Generation method and device for digital twinning of power grid, terminal equipment and storage medium

Abstract

The invention discloses a power grid digital twinning generation method and device, terminal equipment and a storage medium. The method comprises the following steps: constructing an association degree topological graph through power grid data with an association relationship, searching a shortest path in the association degree topological graph, taking the power grid data on the shortest path as indwelling data, and taking other data as candidate data. Due to the fact that the digital twinning process is complex, the step of twinning mapping is skipped from indwelling data, corresponding filling data are obtained through a data filling method, twinning data are still generated from candidate data according to the step of twinning mapping, and then digital twinning of a power grid is constructed according to the filling data and the twinning data. According to the method, digital twinning of the indwelling data is omitted, the number of data exchange times of twinning mapping is reduced, and the performance of a twinning system is improved.

Description

Technical Field

[0001] The present invention relates to the field of smart grid technology, and in particular to a method, device, terminal equipment and storage medium for generating a digital twin of a power grid. Background Art

[0002] With the transformation and upgrading of the energy system, the construction of smart grids is accelerating around the world. The digitization and informatization of power grids have become important means for power companies to achieve intelligent management and refined operation and maintenance. At the same time, digital twins, as a cutting-edge digital technology, have also received great attention from the power industry. Digital twins achieve virtual-real mapping and virtual miniatures of physical assets through data, models, algorithms and computing platforms. At present, power grids involve multiple scenarios such as substations and transmission and distribution lines, and the data types are complex. When all real-time data collected by sensors in the physical world are mapped to digital twins in the digital twin mapping mechanism, the interaction between virtual models and physical entities requires frequent exchange of large amounts of data, resulting in high system pressure and affecting the performance of the power grid digital twin system. Summary of the invention

[0003] The present invention provides a method, device, terminal equipment and storage medium for generating a digital twin of a power grid to solve the technical problem in the prior art that mapping all real-time data to a digital twin requires frequent interaction of large amounts of data, resulting in damage to system performance.

[0004] In order to solve the above technical problems, an embodiment of the present invention provides a method for generating a digital twin of a power grid, comprising:

[0005] Obtain power grid data;

[0006] All power grid data that are associated with other power grid data are stored in a first associated data set, and the remaining power grid data are stored in a non-associated data set;

[0007] According to the first associated data set, a correlation topology graph is constructed; wherein the nodes of the correlation topology graph are the power grid data in the first associated data set, and the edge weights between the nodes are negatively correlated with the correlation between the power grid data corresponding to the nodes;

[0008] Determine the shortest path between two preset nodes in the association topology graph, and use the power grid data corresponding to the nodes on the shortest path as retained data, and use the power grid data other than the retained data in the first association data set as candidate data;

[0009] Performing twin mapping on the power grid data in the non-associated data set and the candidate data to obtain twin data;

[0010] Performing data filling on the first associated data set having only candidate data to generate filled data;

[0011] A digital twin of the power grid is constructed based on the twin data and the filling data.

[0012] As a preferred solution, constructing a correlation topology map according to the first correlation data set includes:

[0013] Calculating the first correlation degree between the power grid data having correlation relationship in the first correlation data set according to the correlation degree calculation formula;

[0014] Calculating edge weights between nodes corresponding to the power grid data according to a first degree of association between the power grid data in the first associated data set;

[0015] The association degree topological graph is generated by taking the power grid data in the first association data set as nodes and the first association degree between the power grid data as edge weights.

[0016] As a preferred solution, the correlation calculation formula is:

[0017]

[0018] X i ={x i (t), t=1,...,M};

[0019] X j ={x j (t), t=1,...,M};

[0020] In the formula, X i represents the i-th power grid data in the first associated data set; X j represents the jth power grid data in the first associated data set; x i (t) represents the power grid data X i The time data of the tth moment in x j (t) represents the power grid data X j The time data of the tth moment in the grid; M represents the number of time data corresponding to each power grid data; R ij Represents power grid data X i and power grid data X i The correlation between i,j (m1,m2) represents the power grid data X i The time data and power grid data X at the m1th moment j The correlation between the data at the m2th moment in ; Cov represents covariance; Var represents variance.

[0021] As a preferred solution, the calculation formula of the edge weight is:

[0022]

[0023] In the formula, ω ij Represents power grid data X i and power grid data X i The edge weights between corresponding nodes; k is a preset constant.

