Power system scene spatial-temporal characteristic quantity extraction method and system based on spectral clustering

Abstract

The embodiment of the invention provides a spectral clustering-based power system scene spatial-temporal characteristic quantity extraction method and system, and belongs to the technical field of Internet and cloud computing. The method comprises the following steps: constructing a space-time incidence matrix according to a geographic position of a new energy electric field and historical data of multiple time scales; determining the initial similarity of every two data points in the space-time incidence matrix to obtain an initial similarity matrix; performing dimension reduction and feature decomposition operation on the initial similarity matrix to obtain a similar feature matrix; and performing normalization operation and clustering operation on the similar feature matrix to obtain a visual power system scene spatial-temporal feature quantity. According to the method and the system, the spatial-temporal characteristic quantity of the historical data of the power system can be accurately extracted, so that the model training efficiency and the accuracy of the trained model are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet and cloud computing, and in particular to a method and system for extracting spatiotemporal feature quantities of power system scenarios based on spectral clustering. Background Art

[0002] With the advancement of the construction of new power systems, the large-scale access of high-proportion renewable energy has made the uncertainty of operation mode faced by power system planning more prominent. Faced with complex and changeable operation modes, the typical scenario operation mode of the traditional power system of "large in winter and small in winter, large in summer and small in summer, high water and low water" can no longer describe the current new power system operation mode with randomness and uncertainty. There are many kinds of characteristic quantities of power grid operation mode, involving load, wind and light, power plants, external interconnection lines, network topology, etc., and are also affected by factors such as energy storage and electricity users. These multidimensional characteristic quantities have nonlinear correlation characteristics in time and space. Their respective characteristics have certain regularities with seasonal climate changes, and they also show their own randomness superimposed on each other, resulting in complex and changeable operation modes, which are difficult to select with traditional experience. Summary of the invention

[0003] The purpose of the embodiments of the present invention is to provide a method and system for extracting spatiotemporal characteristics of power system scenarios based on spectral clustering. The method and system can accurately extract the spatiotemporal characteristics of power system historical data, thereby improving the efficiency of model training and the accuracy of the trained model.

[0004] In order to achieve the above object, an embodiment of the present invention provides a method for extracting spatiotemporal features of power system scenarios based on spectral clustering, comprising: Construct a spatiotemporal correlation matrix based on the geographical location of the renewable energy power plant and historical data at multiple time scales; Determining the initial similarity of every two data points in the spatiotemporal correlation matrix to obtain an initial similarity matrix; Performing dimension reduction and feature decomposition operations on the initial similarity matrix to obtain a similar feature matrix; The similar feature matrix is ​​normalized and clustered to obtain a visualized spatiotemporal feature of the power system scenario.

[0005] Optionally, a spatiotemporal correlation matrix is ​​constructed based on the geographical location of the new energy power plant and historical data at multiple time scales, including: Acquire the geographical location and historical data of the new energy power plant, wherein the historical data includes output data and meteorological data, and the meteorological data includes at least one of wind direction, wind speed, light intensity and temperature; The geographical location and historical data are used as labels of rows and columns of the space-time association matrix, and the values ​​of the corresponding historical data are the values ​​of the elements of the space-time association matrix, so as to obtain the space-time association matrix.

[0006] Optionally, determining the initial similarity of every two data points in the spatiotemporal correlation matrix to obtain an initial similarity matrix includes: Calculate the Manhattan distance between each two data points; Determine whether the Manhattan distance is less than or equal to a preset first threshold; When it is determined that the Manhattan distance is less than or equal to the first threshold, determining that there is a neighbor relationship between the two data points; Traversing each data point in the spatiotemporal association matrix, and selecting a first preset number of data points from the corresponding data points having the neighbor relationship as neighbor points of the current data point; An initial similarity matrix of each of the data points is determined according to the neighboring points.

[0007] Optionally, performing dimension reduction and feature decomposition operations on the initial similarity matrix to obtain a similar feature matrix includes: Traversing each of the data points to determine whether every two data points are neighboring points of each other; In the case where two data points are both neighboring points, extracting elements with a similarity of 1 from the similarity matrices corresponding to the two data points to obtain a combination matrix; Performing a symmetric transformation operation on the combined matrix to obtain a symmetric matrix; Calculate a singular matrix of a similarity matrix according to the symmetric matrix; Compression and clustering operations are performed according to the singular matrix to obtain the similar feature matrix.

