A virtual power plant user sensitive data identification method, system, medium and terminal

Abstract

The application discloses a kind of virtual power plant user sensitive data identification method, system, medium and terminal, comprising: obtaining the user data graph to be identified of virtual power plant;The user data graph to be identified is input to sensitive data identification model, to make model to the user data graph to be identified Identification processing and output the sensitive level of user data graph to be identified;Wherein, sensitive data identification model, is introduced on the basis of the graph convolutional layer of graph convolutional neural network model Graph attention mechanism is constituted;The graph convolutional layer of introduction graph attention mechanism is used to based on attention mechanism, and the attention coefficient of adjacent node is combined, to extract the node feature in the user data graph to be identified of input sensitive data identification model node feature in the user data graph to be identified of input sensitive data identification model Node is assigned weight, to obtain.The application introduces graph attention mechanism on the basis of graph convolutional neural network model, to improve the generalization ability of graph neural network model, and then improve the accuracy of user sensitive data identification.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power grid data security, and particularly relates to a virtual power plant user sensitive data identification method, system, medium and terminal. BACKGROUND

[0002] The energy transformation strategy brings different challenges and opportunities to the development of various industries. For the power industry, the challenge lies in how to promote clean energy consumption and optimize existing resource allocation, and the opportunity lies in how to seize the opportunity and stimulate the potential of power energy interactive consumption by deepening the reform of the power market. Many experts and scholars fully exert the spirit of "pioneer", and practice has proved that the virtual power plant model is a favorable weapon for the joint development of challenges and opportunities. The security protection of complex virtual power plant user data has become an important basis for the national grid data security protection work. With the publication of various data protection laws, data identification requirements are clearly proposed. In recent years, power grid companies have actively made a series of preliminary explorations around data identification, and have preliminarily formed identification methods and practical operation basis.

[0003] The existing technology for identifying sensitive data in a virtual power plant mainly generates a to-be-identified graph network according to the relationship of keywords in to-be-identified user data, then inputs the to-be-identified graph network into a graph convolutional neural network to obtain a minimum connected dominating set corresponding to the to-be-identified graph network, and then generates a simplified graph of the to-be-identified graph network based on the minimum connected dominating set, and finally calculates the matching degree of the simplified graph of the to-be-identified graph network and the simplified graph corresponding to the standard sensitive level data to obtain the sensitive identification result of the to-be-identified user data. However, the above scheme has poor generalization ability when identifying sensitive data in a virtual power plant, and cannot meet the accuracy requirements of virtual power plant sensitive data identification. SUMMARY

[0004] The embodiments of the present application provide a virtual power plant user sensitive data identification method, system, medium and terminal, which introduces a graph attention mechanism on the basis of the graph convolution layer of the graph convolutional neural network model to improve the generalization ability of the graph neural network model and thus improve the accuracy of user sensitive data identification.

[0005] To solve the above technical problems, the embodiments of the present application provide a virtual power plant user sensitive data identification method, comprising:

[0006] obtaining a to-be-identified user data graph of a virtual power plant;

[0007] inputting the to-be-identified user data graph into a pre-constructed sensitive data identification model, so that the sensitive data identification model performs identification processing on the to-be-identified user data graph and outputs the sensitive level of the to-be-identified user data graph as a sensitive data identification result;

[0008] The sensitive data identification model is constructed by introducing a graph attention mechanism on the graph convolutional layer of the graph convolutional neural network model. The graph convolutional layer with the graph attention mechanism is used to assign weights to the nodes of the user data graph to be identified, which is input to the sensitive data identification model, based on the attention mechanism and the attention coefficients of adjacent nodes, so as to extract the node features in the user data graph to be identified.