[0024] As a preferred solution, after generating the filling data, the method further includes:

[0025] Replace the retained data in the first associated data set with the filling data to obtain a second associated data set;

[0026] Calculating the second correlation degree between the power grid data having correlation relationship in the second correlation data set according to the correlation degree calculation formula;

[0027] Calculate the first correlation degree average value of all first correlation degrees and the second correlation degree average value of all second correlation degrees;

[0028] Calculating a correlation error according to the first correlation average value and the second correlation average value;

[0029] Determine whether the correlation error is less than a preset error threshold. If so, continue to execute subsequent operations. If not, construct a digital twin of the power grid based on the twin data and the retained data, and stop executing subsequent operations.

[0030] As a preferred solution, performing data filling on the first associated data set having only candidate data to generate filled data includes:

[0031] Obtaining historical data corresponding to each type of retained data; wherein the historical data is power grid data of a historical period before the current moment;

[0032] Inputting the historical data corresponding to each type of retained data into the first prediction model corresponding to the retained data, so that the prediction model predicts the retained data according to the historical data to obtain first prediction data;

[0033] Using the first predicted data corresponding to all the retained data as filling data;

[0034] The training process of the first prediction model corresponding to each type of retained data includes:

[0035] For each type of retained data, a plurality of first sample data corresponding to the retained data is obtained; wherein each of the first sample data includes: first historical data of a historical period and second historical data at a moment after the historical period;

[0036] For each of the first sample data, the first historical data is input into the first model corresponding to the retained data, so that the first model predicts the power grid data at the next moment corresponding to the first historical data according to the first historical data, and obtains a first historical data prediction result; a first loss value is calculated according to the first historical data prediction result and the second historical data, and the first model is adjusted according to the first loss value;

[0037] The adjusted first model is used as the first prediction model.

[0038] As a preferred solution, performing data filling on the first associated data set having only candidate data to generate filled data includes:

[0039] Inputting the candidate data into a second prediction model so that the second prediction model predicts the retained data according to the candidate data to obtain second prediction data;

[0040] Using the second predicted data as filling data;

[0041] The training process of the second prediction model includes:

[0042] Acquire a plurality of second sample data; wherein each of the second sample data includes: first power grid data corresponding to the candidate data at a certain historical moment and second power grid data corresponding to the retained data;

[0043] For each of the second sample data, the first power grid data is input into the second model, so that the second model predicts the retained data corresponding to the first power grid data according to the first power grid data to obtain the first power grid data prediction result; according to the first power grid data prediction result and the second power grid data, a second loss value is calculated, and the second model is adjusted according to the second loss value;

[0044] The adjusted second model is used as the second prediction model.

[0045] On the basis of the above-mentioned embodiment, another embodiment of the present invention provides a device for generating a digital twin of a power grid, comprising: a data acquisition module, a retention and candidate data determination module, a twin mapping module and a digital twin supplement module;

[0046] The data acquisition module is used to acquire power grid data;

[0047] The retained and candidate data determination module is used to store all power grid data that are associated with other power grid data into a first associated data set, and the remaining power grid data into a non-associated data set; construct an association topology map based on the first associated data set; wherein the nodes of the association topology map are the power grid data in the first associated data set, and the edge weights between the nodes are negatively correlated with the association between the power grid data corresponding to the nodes; determine the shortest path between two preset nodes in the association topology map, and use the power grid data corresponding to the nodes on the shortest path as retained data, and use the power grid data other than the retained data in the first associated data set as candidate data;

[0048] The twin mapping module is used to perform twin mapping on the power grid data in the non-associated data set and the candidate data to obtain twin data;

[0049] The digital twin supplement module is used to fill in data for a first associated data set that only has candidate data to generate filled data; and to construct a digital twin of the power grid based on the twin data and the filled data.

[0050] Based on the above embodiments, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, the method for generating a digital twin of the power grid described in the above invention embodiments is implemented.

[0051] Based on the above embodiments, another embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the method for generating a digital twin of the power grid described in the above invention embodiment.

