Federal map learning-based power Internet of Things equipment anomaly detection method and system

Abstract

The invention discloses an electric power Internet of Things equipment anomaly detection method and system based on federal map learning. The method comprises the following steps: constructing graph structure data; in a client layer, using a graph neural network as a local node classifier of a downstream task, and using an adaptive graph interpolation generator to repair sub-graphs and lost cross-sub-graph connections; adopting a plurality of servers as central nodes, and uploading obtained model parameters to the central server of a local area after a client is locally trained to converge; in the server layer, potential global features of covered clients are obtained according to an automatic encoder, model parameters are allowed to be transmitted between adjacent servers, and a global model is obtained by adopting an aggregation algorithm and issued to each client; and after each client updates the model, the next round of training is continued until the optimal detection model is obtained. Potential connections between sub-graphs are mined by using an adaptive graph interpolation generator, and the blank of lack of cross-client topological information is filled.

Description

Technical Field

[0001] The present invention belongs to the technical field of power grid equipment detection, and particularly relates to a method and system for detecting anomalies in power Internet of Things devices based on federated graph learning. Background Art

[0002] In the power Internet of Things environment, there are a large number of edge devices. Each edge device generates a large amount of time-series data, which may contain sensitive information such as users' electricity consumption habits and geographical locations. Moreover, traditional centralized learning methods require a large amount of data input. If this data is transmitted over the network without proper encryption or security measures, it may be intercepted by malicious third parties, resulting in privacy leakage. In addition, most existing federated learning frameworks cannot handle heterogeneous data, and in most methods using graph neural networks as detection models, the neighbor nodes of cross-client subgraphs are not considered, making it impossible to extract more comprehensive features and correlation relationships. Summary of the Invention

[0003] The present invention provides a method and system for detecting anomalies in power Internet of Things devices based on federated graph learning, which is used to solve the technical problem that in most methods using graph neural networks as detection models, the neighbor nodes of cross-client subgraphs are not considered, making it impossible to extract more comprehensive features and correlation relationships.

[0004] In a first aspect, the present invention provides a method for detecting anomalies in power Internet of Things devices based on federated graph learning, including: Constructing graph structure data, where each node in the graph structure data represents a power Internet of Things device, and each edge represents the correlation relationship between two power Internet of Things devices; At the client layer, using a graph neural network as a local node classifier for downstream tasks, and using an adaptive graph imputation generator to repair subgraphs and missing cross-subgraph connections; Using multiple servers as central nodes. After local training at the client converges, the obtained model parameters are uploaded to the central server in this region; At the server layer, obtaining the potential global features of the covered clients according to an autoencoder, and allowing adjacent servers to transmit model parameters, and using an aggregation algorithm to obtain a global model and distribute it to each client; After each client updates the model, continue the next round of training until the optimal detection model is obtained; Obtaining the to-be-tested graph structure data, inputting the to-be-tested graph structure data into the optimal detection model, and the optimal detection model outputs the anomaly result of the power Internet of Things device.

[0005] In a second aspect, the present invention provides a system for detecting anomalies in power Internet of Things devices based on federated graph learning, including: A construction module is configured to construct graph structure data, where each node in the graph structure data represents an electric power Internet of Things device, and each edge represents an association relationship between two electric power Internet of Things devices; The repair module is configured to use the graph neural network as a local node classifier for downstream tasks at the client layer and an adaptive graph interpolation generator to repair subgraphs and lost cross-subgraph connections; The upload module is configured to use multiple servers as central nodes. After the client trains locally until convergence, the obtained model parameters are uploaded to the central server in the region; The acquisition module is configured to obtain the potential global features of the covered clients according to the autoencoder at the server layer, and allow the model parameters to be transmitted between adjacent servers, and adopt the aggregation algorithm to obtain the global model and send it to each client; The training module is configured to continue the next round of training after each client updates the model until the optimal detection model is obtained; The output module is configured to obtain the graph structure data to be tested, input the graph structure data to be tested into the optimal detection model, and the optimal detection model outputs the abnormal result of the power Internet of Things equipment.

[0006] According to a third aspect, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the steps of the method for anomaly detection of electric power Internet of Things equipment based on federated graph learning according to any embodiment of the present invention.

[0007] In a fourth aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon. When the program instructions are executed by a processor, the processor executes the steps of the method for detecting anomalies of electric power Internet of Things devices based on federated graph learning according to any embodiment of the present invention.

[0008] The present application discloses a method and system for detecting anomalies of power Internet of Things devices based on federated graph learning. A federated graph learning structure is used to perform anomaly detection on power Internet of Things devices. A graph neural network is used as an initial model to process heterogeneous data between devices. At the same time, in the process of constructing subgraphs, an adaptive graph interpolation generator is used to mine potential connections between subgraphs to fill the gap of lack of cross-client topology information. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 Flowchart of a method for anomaly detection of power IoT devices based on federated graph learning provided by an embodiment of the present invention; Figure 2 Architecture diagram of a multi-edge server federated graph learning system provided by an embodiment of the present invention; Figure 3 Block diagram of the structure of an anomaly detection system for power IoT devices based on federated graph learning provided by an embodiment of the present invention; Figure 4 Schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. Detailed implementation manners

[0011] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0012] Please refer to Figure 1 , which shows a flowchart of a method for anomaly detection of power IoT devices based on federated graph learning of the present application.