[0008] Optionally, performing a symmetric transformation operation on the combined matrix to obtain a symmetric matrix includes: The symmetric transformation operation is performed according to formula (1): , (1) in, is the symmetric matrix, is the combination matrix, is the transpose of the combination matrix.

[0009] Optionally, calculating a singular matrix of the similarity matrix according to the symmetric matrix includes: The singular matrix is ​​calculated according to formula (2) to formula (9): , (2) , (3) , (4) , (5) , (6) , (7) , (8) , (9) in, is the singular matrix, is the transpose of the singular matrix, is the eigenvector of the symmetric matrix, is the feature vector The characteristic value of is a singular matrix rank, is a singular matrix The desired rank, for The desired rank of a square matrix of order, and are singular vector matrices, and are transposed of each other.

[0010] Optionally, performing compression and clustering operations according to the singular matrix to obtain the similar feature matrix includes: According to formula (10) and formula (11), the similarity feature matrix is ​​obtained: , (10) , (11) in, is the similarity feature matrix, is a singular matrix Center front The matrix consists of elements, is the singular vector matrix Center front The matrix consists of elements, is the singular vector matrix Center front The matrix consists of elements, For the matrix The transpose of For the matrix The transpose of For the matrix The transpose of .

[0011] Optionally, normalization and clustering operations are performed on the similar feature matrix to obtain visualized spatiotemporal feature quantities of the power system scenario, including: Construct the unnormalized Laplacian matrix; Construct an intermediate matrix from the unnormalized Laplacian matrix; Determining a normalized Laplace matrix according to the intermediate matrix; Construct the P matrix according to the eigenvectors corresponding to the eigenvalues ​​of the normalized Laplace matrix; Normalize the row vectors of the P matrix to obtain the Q matrix; The Q matrix is ​​visualized and the results of the visualization are clustered to obtain visualized spatiotemporal characteristics of the power system scenario.

[0012] Optionally, construct the unnormalized Laplacian matrix, including: The unnormalized Laplacian matrix is ​​determined according to formula (12): , (12) in, is the unnormalized Laplacian matrix, is the degree matrix, is the similarity feature matrix; Construct an intermediate matrix based on the unnormalized Laplacian matrix, including: The intermediate matrix is ​​determined according to formula (13): , (13) in, is the intermediate matrix, is the similarity feature matrix; Determining a normalized Laplace matrix according to the intermediate matrix includes: The normalized Laplace matrix is ​​determined according to formula (14): , (14) in, is the normalized Laplacian matrix, is the identity matrix; The P matrix is ​​constructed according to the eigenvectors corresponding to the eigenvalues ​​of the normalized Laplace matrix, including: Selecting a second preset number of eigenvectors from the eigenvectors corresponding to the eigenvalues ​​of the normalized Laplace matrix to obtain the P matrix; Normalize the row vectors of the P matrix to obtain the Q matrix, including: Calculating the Euclidean norm of the P matrix; The vectors in the P matrix are divided by the corresponding Euclidean norm to obtain the Q matrix.

[0013] On the other hand, the present invention further provides a system for extracting spatiotemporal features of power system scenarios based on spectral clustering, the system comprising a processor, and the processor is used to execute any of the methods described above.

[0014] Through the above technical scheme, the embodiment of the present invention provides a method and system for extracting spatiotemporal features of power system scenarios based on spectral clustering. The method and system realize the feature extraction of visual data of spatiotemporal features of power system scenarios by constructing a spatiotemporal association matrix and adopting dimension reduction and feature decomposition operations, normalization operations and classification operations of the similarity matrix. Compared with the prior art, the method and system provided by the present invention clusters the massive variables with both temporal and spatial features scattered in the power grid according to the similarity of their influence on the operation mode by reducing the dimension and associating massive high-dimensional data, extracts spatiotemporal features suitable for power grid scenario analysis, realizes a more accurate feature extraction process, makes the subsequent model training more efficient and the trained model more accurate.