[0009] This invention introduces a graph attention mechanism into the graph convolutional layer of a graph convolutional neural network (GNN) model to construct a sensitive data identification model. The graph convolutional layer with the graph attention mechanism is used to assign weights to each node of the user data graph to be identified, based on the attention mechanism and the attention coefficients of adjacent nodes, thereby extracting node features from the user data graph. By introducing a graph attention mechanism into the GNN model, the limitation of the GNN model in assigning equal weights to adjacent nodes of the same order is overcome, resulting in a more advanced generalization ability and improved accuracy in identifying sensitive user data. After optimizing the graph convolutional layer of the GNN model to obtain the sensitive data identification model, the user data graph of the virtual power plant is input into the sensitive data identification model. The model then processes the user data graph and outputs its sensitivity level as the sensitive data identification result. This allows operation and maintenance personnel to promptly and accurately grasp the data security status of the virtual power plant, providing data support for the development of optimization measures for the virtual power plant. Furthermore, through graph convolutional layers, the model can utilize the node and edge information in the graph structure to capture the contextual relationships between nodes, thereby enabling the model to more accurately identify and predict complex relationships and patterns, especially in dynamically changing scenarios.

[0010] As a preferred embodiment, the extraction of node features from the user data graph to be identified specifically includes:

[0011] Using the graph attention mechanism, the correlation between each node is calculated for all nodes of the user data graph to be identified input into the sensitive data identification model.

[0012] Traverse each node of the user data graph to be identified, and perform softmax normalization on the correlation between the currently traversed node and its neighboring nodes to obtain the attention coefficient corresponding to the currently traversed node.

[0013] According to the preset attention mechanism algorithm, the output vector of the current traversed node is calculated based on the attention coefficient corresponding to the current traversed node and the output vector matrix of the neighboring nodes of the current traversed node, and is used as the node feature corresponding to the current traversed node.

[0014] In a preferred embodiment of the present invention, a graph attention mechanism allows the model to adaptively assign different attention weights to the neighboring nodes of each node, achieving dynamic weight allocation. This enables the model to focus more on neighboring nodes that have a greater impact on the currently traversed node, thereby improving the accuracy of feature extraction. Furthermore, through softmax normalization, the attention coefficients can more clearly reflect the correlation between nodes.

[0015] As a preferred embodiment, the step of calculating the output vector of the currently traversed node according to a preset attention mechanism algorithm, based on the attention coefficient corresponding to the currently traversed node and the output vector matrix of the neighboring nodes of the currently traversed node, is used as the node feature corresponding to the currently traversed node. Specifically:

[0016] h′ i =σ(∑a ij Wh j );

[0017] In the formula, h′ i Let σ represent the output vector of the i-th node; σ represents the activation function; a ij Let h be the attention coefficients of the i-th node and its neighbor j; W is the weight parameter matrix; h is the attention coefficients of the i-th node and its neighbor j. j Let be the output vector matrix of node j.

[0018] The preferred embodiment of this invention, through the combination of activation functions and weight parameter matrices, enables the model to better capture the complex relationships between nodes, thereby improving the overall performance of the model, especially on high-dimensional or complex datasets. Furthermore, in traditional graph neural networks, all neighboring nodes contribute equally to the central node; however, by introducing an attention mechanism, the model can automatically filter out noisy nodes, reducing the impact of noise on the model's prediction results and improving the accuracy and reliability of predictions.

[0019] As a preferred embodiment, the expression for the graph attention mechanism is specifically as follows:

[0020] Attention(Q, S) = Σ <Q,S i >·S i ;

[0021] In the formula, S represents the user data map to be identified in the virtual power plant, Q represents the known user sensitive data map of the virtual power plant, and S i This represents the i-th piece of information in the user data graph to be identified in the virtual power plant.

[0022] In a preferred embodiment of the present invention, by combining the user data map to be identified with the known user sensitive data map, the model can better integrate and utilize known information to identify and predict unknown data. This integration capability helps to improve the accuracy and efficiency of sensitive data identification.