[0052] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0053] The present invention constructs a correlation topology map through the power grid data with an associated relationship, finds the shortest path in the correlation topology map, uses the power grid data on the shortest path as retained data, and other data as candidate data. Since the process of digital twins is relatively complicated, the present invention skips the twin mapping step for the retained data, uses the data filling method to obtain the corresponding filling data, and still generates twin data for the candidate data according to the twin mapping step, and then constructs the digital twin of the power grid based on the filling data and the twin data. The present invention omits the digital twin of the retained data, reduces the number of data exchanges in the twin mapping, and improves the performance of the twin system. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1It is a flowchart of a method for generating a digital twin of a power grid provided by an embodiment of the present invention;

[0055] Figure 2 It is a structural schematic diagram of a power grid digital twin generation device provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0057] Embodiment 1

[0058] Please refer to Figure 1 , a method for generating a digital twin of a power grid provided by an embodiment of the present invention, comprising:

[0059] S1. Obtain power grid data.

[0060] It should be noted that power grid data refers to data that needs to be mapped when generating the digital twin of the power grid, including but not limited to: power grid physical entity data, power grid equipment status and parameter data, power flow data, load data, equipment health data and environmental monitoring data. Among them, power grid physical entity data refers to the location and connection method of physical entities such as substations, transmission lines, distribution lines and transformers, and power flow data refers to current and voltage data.

[0061] S2. All power grid data that are associated with other power grid data are stored in a first associated data set, and the remaining power grid data are stored in a non-associated data set.

[0062] It should be noted that the association relationship of power grid data is preset artificially based on actual conditions. If the change of data A will cause the change of data B, it is considered that there is an association relationship between data A and data B. For example, environmental monitoring data will affect equipment health data, so there is an association relationship between environmental monitoring data and equipment health data.

[0063] In step S2, all the power grid data that are associated with other power grid data are stored in the first associated data set, and the rest of the power grid data are stored in the non-associated data set. That is, as long as a certain power grid data is associated with any other power grid data, the power grid data is stored in the first associated data set, and if a certain data is not associated with any other power grid data, the power grid data is stored in the non-associated data set.

[0064] S3. Construct a correlation topology map based on the first correlation data set; wherein the nodes of the correlation topology map are the power grid data in the first correlation data set, and the edge weights between the nodes are negatively correlated with the correlation between the power grid data corresponding to the nodes.

[0065] It should be noted that the correlation degree is used to characterize the degree of correlation between power grid data. The greater the correlation degree, the higher the degree of correlation.

[0066] In step S3, an undirected connected graph is constructed according to the first associated data set as an association topology graph. The power grid data is taken as nodes, and there are edges between the nodes corresponding to the power grid data with associated relationships, and there are no edges between the nodes corresponding to the power grid data without associated relationships. The degree of association between the power grid data is negatively correlated with the edge weights between the corresponding nodes, that is, the greater the degree of association, the smaller the edge weight corresponding to the power grid data.

[0067] In a preferred embodiment, constructing a correlation topology map according to the first correlation data set includes:

[0068] Calculating the first correlation degree between the power grid data having correlation relationship in the first correlation data set according to the correlation degree calculation formula;

[0069] Calculating edge weights between nodes corresponding to the power grid data according to a first degree of association between the power grid data in the first associated data set;

[0070] The association degree topological graph is generated by taking the power grid data in the first association data set as nodes and the first association degree between the power grid data as edge weights.

[0071] In a preferred embodiment, the correlation calculation formula is:

[0072]

[0073] X i ={x i (t), t=1,...,M};

[0074] X j ={x j (t), t=1,...,M};

[0075] In the formula, X i represents the i-th power grid data in the first associated data set; X j represents the jth power grid data in the first associated data set; x i (t) represents the power grid data X i The time data of the tth moment in x j (t) represents the power grid data X jThe time data of the tth moment in the grid; M represents the number of time data corresponding to each power grid data; R ij Represents power grid data X i and power grid data X i The correlation between i,j (m1,m2) represents the power grid data X i The time data and power grid data X at the m1th moment j The correlation between the data at the m2th moment in ; Cov represents covariance; Var represents variance.

[0076] It should be noted that when calculating the correlation, some non-quantitative data, such as power grid equipment status data, can be analyzed and processed using nlp text quantification tools to convert data text that is not a quantitative indicator from a qualitative description into a quantitative indicator. The power grid data collected by the present invention is data of the same period. When calculating the correlation, each power grid data is divided based on the time series to obtain the moment data of all moments in the period. The moment data refers to the power grid data at a certain moment. The period contains a total of M moments, and each power grid data is divided to obtain M moment data.