[0015] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings: Figure 1 is a flow chart of a method for extracting spatiotemporal characteristics of power system scenarios based on spectral clustering according to an embodiment of the present invention; Figure 2 is a flow chart of a method for determining an initial similarity according to an embodiment of the present invention; Figure 3 is a flow chart of a method for obtaining a similar feature matrix according to an embodiment of the present invention; Figure 4 is a flow chart of a method of normalization operation according to one embodiment of the present invention; Figure 5 This is an example diagram of visualized spatiotemporal characteristic quantities of power system scenarios generated by clustering according to an example of the present invention; Figure 6 This is an example diagram of visualized spatiotemporal characteristic quantities of a power system scenario generated by clustering according to an example of the present invention. DETAILED DESCRIPTION

[0017] The specific implementation of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the embodiment of the present invention, and is not used to limit the embodiment of the present invention.

[0018] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application are in compliance with the relevant provisions of national laws and regulations. In the embodiments of this application, some existing solutions in the industry such as certain software, components, and models may be mentioned, which should be considered as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of this application, but it does not mean that the applicant has or will necessarily use the solution.

[0019] like Figure 1 FIG. 1 is a flow chart of a method for extracting spatiotemporal characteristics of power system scenarios based on spectral clustering according to an embodiment of the present invention. Figure 1 In the method, the method may include the following steps: In step S10, a spatiotemporal correlation matrix is ​​constructed based on the geographical location of the new energy power plant and historical data at multiple time scales; In step S11, the initial similarity between every two data points in the spatiotemporal correlation matrix is ​​determined to obtain an initial similarity matrix; In step S12, the initial similarity matrix is ​​subjected to dimensionality reduction and feature decomposition operations to obtain a similar feature matrix; In step S13, normalization and clustering operations are performed on the similar feature matrix to obtain visualized spatiotemporal feature quantities of the power system scenario.

[0020] In this Figure 1In the method shown, step S10 is used to construct a spatiotemporal association matrix based on the geographical location of the new energy power field and historical data at multiple time scales. The geographical location may include the location coordinates of each new energy power field. The historical data may include output data and meteorological data, and the meteorological data includes at least one of wind direction, wind speed, light intensity and temperature. Specifically, the specific method for constructing the spatiotemporal association matrix may be in various forms known to those skilled in the art. In one example of the present invention, the method for constructing the spatiotemporal association matrix may be to first obtain the geographical location and historical data of the new energy power field, and then use the geographical location and historical data as the labels of the rows and columns of the spatiotemporal association matrix, and the corresponding historical data values ​​are the values ​​of the elements of the spatiotemporal association matrix to obtain the spatiotemporal association matrix. In addition, when obtaining the historical data, it may be obtained by searching the historical ledger in the power system, and in order to avoid the loss of extracted features, when obtaining the ledger, the missing values ​​or fault values ​​may be processed and deleted for abnormal values ​​to prevent large errors after the historical data is integrated, and finally the obtained load data is matched and aligned according to the time and space location dimensions. In this example, the spatiotemporal association matrix can be expressed as ,in, is the number of geographical locations, is the number of time positions, and the elements in the spatiotemporal incidence matrix Indicates Geographic location in The data value at a point in time.

[0021] Step S11 can be used to determine the initial similarity of every two data points in the spatiotemporal correlation matrix to obtain an initial similarity matrix. The initial similarity matrix can correspond to each data point one by one. The method for determining the initial similarity matrix can be in various forms known to those skilled in the art. In one example of the present invention, the method for determining the initial similarity can further include: Figure 2 In the steps shown in Figure 2 In the method for determining the initial similarity, the method may include the following steps: In step S20, the Manhattan distance between every two data points is calculated respectively; In step S21, it is determined whether the Manhattan distance is less than or equal to a preset first threshold; In step S22, when it is determined that the Manhattan distance is less than or equal to the first threshold, it is determined that there is a neighbor relationship between the two data points; In step S23, each data point in the spatiotemporal correlation matrix is ​​traversed, and a first preset number of data points are selected from the corresponding data points with neighbor relationships as neighbor points of the current data point; In step S24, an initial similarity matrix of each data point is determined based on the neighboring points.