[0023] As a preferred embodiment, the graph convolutional neural network model includes: an input layer, a graph convolutional layer, a fully connected layer, a batch normalization layer, a spatiotemporal convolutional layer, an average pooling layer, a max pooling layer, and an output layer.

[0024] The spatiotemporal convolutional layer is used to perform convolution operations on the spatiotemporal data of each node; the spatiotemporal data of each node is spatiotemporal feature data formed by introducing time and space dimensions into the node features of each node.

[0025] In a preferred embodiment of the present invention, the introduction of a spatiotemporal convolutional layer enables the model to process data in both temporal and spatial dimensions simultaneously. This means the model can not only capture spatial relationships between nodes (such as geographical locations) but also capture behavioral patterns of nodes over time, thereby achieving multi-dimensional feature extraction for a more comprehensive understanding and analysis of the data. Furthermore, the fully connected layer can integrate the features extracted from previous layers to generate the final prediction result. This integrated processing capability allows the model to better adapt to different task requirements, such as classification, regression, and anomaly detection. The batch normalization layer accelerates the training process and reduces the risk of overfitting, while the average pooling layer and max pooling layer further extract important features, reduce data dimensionality, and decrease computational complexity.

[0026] As a preferred embodiment, the acquisition of the user data map to be identified in the virtual power plant specifically includes:

[0027] Real-time collection of virtual power plant user data within a preset time period;

[0028] Keyword extraction is performed on the virtual power plant user data, and the extracted keywords are used as nodes. The contextual relationships between each keyword are used as edges between each node. Based on all nodes and the edges between each node, a user data graph to be identified for the virtual power plant is constructed.

[0029] A preferred embodiment of the present invention involves real-time collection of virtual power plant user data and extraction of keywords. The extracted keywords are used as nodes, and the contextual relationships between the keywords are used as edges between the nodes. Based on all nodes and the edges between them, a graph of the user data to be identified in the virtual power plant is constructed. This integrates user data scattered across different systems into a unified graph structure, so as to intuitively display the correlation between the user data of the virtual power plant, thereby enabling more rapid and accurate identification of abnormal user data.

[0030] To address the same technical problem, embodiments of the present invention also provide a virtual power plant user sensitive data identification system, comprising:

[0031] The data acquisition module is used to acquire the user data map to be identified in the virtual power plant;

[0032] A sensitive data identification module is used to input the user data graph to be identified into a pre-built sensitive data identification model, so that the sensitive data identification model can identify the user data graph and output the sensitivity level of the user data graph as the sensitive data identification result; wherein, the sensitive data identification model is constructed by introducing a graph attention mechanism on the graph convolutional layer of a graph convolutional neural network model; the graph convolutional layer with the graph attention mechanism is used to assign weights to the nodes of the user data graph to be identified input into the sensitive data identification model based on the attention mechanism and combined with the attention coefficients of adjacent nodes, so as to extract the node features in the user data graph to be identified.

[0033] As a preferred embodiment, the data acquisition module specifically includes:

[0034] The data acquisition unit is used to collect virtual power plant user data in real time within a preset time period;

[0035] The data graph construction unit is used to extract keywords from the virtual power plant user data, and use the extracted keywords as nodes, the contextual relationships between each keyword as edges between each node, and construct the user data graph to be identified for the virtual power plant based on all nodes and the edges between each node.

[0036] To address the same technical problem, the present invention also provides a computer-readable storage medium comprising a stored computer program; wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the aforementioned method for identifying sensitive user data in a virtual power plant.

[0037] To address the same technical problem, the present invention also provides a terminal, including a processor, a memory, and a computer program stored in the memory; wherein the computer program can be executed by the processor to implement the aforementioned method for identifying sensitive user data of a virtual power plant. Attached Figure Description

[0038] Figure 1 This is a flowchart illustrating a method for identifying sensitive user data in a virtual power plant, as provided in Embodiment 1 of the present invention.