[0013] As Figure 1 shown, the method for anomaly detection of power IoT devices based on federated graph learning specifically includes the following steps: Step S101: Construct graph structure data, where each node in the graph structure data represents a power IoT device, and each edge represents the association relationship between two power IoT devices.

[0014] Step S102: At the client layer, use a graph neural network as a local node classifier for downstream tasks, and use an adaptive graph imputation generator to repair subgraphs and missing cross-subgraph connections.

[0015] Current federated graph learning fails to fully consider the problem of missing topological information between clients in real scenarios, resulting in insufficient aggregation of multi-hop neighbor client features. By constructing a global topological graph and selecting the most similar nodes as potential links, this method can effectively repair the missing links between clients. And the generated potential feature vectors help to perform more effective feature propagation in the local model, thereby improving the training effect of the global model. At the same time, the use of autoencoders and random noise input ensures the privacy of the original data, which conforms to the principles of federated learning.

[0016] The adaptive graph interpolation generator includes: an input layer, an encoding layer, a multi-granularity pooling layer, a graph matching layer, a decoding layer, and a topological regularization layer.

[0017] Input layer: The original graph containing missing subgraphs is input. Through graph collapse preprocessing, the high-density subgraphs are compressed into multiple nodes, reducing the computational complexity.

[0018] Encoding layer: Fuses candidate node and neighborhood features. According to the context information of the missing region of the subgraph, the weights of the interpolation path are dynamically generated to control the fusion ratio of candidate nodes and existing nodes, and the enhanced features are output. The formula is: , where, is the output gating value, is the input candidate node feature, is the input adjacent subgraph feature, represents element-wise multiplication, is the output enhanced feature. Compared with the traditional generator's method of directly splicing or averaging neighborhood features, the gating mechanism can adaptively adjust information fusion and avoid noise interference.

[0019] Multi-granularity pooling layer: Performs local and global feature extraction on the enhanced feature . In local pooling, a weighted summation method based on node attention weights is used, while in global pooling, subgraph-level graph convolution is used to capture the associations of local details and global patterns respectively, obtaining a multi-scale representation. The multi-granularity fusion formula is: , where, is the feature after local pooling, representing the weighted aggregation feature of the candidate node set, is the feature after global pooling, representing the overall pattern of the subgraph. MLP is a multi-layer perceptron, || is the concatenation operation, is the multi-granularity pooling feature. Compared with the existing methods that use a single pooling strategy, multi-granularity pooling enhances the modeling ability for complex topologies by jointly modeling local and global features.

[0020] Graph matching layer: Inputs the multi-granularity pooling feature . Through the attention mechanism, the similarity of cross-subgraph nodes is calculated to generate adaptive weights to guide the generation of missing connections. For subgraph repair, the context information of adjacent subgraphs is combined to generate candidate node representations, the association probability between candidate nodes and existing nodes is calculated, and the optimal connection method is selected. For cross-subgraph connection repair, cross-graph convolution in the graph matching layer is used to achieve two-way heterogeneous message passing, fusing the information of the source subgraph and the target subgraph. Compared with the disadvantage of traditional graph convolution that only performs single-graph message passing, the graph matching layer realizes cross-graph feature distillation.

[0021] Decoding layer: Generate the nodes and edges of the missing subgraph according to the reconstructed adjacency matrix generated by the graph matching layer.

[0022] Topological regularization layer: Constrain the degree distribution and loop structure of the graph generated by the decoding layer to improve the topological rationality of the generated subgraph. Calculate the node degree distribution loss between the generated graph and the real graph using KL divergence; at the same time, detect the loop structure in the generated graph through spectral analysis and penalize redundant loops. Compared with traditional generators that rely on implicit learning, topological regularization ensures the physical rationality of the generated structure through explicit topological constraints.

[0023] Step S103: Use multiple servers as central nodes. After the client locally trains until convergence, upload the obtained model parameters to the central server in this region.

[0024] Step S104: At the server layer, obtain the potential global features of the covered clients according to the autoencoder, and allow the transmission of model parameters between adjacent servers. Use the aggregation algorithm to obtain the global model and distribute it to each client.

[0025] Step S105: After each client updates the model, continue the next round of training until the optimal detection model is obtained.

[0026] Step S106: Obtain the to-be-tested graph structure data, input the to-be-tested graph structure data into the optimal detection model, and the optimal detection model outputs the abnormal result of the power Internet of Things device.