[0022] In this Figure 2 In the method shown, the Manhattan distance calculated in step S20 is used to represent the similarity between two data points. If step S21 determines that the Manhattan distance is less than the preset first threshold, it means that the similarity between the two data points is high and can be used as neighboring points (i.e., there is a neighboring relationship). However, since the number of data points with neighboring relationships for each data point is different, in order to ensure that the number of neighboring points included in the constructed initial similarity matrix is ​​consistent, step S23 is also required, that is: select a first preset number of data points from the corresponding data points with neighboring relationships as neighboring points of the current data point. Finally, when constructing the initial similarity matrix, the value of the element corresponding to the neighboring point can be set to 1, and the value of the element corresponding to the remaining data points can be set to 0.

[0023] Step S12 may be used to perform dimensionality reduction and feature decomposition operations on the initial similarity matrix to obtain a similar feature matrix. Specifically, in this embodiment, the method for obtaining the similar feature matrix may further include: Figure 3 In the steps shown in Figure 3 In the method of obtaining the similarity feature matrix, the method may include the following steps: In step S30, each data point is traversed to determine whether every two data points are neighboring points to each other; In step S31, when two data points are both neighboring points, elements with a similarity of 1 are extracted from the similarity matrices corresponding to the two data points to obtain a combination matrix; In step S32, a symmetric transformation operation is performed on the combined matrix to obtain a symmetric matrix. Specifically, in an example of the present invention, step S32 may be performed according to the following formula (1): , (1) in, is a symmetric matrix, is the combination matrix, is the transpose of the combined matrix.

[0024] In step S33, a singular matrix of the similarity matrix is ​​calculated based on the symmetric matrix. Specifically, in an example of the present invention, step S32 may be to calculate the singular matrix according to the following formulas (2) to (9): , (2) , (3) , (4) , (5) , (6) , (7) , (8) , (9) in, is a singular matrix, is the transpose of a singular matrix, is the eigenvector of a symmetric matrix, is the feature vector The characteristic value of is a singular matrix The definition of rank is, is a singular matrix rank, for The rank of a square matrix, and are singular vector matrices, and are transposed of each other.

[0025] In step S34, compression and clustering operations are performed on the singular matrix to obtain a similar feature matrix. Specifically, in one example of the present invention, step S34 may be to obtain a similar feature matrix using the following formula (10) and formula (11): , (10) , (11) in, is the similarity feature matrix, is a singular matrix Center front The matrix consists of elements, is the singular vector matrix Center front The matrix consists of elements, is the singular vector matrix Center front The matrix consists of elements, For the matrix The transpose of For the matrix The transpose of For the matrix The transpose of .

[0026] Step S13 can be used to perform normalization and clustering operations on the similar feature matrix to obtain a visualized spatiotemporal feature of the power system scenario. The specific method of the normalization operation can be in various forms known to those skilled in the art. In one example of the present invention, the normalization operation method may include: Figure 4 In the steps shown in Figure 4 In the method of normalization operation, the method may include the following steps: In step S40, an unnormalized Laplacian matrix is ​​constructed. Specifically, in this example, the unnormalized Laplacian matrix may be constructed according to the following formula (12): , (12) in, is the unnormalized Laplacian matrix, is the degree matrix, is the similarity feature matrix.

[0027] In step S41, an intermediate matrix is ​​constructed according to the unnormalized Laplacian matrix. Specifically, in this example, the intermediate matrix may be constructed according to the following formula (13): , (13) in, is the intermediate matrix, is the similarity feature matrix.

[0028] In step S42, a normalized Laplacian matrix is ​​determined according to the intermediate matrix. Specifically, in this example, the normalized Laplacian matrix may be determined according to the following formula (14): , (14) in, is the normalized Laplacian matrix, is the identity matrix.

[0029] In step S43, a P matrix is ​​constructed according to the eigenvectors corresponding to the eigenvalues ​​of the normalized Laplace matrix. Specifically, in this example, step S43 may be to select a second preset number of eigenvectors from the eigenvectors corresponding to the eigenvalues ​​of the normalized Laplace matrix to obtain the P matrix. More specifically, in this example, when using the intermediate matrix After replacing the normalized Laplace matrix (Formula (14)), its eigenvalues ​​are changed from 1 to becomes , and the corresponding eigenvector remains unchanged, because the Laplace matrix is ​​semi-positive definite, and the corresponding n real eigenvalues ​​are greater than or equal to 0, that is: , and the smallest eigenvalue is 0. Therefore, in this example, the normalized Laplace matrix The corresponding eigenvector Construct the P matrix.