[0039] Figure 2This is a schematic diagram of the structure of a virtual power plant user sensitive data identification system provided in Embodiment 1 of the present invention. Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] Example 1:

[0042] Please refer to Figure 1 This invention provides a method for identifying sensitive user data in a virtual power plant, comprising steps S1 to S2, the specific details of which are as follows:

[0043] Step S1: Obtain the user data map of the virtual power plant to be identified.

[0044] As a preferred embodiment, step S1 includes steps S11 to S12, each of which is detailed below:

[0045] Step S11: Collect virtual power plant user data in real time within a preset time period.

[0046] In this embodiment, a power data acquisition layer is used to extract data such as the operating status of user equipment, transmission quota, energy type, electricity price, source address, and reception time of the virtual power plant over a certain period of time, which serves as the virtual power plant user data for the preset time period. The power data acquisition layer is the physical area and basic module of the virtual power plant power data management platform.

[0047] Step S12: Extract keywords from the virtual power plant user data, use the extracted keywords as nodes, use the contextual relationships between keywords as edges between nodes, and construct the user data graph of the virtual power plant based on all nodes and the edges between nodes.

[0048] It should be noted that a graph is a set of vertices and edges, or a set of nodes and their relationships. Nodes generally point to entities under study, while edges point to the connections between these entities; these connections can be entity connections.

[0049] Step S2: Input the user data graph to be identified into the pre-built sensitive data identification model so that the sensitive data identification model can identify the user data graph and output the sensitivity level of the user data graph to be identified as the sensitive data identification result.

[0050] In this embodiment, the sensitivity levels include four levels: severe, high, medium, and low.

[0051] As an example, the process of "sensitive data identification model identifying and processing user data graphs to be identified" is as follows: Based on the user data graph to be identified, the adjacent node feature information of the target node's neighboring nodes is generated. Then, the target node's own feature information is updated through the adjacent node feature information. This allows the coal seam of the graph neural network model to radiate its own information to first-order adjacent nodes after information aggregation and transformation, thereby improving the graph convolutional neural network model's ability to diffuse local information to the whole.

[0052] It's important to note that graph convolutional neural network models primarily utilize graph Fourier transform to achieve convolution. They derive the Laplacian operator in the frequency domain from the graph's Laplacian matrix, and then, analogous to convolution in Euclidean space in the frequency domain, derive the formula for graph convolution. Compared to spectral domain convolution, spatial domain convolution is more easily understood in its design philosophy. Based on the principle of spatial correlation, spatial domain convolution uses the target node's own feature information and the feature information of its neighboring nodes to update the target node's feature information, completing the "classification" operation by aggregating the information of surrounding nodes into itself. In practical implementation, graph convolutional layers and graph pooling modules can be interleaved to ultimately extract high-dimensional features from graph data. The initial input graph is processed through graph convolution into high-level substructures, i.e., various levels of graph representation or feature information extraction. The extracted information is then used to output the classification result.

[0053] The sensitive data identification model is constructed by introducing a graph attention mechanism on the graph convolutional layer of the graph convolutional neural network model to improve the adaptive capability of the graph convolutional neural network model. The graph convolutional layer with the graph attention mechanism is used to assign weights to the nodes of the user data graph to be identified by the sensitive data identification model based on the attention mechanism and the attention coefficients of the adjacent nodes, so as to extract the node features in the user data graph to be identified.

[0054] It's important to note that graph attention mechanisms originate from the field of computer vision and are inspired by how human cognitive neurons process information. When humans process information, they selectively focus on more important parts and ignore less important parts to reduce processing effort and improve efficiency. Applying this mechanism to computers requires filtering out useful information from a large amount of data and assigning weights to the information to be processed, giving high weights to parts that require special attention and low weights to useless parts.

[0055] As a preferred embodiment, the above-mentioned process of "extracting node features from the user data graph to be identified" includes steps S21 to S23, each of which is as follows:

[0056] Step S21: Using the graph attention mechanism, calculate the correlation between all nodes of the user data graph to be identified in the input sensitive data identification model.