[0077] In this embodiment, the i-th power grid data X in the first associated data set is calculated. i and the jth power grid data X j The correlation between them, first, according to the divided X i and X j Time data: X i ={x i (t), t=1,...,M} and X j ={x j (t), t=1,...,M}, calculate X i and X j The correlation between all two-time data is r i,j (m1,m2),m1=1...M,m2=1...M,getX i and X j The correlation matrix between:

[0078]

[0079] Then, calculate the mean of the correlation matrix: Get X i and X j The degree of correlation between them.

[0080] It should also be noted that the correlation degree calculated based on the acquired power grid data is the first correlation degree, which is distinguished from the second correlation degree recalculated later.

[0081] In a preferred embodiment, the edge weight is calculated as follows:

[0082]

[0083] In the formula, ω ij Represents power grid data X i and power grid data X i The edge weights between corresponding nodes; k is a preset constant.

[0084] It should be noted that the edge weight is negatively correlated with the association degree. In this embodiment, an inverse formula is used to characterize the relationship between the two.

[0085] S4. Determine the shortest path between two preset nodes in the association topology graph, and use the power grid data corresponding to the nodes on the shortest path as retained data, and use the power grid data other than the retained data in the first association data set as candidate data.

[0086] In step S4, two nodes corresponding to the power grid data are preset as the starting point and the end point, and the ant colony algorithm is used to determine the shortest path between the starting point and the end point.

[0087] First, according to the starting point and the end point, the parameters of the ant colony algorithm are initialized and set, and a preset number of ants are placed at the starting point; then each ant selects the next node according to the probability of the pheromone concentration and the heuristic factor. The ants move in the planar topological graph structure and gradually build a path. After the ants complete the path, the pheromone concentration is updated according to the path length. The process of path selection and pheromone concentration update is repeated until the predetermined number of iterations is reached; in all iterations, the shortest path found by each ant is recorded, and the shortest path among all paths is selected as the optimal solution and output.

[0088] Each ant selects the next node based on the probability of pheromone concentration and heuristic factor. The probability P(a,b) of an ant going from node a to node b is expressed as:

[0089]

[0090] In the formula, α and β represent the relative importance of pheromone heuristic factor and expected heuristic factor respectively; τ(a,b) represents the number of pheromones in the node path (a,b); μ(a,b) represents the heuristic factor for path node selection; S next Represents the set from node a to the next node.

[0091] When all ants complete a path traversal, the pheromones on each path are updated. The pheromone update is expressed as:

[0092] τ ab (t+n)=(1-ρ)·τab (t)+Δτ ab ;

[0093]

[0094] Where ρ represents the evaporation coefficient of pheromone on the path, ρ∈[0,1]; 1-ρ represents the persistence coefficient of pheromone; Δτ ab represents the pheromone increment on the path (a, b); z represents the number of ants; Δτ ab k represents the amount of pheromone left by the kth ant on the path (a, b) in this iteration; Q represents a positive constant; L k It represents the length of the path traveled by the kth ant in this traversal.

[0095] S5. Perform twin mapping on the power grid data in the non-associated data set and the candidate data to obtain twin data.

[0096] In step S5, a data mapping model is used to perform digital twin mapping on the power grid data and candidate data in the non-associated data set to obtain twin data.

[0097] S6. Perform data filling on the first associated data set that only has candidate data to generate filled data.

[0098] In step S6, the retained data in the first associated data set is removed, and the first associated data set containing only the candidate data is filled to generate filled data.

[0099] It should be noted that the fill data can be understood as new retained data, which is used to replace the original retained data that has been deleted.

[0100] In a preferred embodiment, after generating the filling data, the method further includes:

[0101] Replace the retained data in the first associated data set with the filling data to obtain a second associated data set;

[0102] Calculating the second correlation degree between the power grid data having correlation relationship in the second correlation data set according to the correlation degree calculation formula;

[0103] Calculate the first correlation degree average value of all first correlation degrees and the second correlation degree average value of all second correlation degrees;

[0104] Calculating a correlation error according to the first correlation average value and the second correlation average value;

[0105] Determine whether the correlation error is less than a preset error threshold. If so, continue to execute subsequent operations. If not, construct a digital twin of the power grid based on the twin data and the retained data, and stop executing subsequent operations.

