The average of the superimposed abnormal characteristic values ​​of all grid parameters in each unit at each time moment is taken as the joint fluctuation characteristic value of each unit at each time moment. The strength of the cooperative effect of each unit at each time step is determined by the following formula: ,in, For the first The unit in the first The strength of the synergistic effect at any given moment. The number of all units in the power grid. To calculate the correlation coefficient, For the first The unit in the first The joint feature sequence at time step, For the first The unit in the first A joint feature sequence at each time step, wherein the joint feature sequence is a sequence composed of the joint fluctuation feature values ​​of multiple time steps prior to each time step of each unit; The fault perception intensity at each time point is the average of the product of the cooperative effect intensity of all units at each time point and the joint fluctuation characteristic value. The construction of the fault prediction model includes: The average value of each power grid parameter in all units at each time moment is used as the average data of each power grid parameter at each time moment; the average data of all power grid parameters at all times moment is used as the training set, and the fault perception intensity at all times moment is used as the training label. Based on the training set and the training labels, the neural network model is trained, and the trained neural network model is denoted as the fault prediction model.

2. The method for constructing a digital twin model of a smart grid as described in claim 1, characterized in that, The process of obtaining each modal component of each grid parameter of each unit at each time step includes: A signal decomposition algorithm is used to decompose the data of each grid parameter of each unit at multiple times before each time, and obtain each modal component of each grid parameter of each unit at each time.

3. The method for constructing a digital twin model of a smart grid as described in claim 1, characterized in that, Determining the distribution fluctuation of each grid parameter in each unit at each time step includes: Calculate the degree of dispersion of each modal component of each grid parameter in each unit at each time step; The mean of the dispersion of all modal components of each grid parameter in each unit at each time step is taken as the distribution fluctuation of each grid parameter in each unit at each time step.

4. The method for constructing a digital twin model of a smart grid as described in claim 1, characterized in that, Determining the modal similarity of each grid parameter in each unit at each time step includes: Calculate the similarity between the first-order difference sequence of each modal component of each grid parameter in each unit at each time step and the first-order difference sequence of its previous modal component; The mean of all similarities of each grid parameter in each unit at each time step is taken as the modal similarity of each grid parameter in each unit at each time step.

5. A digital twin model construction system for a smart grid, implementing the method as described in claim 1, characterized in that, The system includes: The power grid data acquisition module acquires the power grid parameters of each unit within the power grid at each time point, including voltage, current, power, and frequency. The fault prediction model construction module obtains each modal component of each grid parameter in each unit at each time step based on the data of each grid parameter in each unit at different times; determines the distribution volatility of each grid parameter in each unit at each time step based on the discreteness of different modal components in each unit at each time step; determines the modal similarity of each grid parameter in each unit at each time step based on the similarity of the changing trends of different modal components in each unit at each time step; determines the joint fluctuation characteristic value of each unit at each time step based on the distribution volatility and the modal similarity; determines the cooperative effect strength of each unit at each time step based on the similarity of the joint fluctuation characteristic value of each unit with the other units at each time step; determines the fault perception strength at each time step based on the joint fluctuation characteristic value and the cooperative effect strength; and constructs a fault prediction model based on the grid parameters of all units in the grid at all times step, combined with the fault perception strength. The digital twin model construction module constructs a physical virtual model and a knowledge model of the power grid based on textual data of the physical structural relationships and inherent attributes of various devices in the power grid, and then fuses these models with the fault prediction model to obtain a digital twin model of the power grid.

6. A digital twin model construction medium for a smart grid, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for constructing a digital twin model of a smart grid as described in any one of claims 1-4.

Images