[0057] As a preferred option, the mathematical expression of the "graph attention mechanism" mentioned in step S21 can be found in equation (1).

[0058] Attention(Q, S) = ∑ <Q,S i >·S i (1)

[0059] In the formula, S represents the user data map to be identified in the virtual power plant, Q represents the known user sensitive data map of the virtual power plant, and S i This represents the i-th piece of information in the user data graph to be identified in the virtual power plant.

[0060] Step S22: Traverse each node of the user data graph to be identified, and perform softmax normalization on the correlation between the currently traversed node and its neighboring nodes to obtain the attention coefficient corresponding to the currently traversed node.

[0061] Step S23: According to the preset attention mechanism algorithm, the output vector of the current traversed node is calculated based on the attention coefficient corresponding to the current traversed node and the output vector matrix of the neighboring nodes of the current traversed node, and is used as the node feature corresponding to the current traversed node.

[0062] As a preferred option, the "attention mechanism algorithm" mentioned in step S23 is specifically referred to in equation (2).

[0063] h′ i =σ(Σa ij Wh j (2)

[0064] In the formula, h′ i Let σ represent the output vector of the i-th node; σ represents the activation function; a ij Let h be the attention coefficients of the i-th node and its neighbor j; W is the weight parameter matrix; h is the attention coefficients of the i-th node and its neighbor j. j Let be the output vector matrix of node j.

[0065] It should be noted that by introducing an attention mechanism, the problem that graph convolutional neural network models can only assign the same weights to adjacent intersection points of the same order is solved, thereby enabling graph neural network models to have better generalization ability and improve the accuracy of user sensitive data identification.

[0066] As a preferred option, the graph convolutional neural network model includes: an input layer, a graph convolutional layer, a fully connected layer, a batch normalization layer, a spatiotemporal convolutional layer, an average pooling layer, a max pooling layer, and an output layer.

[0067] The spatiotemporal convolutional layer is used to perform convolution operations on the spatiotemporal data of each node; the spatiotemporal data of each node is spatiotemporal feature data formed by introducing time and space dimensions into the node features of each node.

[0068] Please refer to Figure 2 This is a schematic diagram of a virtual power plant user sensitive data identification system provided in an embodiment of the present invention. The system includes a data acquisition module M1 and a sensitive data identification module M2, and the specific details of each module are as follows:

[0069] The data acquisition module M1 is used to acquire the user data map to be identified in the virtual power plant;

[0070] The sensitive data identification module M2 is used to input the user data graph to be identified into a pre-built sensitive data identification model, so that the sensitive data identification model can identify and process the user data graph to be identified and output the sensitivity level of the user data graph to be identified as the sensitive data identification result. The sensitive data identification model is constructed by introducing a graph attention mechanism on the graph convolutional layer of the graph convolutional neural network model. The graph convolutional layer with the graph attention mechanism is used to assign weights to the nodes of the user data graph to be identified input into the sensitive data identification model based on the attention mechanism and combined with the attention coefficients of neighboring nodes, so as to extract the node features in the user data graph to be identified.

[0071] As a preferred embodiment, the data acquisition module M1 specifically includes a data acquisition unit 11 and a data graph construction unit 12, with each unit as follows:

[0072] Data acquisition unit 11 is used to collect virtual power plant user data in real time within a preset time period;

[0073] The data graph construction unit 12 is used to extract keywords from the virtual power plant user data, and use the extracted keywords as nodes, and the contextual relationships between each keyword as edges between each node. Based on all nodes and the edges between each node, a user data graph of the virtual power plant to be identified is constructed.

[0074] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0075] This invention also provides a computer-readable storage medium, which includes a stored computer program; wherein, when the computer program is running, it controls the device where the computer-readable storage medium is located to execute the virtual power plant user sensitive data identification method described in Embodiment 1.