[0106] In this embodiment, after the filling data is generated, the retained data in the first associated data set is replaced with the filling data to obtain a second associated data set. The second associated data set is composed of candidate data and filling data. It is also necessary to calculate the second degree of association between the power grid data with associated relationships in the second associated data set according to the correlation calculation formula. It should be noted that the filling data inherits the association relationship of the retained data. According to whether the correlation error between the first correlation degree and the second correlation degree is less than the error threshold, it is determined whether the generated filling data can be used to replace the retained data. If it is less than the error threshold, it means that the filling data can replace the retained data, and the subsequent operations are continued. If it is not less than the error threshold, it means that the filling data cannot replace the retained data, the retained data is restored, and the digital twin of the power grid is formed according to the retained data and the twin data, and the subsequent operations are stopped.

[0107] In a preferred embodiment, performing data filling on the first associated data set having only candidate data to generate filled data includes:

[0108] Obtaining historical data corresponding to each type of retained data; wherein the historical data is power grid data of a historical period before the current moment;

[0109] Inputting the historical data corresponding to each type of retained data into the first prediction model corresponding to the retained data, so that the prediction model predicts the retained data according to the historical data to obtain first prediction data;

[0110] Using the first predicted data corresponding to all the retained data as filling data;

[0111] The training process of the first prediction model corresponding to each type of retained data includes:

[0112] For each type of retained data, a plurality of first sample data corresponding to the retained data is obtained; wherein each of the first sample data includes: first historical data of a historical period and second historical data at a moment after the historical period;

[0113] For each of the first sample data, the first historical data is input into the first model corresponding to the retained data, so that the first model predicts the power grid data at the next moment corresponding to the first historical data according to the first historical data, and obtains a first historical data prediction result; a first loss value is calculated according to the first historical data prediction result and the second historical data, and the first model is adjusted according to the first loss value;

[0114] The adjusted first model is used as the first prediction model.

[0115] In this embodiment, a prediction model based on the C3D-LSTM framework is adopted, and the C3D module is used to extract the spatiotemporal features of the historical data of the retained data. The calculation process is:

[0116]

[0117] In the formula, * represents the three-dimensional convolution operation; represents the output of the i-th layer; represents the input of the i-th layer and the output of the i-1-th layer. When i=1, X∈R ξ×S×F×c , X is the original input data, ξ, S and F represent time, space and features respectively, c represents the number of channels, and c = 1, Represents the weight parameter of the i-th convolutional layer; represents the bias parameter; f represents the activation function.

[0118] The extracted features are transformed into time series data through dimension transformation and input into the LSTM network, and the time series features used to predict the data are extracted from the sequence; the extracted time series features are passed through a fully connected network and the predicted data is output.

[0119] It should be noted that the data filling method provided in this embodiment is to predict the retained data of the current period based on the historical data corresponding to the retained data, and the current period refers to the period corresponding to the power grid data obtained in step S1.

[0120] In a preferred embodiment, performing data filling on the first associated data set having only candidate data to generate filled data includes:

[0121] Inputting the candidate data into a second prediction model so that the second prediction model predicts the retained data according to the candidate data to obtain second prediction data;

[0122] Using the second predicted data as filling data;

[0123] The training process of the second prediction model includes:

[0124] Acquire a plurality of second sample data; wherein each of the second sample data includes: first power grid data corresponding to the candidate data at a certain historical moment and second power grid data corresponding to the retained data;

[0125] For each of the second sample data, the first power grid data is input into the second model, so that the second model predicts the retained data corresponding to the first power grid data according to the first power grid data to obtain the first power grid data prediction result; according to the first power grid data prediction result and the second power grid data, a second loss value is calculated, and the second model is adjusted according to the second loss value;

[0126] The adjusted second model is used as the second prediction model.

[0127] It should be noted that the data filling method provided in this embodiment predicts the retained data based on the candidate data. Because there is an association relationship between the candidate data and the retained data, the second prediction model trained by the association relationship between the historical candidate data and the retained data can realize the function of taking the candidate data as input and predicting the corresponding retained data.

[0128] S7. Construct a digital twin of the power grid based on the twin data and the filling data.