[0030] In step S44, the row vectors of the P matrix are normalized to obtain the Q matrix. Specifically, in this example, the method for normalizing the row vectors of the P matrix may be to first calculate the Euclidean norm of the P matrix according to the following formula (15): , (15) in, is the Euclidean norm. Then, the eigenvector Each component of is divided by its corresponding Euclidean norm, thus obtaining the normalized vector shown in the following formula (16): : , (16) in, For vector , , , …, They are The corresponding Euclidean norm.

[0031] In addition, in order to make the spatiotemporal characteristics of the power system scene represented by the Q matrix more intuitive, in this embodiment, the Q matrix can also be clustered and visualized. Specifically, a preset clustering method can be used to take each row of the Q matrix as a data point, thereby obtaining a data set , each of which For one dimensional vector. When clustering, firstly Each cluster is treated as a separate cluster, and then the two closest clusters are found based on the Euclidean distance. The two clusters are combined into a new cluster, and the above steps are repeated until the condition of generating n clusters is met, thereby generating a complete hierarchical clustering tree. By observing the changes in the length of the branches in the hierarchical clustering tree or directly setting the number of clusters in the clustering process, the appropriate truncation point is selected to ensure that the final result can intuitively display the distribution of each cluster and the attribution of data points. The example diagram of the visualized spatiotemporal characteristics of the power system scenario generated by the corresponding clustering is as follows: Figure 5 and Figure 6 shown.

[0032] In step S45, the Q matrix is ​​visualized, and a clustering operation is performed on the result of the visualization processing to obtain visualized spatiotemporal characteristics of the power system scenario.

[0033] On the other hand, the present invention further provides a system for extracting spatiotemporal features of power system scenarios based on spectral clustering, the system comprising a processor, and the processor is used to execute any of the methods described above.

[0034] Through the above technical scheme, the embodiment of the present invention provides a method and system for extracting spatiotemporal features of power system scenarios based on spectral clustering. The method and system realize the feature extraction of visual data of spatiotemporal features of power system scenarios by constructing a spatiotemporal association matrix and adopting dimension reduction and feature decomposition operations, normalization operations and classification operations of the similarity matrix. Compared with the prior art, the method and system provided by the present invention clusters the massive variables with both temporal and spatial features scattered in the power grid according to the similarity of their influence on the operation mode by reducing the dimension and associating massive high-dimensional data, extracts spatiotemporal features suitable for power grid scenario analysis, realizes a more accurate feature extraction process, makes the subsequent model training more efficient and the trained model more accurate.

[0035] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0036] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0037] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1A function specified in one or more boxes.

[0038] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0039] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0040] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0041] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0042] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0043] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A method for extracting spatiotemporal characteristics of power system scenarios based on spectral clustering, characterized in that: The method comprises: Construct a spatiotemporal correlation matrix based on the geographical location of the renewable energy power plant and historical data at multiple time scales; Determine the initial similarity of every two data points in the spatiotemporal correlation matrix to obtain an initial similarity matrix; Performing dimension reduction and feature decomposition operations on the initial similarity matrix to obtain a similar feature matrix; The similar feature matrix is ​​normalized and clustered to obtain a visualized spatiotemporal feature of the power system scenario.

2. The method according to claim 1, characterized in that The spatiotemporal correlation matrix is ​​constructed based on the geographical location of the new energy power plant and historical data at multiple time scales, including: Acquire the geographical location and historical data of the new energy power plant, wherein the historical data includes output data and meteorological data, and the meteorological data includes at least one of wind direction, wind speed, light intensity and temperature; The geographical location and historical data are used as labels of rows and columns of the space-time association matrix, and the values ​​of the corresponding historical data are the values ​​of the elements of the space-time association matrix, so as to obtain the space-time association matrix.

3. The method according to claim 1, characterized in that Determining the initial similarity of every two data points in the spatiotemporal correlation matrix to obtain an initial similarity matrix includes: Calculate the Manhattan distance between each two data points; Determine whether the Manhattan distance is less than or equal to a preset first threshold; When it is determined that the Manhattan distance is less than or equal to the first threshold, determining that there is a neighbor relationship between the two data points; Traversing each data point in the spatiotemporal association matrix, and selecting a first preset number of data points from the corresponding data points having the neighbor relationship as neighbor points of the current data point; An initial similarity matrix of each of the data points is determined according to the neighboring points.