[0076] This invention also provides a terminal, including a processor, a memory, and a computer program stored in the memory; wherein the computer program can be executed by the processor to realize the identification of sensitive user data of a virtual power plant as described in Embodiment 1.

[0077] Preferably, the computer program can be divided into one or more modules / units (such as a computer program, a computer program), and one or more modules / units are stored in memory and executed by a processor to complete the present invention. One or more modules / units can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in a terminal.

[0078] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor, or any conventional processor. The processor is the control center of the terminal, connecting various parts of the terminal through various interfaces and lines.

[0079] The memory mainly consists of a program storage area and a data storage area. The program storage area can store the operating system, applications required for at least one function, etc., while the data storage area can store related data. Furthermore, the memory can be high-speed random access memory, or non-volatile memory, such as plug-in hard drives, smart media cards (SMC), secure digital cards (SD), and flash cards, or other volatile solid-state storage devices.

[0080] It should be noted that the above-mentioned terminal may include, but is not limited to, processors and memory. Those skilled in the art will understand that the above-mentioned terminal is merely an example and does not constitute a limitation on the terminal. It may include more or fewer components, or combine certain components, or different components.

[0081] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0082] This invention provides a method, system, medium, and terminal for identifying sensitive user data in a virtual power plant. It introduces a graph attention mechanism into the graph convolutional layer of a graph convolutional neural network model to construct a sensitive data identification model. The graph convolutional layer with the graph attention mechanism is used to assign weights to each node of the user data graph to be identified, based on the attention mechanism and the attention coefficients of adjacent nodes, in order to extract node features from the user data graph. By introducing a graph attention mechanism into the graph convolutional neural network model, the limitation of the model in assigning equal weights to adjacent nodes of the same order is overcome, resulting in a more superior generalization ability and improved accuracy in identifying sensitive user data. After optimizing the graph convolutional layers of the graph convolutional neural network model to obtain a sensitive data identification model, the user data graph of the virtual power plant to be identified is input into the sensitive data identification model. The model then processes the user data graph and outputs its sensitivity level as the sensitive data identification result. This allows operation and maintenance personnel to promptly and accurately grasp the data security status of the virtual power plant, providing data support for developing optimization measures. Furthermore, through the graph convolutional layers, the model can utilize the node and edge information in the graph structure to capture the contextual relationships between nodes, enabling the model to more accurately identify and predict complex relationships and patterns, especially in dynamically changing scenarios.

[0083] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for identifying sensitive user data in a virtual power plant, characterized in that, include: Obtain the user data map to be identified in the virtual power plant; The user data map to be identified is input into a pre-built sensitive data identification model, so that the sensitive data identification model can identify the user data map and output the sensitivity level of the user data map as the sensitive data identification result. The sensitive data identification model is constructed by introducing a graph attention mechanism on the graph convolutional layer of the graph convolutional neural network model. The graph convolutional layer with the graph attention mechanism is used to assign weights to the nodes of the user data graph to be identified, which is input to the sensitive data identification model, based on the attention mechanism and combined with the attention coefficients of the adjacent nodes, so as to extract the node features in the user data graph to be identified. Specifically, the extraction of node features in the user data graph to be identified involves: using a graph attention mechanism, calculating the correlation between all nodes in the user data graph to be identified input to the sensitive data identification model; traversing each node in the user data graph to be identified, performing softmax normalization on the correlation between the currently traversed node and its neighboring nodes to obtain the attention coefficient corresponding to the currently traversed node; and calculating the output vector of the currently traversed node based on the attention coefficient and the output vector matrix of its neighboring nodes, according to a preset attention mechanism algorithm, using this as the node feature corresponding to the currently traversed node.