[0129] It should be noted that the process of digital twin is relatively complicated. The retained data skips the twin mapping step, uses the data filling method to obtain the corresponding filling data, and generates twin data according to the twin mapping step for the candidate data. Then, the digital twin of the power grid is constructed based on the filling data and the twin data. The present invention eliminates the need for digital twins of retained data, reduces the number of data exchanges in twin mapping, and improves the performance of the twin system.

[0130] Embodiment 2

[0131] Please refer to Figure 2 , a power grid digital twin generation device provided by an embodiment of the present invention, comprising: a data acquisition module, a retention and candidate data determination module, a twin mapping module and a digital twin supplement module;

[0132] The data acquisition module is used to acquire power grid data;

[0133] The retained and candidate data determination module is used to store all power grid data that are associated with other power grid data into a first associated data set, and the remaining power grid data into a non-associated data set; construct an association topology map based on the first associated data set; wherein the nodes of the association topology map are the power grid data in the first associated data set, and the edge weights between the nodes are negatively correlated with the association between the power grid data corresponding to the nodes; determine the shortest path between two preset nodes in the association topology map, and use the power grid data corresponding to the nodes on the shortest path as retained data, and use the power grid data other than the retained data in the first associated data set as candidate data;

[0134] The twin mapping module is used to perform twin mapping on the power grid data in the non-associated data set and the candidate data to obtain twin data;

[0135] The digital twin supplement module is used to fill in data for a first associated data set that only has candidate data to generate filled data; and to construct a digital twin of the power grid based on the twin data and the filled data.

[0136] Embodiment 3

[0137] Accordingly, an embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements the method for generating a digital twin of a power grid as described in the above-mentioned embodiment of the invention.

[0138] Embodiment 4

[0139] Accordingly, an embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the method for generating a digital twin of a power grid described in the above-mentioned embodiment of the invention.

[0140] It should be noted that the device embodiments described above are merely schematic, 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. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art may understand and implement it without paying any creative effort.

[0141] Those skilled in the art can clearly understand that for the sake of convenience and brevity, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0142] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0143] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the device, and various interfaces and lines are used to connect various parts of the entire device.

[0144] The memory can be used to store the computer program, and the processor realizes various functions of the device by running or executing the computer program stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0145] The storage medium is a storage medium, and the computer program is stored in the storage medium. When the computer program is executed by the processor, the steps of each method embodiment described above can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer readable media do not include electric carrier signals and telecommunication signals.

[0146] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for generating a digital twin of a power grid, characterized in that: include: Obtain power grid data; All power grid data that are associated with other power grid data are stored in a first associated data set, and the remaining power grid data are stored in a non-associated data set; According to the first associated data set, a correlation topology graph is constructed; wherein the nodes of the correlation topology graph are the power grid data in the first associated data set, and the edge weights between the nodes are negatively correlated with the correlation between the power grid data corresponding to the nodes; Determine the shortest path between two preset nodes in the association topology graph, and use the power grid data corresponding to the nodes on the shortest path as retained data, and use the power grid data other than the retained data in the first association data set as candidate data; Performing twin mapping on the power grid data in the non-associated data set and the candidate data to obtain twin data; Performing data filling on the first associated data set having only candidate data to generate filled data; A digital twin of the power grid is constructed based on the twin data and the filling data.

2. The method for generating a digital twin of a power grid according to claim 1, characterized in that: The step of constructing a correlation topology graph according to the first correlation data set includes: Calculating the first correlation degree between the power grid data having correlation relationship in the first correlation data set according to the correlation degree calculation formula; Calculating edge weights between nodes corresponding to the power grid data according to a first degree of association between the power grid data in the first associated data set; The association degree topological graph is generated by taking the power grid data in the first association data set as nodes and the first association degree between the power grid data as edge weights.

3. The method for generating a digital twin of a power grid according to claim 2, characterized in that: The correlation calculation formula is: X i ={x i (t),t=1,...,M}; X j ={x j (t),t=1,...,M}; Where, X i represents the i-th power grid data in the first associated data set; X j represents the jth power grid data in the first associated data set; x i (t) represents the power grid data X i The time data of the tth moment in x j (t) represents the power grid data X j The time data of the tth moment in the grid; M represents the number of time data corresponding to each power grid data; R ij Represents power grid data X i and power grid data X i The correlation between i,j (m1,m2) represents the power grid data X i The time data and power grid data X at the m1th moment j The correlation between the data at the m2th moment in ; Cov represents covariance; Var represents variance.