4. The method according to claim 1, characterized in that: Performing dimension reduction and feature decomposition operations on the initial similarity matrix to obtain a similar feature matrix, including: Traversing each of the data points to determine whether every two data points are neighboring points of each other; In the case where two data points are both neighboring points, extracting elements with a similarity of 1 from the similarity matrices corresponding to the two data points to obtain a combination matrix; Performing a symmetric transformation operation on the combined matrix to obtain a symmetric matrix; Calculate a singular matrix of a similarity matrix according to the symmetric matrix; Compression and clustering operations are performed according to the singular matrix to obtain the similar feature matrix.

5. The method according to claim 4, characterized in that Performing a symmetric transformation operation on the combined matrix to obtain a symmetric matrix includes: The symmetric transformation operation is performed according to formula (1): ，（1） in, is the symmetric matrix, is the combination matrix, is the transpose of the combination matrix.

6. The method according to claim 5, characterized in that The singular matrix of the similarity matrix is ​​calculated according to the symmetric matrix, including: The singular matrix is ​​calculated according to formula (2) to formula (9): ，（2） ，（3） ，（4） ，（5） ，（6） ，（7） ,（8） ，（9） in, is the singular matrix, is the transpose of the singular matrix, is the eigenvector of the symmetric matrix, is the feature vector The characteristic value of is a singular matrix rank, is a singular matrix The desired rank, for The rank of the square matrix is and are singular vector matrices, and are transposed of each other.

7. The method according to claim 6, characterized in that Performing compression and clustering operations according to the singular matrix to obtain the similar feature matrix includes: According to formula (10) and formula (11), the similarity feature matrix is ​​obtained: ，（10） ，（11） in, is the similarity feature matrix, is a singular matrix Center front The matrix consists of elements, is the singular vector matrix Center front The matrix consists of elements, is the singular vector matrix Center front The matrix consists of elements, For the matrix The transpose of For the matrix The transpose of For the matrix The transpose of .

8. The method according to claim 1, characterized in that The similar feature matrix is ​​normalized and clustered to obtain visualized spatiotemporal feature quantities of the power system scenario, including: Construct the unnormalized Laplacian matrix; Construct an intermediate matrix from the unnormalized Laplacian matrix; Determining a normalized Laplace matrix according to the intermediate matrix; Construct the P matrix according to the eigenvectors corresponding to the eigenvalues ​​of the normalized Laplace matrix; Normalize the row vectors of the P matrix to obtain the Q matrix; The Q matrix is ​​visualized and the results of the visualization are clustered to obtain visualized spatiotemporal characteristics of the power system scenario.

9. The method according to claim 8, characterized in that Construct the unnormalized Laplacian matrix, including: The unnormalized Laplacian matrix is ​​determined according to formula (12): ，（12） in, is the unnormalized Laplacian matrix, is the degree matrix, is the similarity feature matrix; Construct an intermediate matrix based on the unnormalized Laplacian matrix, including: The intermediate matrix is ​​determined according to formula (13) and formula (14): ，（13） in, is the intermediate matrix, is the similarity feature matrix; Determining a normalized Laplace matrix according to the intermediate matrix includes: The normalized Laplacian matrix is ​​determined according to formula (15): ，（15） in, is the normalized Laplacian matrix, is the identity matrix; The P matrix is ​​constructed according to the eigenvectors corresponding to the eigenvalues ​​of the normalized Laplace matrix, including: Selecting a second preset number of eigenvectors from the eigenvectors corresponding to the eigenvalues ​​of the normalized Laplace matrix to obtain the P matrix; Normalize the row vectors of the P matrix to obtain the Q matrix, including: Calculating the Euclidean norm of the P matrix; The vectors in the P matrix are divided by the corresponding Euclidean norm to obtain the Q matrix.

10. A system for extracting spatiotemporal characteristics of power system scenarios based on spectral clustering, characterized in that: The system comprises a processor, and the processor is configured to execute the method according to any one of claims 1 to 9.

Images