2. The method for identifying sensitive user data in a virtual power plant as described in claim 1, characterized in that, The process involves calculating the output vector of the currently traversed node based on the attention coefficient corresponding to the current traversed node and the output vector matrix of the neighboring nodes, according to a preset attention mechanism algorithm. This output vector serves as the node feature corresponding to the current traversed node. Specifically: h′ i =σ(∑a ij Wh j ); In the formula, h′ i Let σ represent the output vector of the i-th node; σ represents the activation function; a ij Let be the attention coefficient between the i-th node and its neighbor node j; W is the weight parameter matrix; h j Let be the output vector matrix of node j.

3. The method for identifying sensitive user data in a virtual power plant as described in claim 1, characterized in that, The expression for the graph attention mechanism is as follows: Attention(Q,S)=∑<Q,S i >·S i ; In the formula, S represents the user data map to be identified in the virtual power plant, Q represents the known user sensitive data map of the virtual power plant, and S i This represents the i-th piece of information in the user data graph to be identified in the virtual power plant.

4. The method for identifying sensitive user data in a virtual power plant as described in claim 1, characterized in that, The graph convolutional neural network model includes: an input layer, a graph convolutional layer, a fully connected layer, a batch normalization layer, a spatiotemporal convolutional layer, an average pooling layer, a max pooling layer, and an output layer. The spatiotemporal convolutional layer is used to perform convolution operations on the spatiotemporal data of each node; the spatiotemporal data of each node is spatiotemporal feature data formed by introducing time and space dimensions into the node features of each node.

5. The method for identifying sensitive user data in a virtual power plant as described in claim 1, characterized in that, The acquisition of the user data map to be identified in the virtual power plant specifically involves: Real-time collection of virtual power plant user data within a preset time period; Keyword extraction is performed on the virtual power plant user data, and the extracted keywords are used as nodes. The contextual relationships between each keyword are used as edges between each node. Based on all nodes and the edges between each node, a user data graph to be identified for the virtual power plant is constructed.

6. A virtual power plant user sensitive data identification system, characterized in that, include: The data acquisition module is used to acquire the user data map to be identified in the virtual power plant; A sensitive data identification module is used to input the user data graph to be identified into a pre-built sensitive data identification model, so that the sensitive data identification model can identify the user data graph and output the sensitivity level of the user data graph as the sensitive data identification result; wherein, the sensitive data identification model is constructed by introducing a graph attention mechanism on the graph convolutional layer of a graph convolutional neural network model; the graph convolutional layer with the graph attention mechanism is used to assign weights to the nodes of the user data graph to be identified input into the sensitive data identification model based on the attention mechanism and combined with the attention coefficients of adjacent nodes, so as to extract the node features in the user data graph to be identified; wherein The extraction of node features in the user data graph to be identified specifically involves: using a graph attention mechanism, calculating the correlation between all nodes in the user data graph to be identified input to the sensitive data identification model; traversing each node in the user data graph to be identified, performing softmax normalization on the correlation between the currently traversed node and its neighboring nodes to obtain the attention coefficient corresponding to the currently traversed node; and calculating the output vector of the currently traversed node based on the attention coefficient corresponding to the currently traversed node and the output vector matrix of the neighboring nodes of the currently traversed node, according to a preset attention mechanism algorithm, and using this as the node feature corresponding to the currently traversed node.

7. The virtual power plant user sensitive data identification system as described in claim 6, characterized in that, The data acquisition module specifically includes: The data acquisition unit is used to collect virtual power plant user data in real time within a preset time period; The data graph construction unit is used to extract keywords from the virtual power plant user data, and use the extracted keywords as nodes, the contextual relationships between each keyword as edges between each node, and construct the user data graph to be identified for the virtual power plant based on all nodes and the edges between each node.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program; wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform a method for identifying sensitive user data of a virtual power plant as described in any one of claims 1 to 5.

9. A terminal, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory; wherein the computer program can be executed by the processor to implement a method for identifying sensitive user data of a virtual power plant as described in any one of claims 1 to 5.

Images