4. The method for generating a digital twin of a power grid according to claim 2, characterized in that: The calculation formula of the edge weight is: In the formula, ω ij Represents power grid data X i and power grid data X i The edge weights between corresponding nodes; k is a preset constant.

5. The method for generating a digital twin of a power grid according to claim 2, characterized in that: After generating the fill data, also include: Replace the retained data in the first associated data set with the filling data to obtain a second associated data set; Calculating the second correlation degree between the power grid data having correlation relationship in the second correlation data set according to the correlation degree calculation formula; Calculate the first correlation degree average value of all first correlation degrees and the second correlation degree average value of all second correlation degrees; Calculating a correlation error according to the first correlation average value and the second correlation average value; Determine whether the correlation error is less than a preset error threshold. If so, continue to execute subsequent operations. If not, construct a digital twin of the power grid based on the twin data and the retained data, and stop executing subsequent operations.

6. The method for generating a digital twin of a power grid according to claim 1, characterized in that: The step of performing data filling on the first associated data set having only candidate data to generate filled data includes: Obtaining historical data corresponding to each type of retained data; wherein the historical data is power grid data of a historical period before the current moment; Inputting the historical data corresponding to each type of retained data into the first prediction model corresponding to the retained data, so that the prediction model predicts the retained data according to the historical data to obtain first prediction data; Using the first predicted data corresponding to all the retained data as filling data; The training process of the first prediction model corresponding to each type of retained data includes: For each type of retained data, a plurality of first sample data corresponding to the retained data is obtained; wherein each of the first sample data includes: first historical data of a historical period and second historical data at a moment after the historical period; For each of the first sample data, the first historical data is input into the first model corresponding to the retained data, so that the first model predicts the power grid data at the next moment corresponding to the first historical data according to the first historical data, and obtains a first historical data prediction result; a first loss value is calculated according to the first historical data prediction result and the second historical data, and the first model is adjusted according to the first loss value; The adjusted first model is used as the first prediction model.

7. The method for generating a digital twin of a power grid according to claim 1, characterized in that: The step of performing data filling on the first associated data set having only candidate data to generate filled data includes: Inputting the candidate data into a second prediction model so that the second prediction model predicts the retained data according to the candidate data to obtain second prediction data; Using the second predicted data as filling data; The training process of the second prediction model includes: Acquire a plurality of second sample data; wherein each of the second sample data includes: first power grid data corresponding to the candidate data at a certain historical moment and second power grid data corresponding to the retained data; For each of the second sample data, the first power grid data is input into the second model, so that the second model predicts the retained data corresponding to the first power grid data according to the first power grid data to obtain the first power grid data prediction result; according to the first power grid data prediction result and the second power grid data, a second loss value is calculated, and the second model is adjusted according to the second loss value; The adjusted second model is used as the second prediction model.

8. A device for generating a digital twin of a power grid, characterized in that: include: Data acquisition module, retention and candidate data determination module, twin mapping module and digital twin supplement module; The data acquisition module is used to acquire power grid data; The retained and candidate data determination module is used to store all power grid data that are associated with other power grid data into a first associated data set, and the remaining power grid data into a non-associated data set; construct an association topology map based on the first associated data set; wherein the nodes of the association topology map are the power grid data in the first associated data set, and the edge weights between the nodes are negatively correlated with the association between the power grid data corresponding to the nodes; determine the shortest path between two preset nodes in the association topology map, and use the power grid data corresponding to the nodes on the shortest path as retained data, and use the power grid data other than the retained data in the first associated data set as candidate data; The twin mapping module is used to perform twin mapping on the power grid data in the non-associated data set and the candidate data to obtain twin data; The digital twin supplement module is used to fill in data for a first associated data set that only has candidate data to generate filled data; and to construct a digital twin of the power grid based on the twin data and the filled data.

9. A terminal device, characterized in that: It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements the method for generating a digital twin of a power grid as described in any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein, when the computer program is running, the device where the storage medium is located is controlled to execute the method for generating a digital twin of a power grid as described in any one of claims 1 to 7.

Images