[0027] In a specific embodiment, the method for detecting anomalies in power Internet of Things devices based on federated graph learning is described in detail from four aspects: data processing, subgraph repair and feature extraction, model update, and anomaly detection. I. Data processing First, it is necessary to collect operation data from each terminal device in the power Internet of Things. This data may include but is not limited to device status information, power consumption, fault records, etc. Divide the power Internet of Things devices into multiple subgraphs according to the actual application scenario, and each subgraph corresponds to a local area or a specific functional group. Each client obtains data related to the power Internet of Things devices in various ways, extracts the data generated by the devices as the node set V, extracts the association relationships between the devices as the edge set E, and extracts the spatio-temporal features of the data as the feature matrix X. Construct a graph network as the global graph. Where , n represents the number of nodes; is the edge set storing the connection relationship between nodes u and v, where ; represents the node feature matrix, where is the i-th node of the node feature vector, and d is the dimension of the feature vector. Suppose there are N edge servers and M clients. The edge servers cover local clients for federated graph learning training, where . The clients have local subgraphs of the global graph, denoted by , where represents the node set of the local subgraph, represents the edge set of the local subgraph, represents the node feature matrix of the local subgraph. To simulate the real scenario of the lack of connections between clients, we assume that there are no shared nodes and connections between clients, denoted by , and are different subgraph node sets. The subgraphs of all clients form a complete graph, defined as .

[0028] II. Subgraph Repair and Feature Extraction To capture potential cross-subgraph connections, we use an adaptive graph imputation generator to repair the subgraphs to explore the learnable potential graph to repair cross-subgraph connections, where is the cross-subgraph connection edge set of the potential subgraph, is the node feature matrix of the potential subgraph. To construct globally shared information without revealing the original data, the clients upload the processed embeddings to the edge servers at every K intervals of edge-client communication, where the original connected nodes remain close in the low-dimensional space. Next, the adaptive graph imputation generator is used to fuse the processed embeddings to obtain the globally shared information , where is the number of all clients covered, and c is the input dimension. Therefore, is denoted as: , where is the processed embedding representation of the first client, is the -th client's processed embedding representation.

[0029] In the actual application scenario of federated graph learning, each node in the client may have potential cross-subgraph connections. If these cross-subgraph connections are missing, the client may not be sufficient to propagate features in the multi-hop neighbors. Therefore, the adaptive graph imputation generator uses distance to evaluate node similarity and constructs a global topology graph, denoted as , where The distance for evaluating node similarity is the transpose of the embedded representation after client processing. Then, k most similar nodes are selected from this topological graph as potential cross-subgraph connections. To generate potential feature vectors under the knowledge of global shared information, an autoencoder parameterized by is used to obtain the complete underlying representation of . To protect data privacy, a random noise vector S is input into the autoencoder, so the output of the autoencoder is reconstructed as , where f(·) and h(·) are the encoder and decoder respectively. And represents the potential features that the encoder expects to extract. Using the autoencoder to map the random noise vector to the same dimension as , the output of the (l + 1)-th layer is defined as: , where and are the weights and biases of a specific layer respectively, represents the activation function.

[0030] Since the conditional distribution of depends on and is independent of S, an evaluator parameterized by is used to supervise the quality of the decoder's reconstructed data, aiming to extract the expected underlying features tailored for node classification . Specifically, the loss function of the autoencoder is defined as: , where is the expectation of the variable in , represents sampled from the distribution of , is the global information evaluator. The loss function of the evaluator is defined as: , III. Model Update The specific architecture of this multi-center server federated graph learning system is as follows Figure 2 shown.

[0031] In practical application scenarios, as the number of clients increases, especially when the client geographical locations are scattered, a single edge server may encounter problems such as excessively high communication costs and excessive performance consumption. And when the central server fails, there is a risk of paralysis in the entire federated graph learning system. Therefore, we adopt multiple edge servers to extend federated graph learning to a multi-edge environment, enabling the federated graph learning system to promote more efficient training and better load balancing in a multi-edge collaborative environment. We consider using N edge servers in this system, and the edge servers are equipped with a global node classifier parameterized by . And clients only communicate with their nearest edge servers, which can reduce unnecessary communication costs. There is a neighbor relationship between servers, represented by the matrix . If server and server are neighbors, then , otherwise . And parameter transmission is allowed between neighbor servers.

[0032] To enhance feature propagation in local tasks and facilitate the inference of the subsequent global model, we adopt an L-layer GNN model with a local node classifier , defined as: , where GNNconv(·) is a GNN model, represents the GNN output of the i-th client covered by . The feature propagation of the (l + 1)-th layer is shown as follows: , where, is the vector of node u in the l-th GNN layer, is the vector of node v in the l-th GNN layer, represents the connection between nodes u and v, AGG(·) is an aggregation function for integrating the neighbor features of node u through , and || represents the concatenation operation.

[0033] During local training, the edge server exchanges information with the covered clients in each client communication. The client and the nearest edge server collaborate to utilize the shared information and adopt an adaptive graph imputation generator to extract potential connections. Specifically, the loss function of the i-th client covered by is defined as: where ​is the loss function of the i-th client, which is used to measure the local empirical risk; is the local node classifier and the learnable weights of. Each client trains the local node classifier parameterized by in parallel during the local training round , and the specific formula is: , where α is the learning rate, t represents the local training round, represents the parameters of the t-th round, where is the loss function of the i-th client, is the node classifier of the i-th client. After local training, aggregates the local parameters to update the global parameters , and then is broadcast to all clients on each edge.

[0034] To better explore potential cross-subgraph connections by leveraging information from other servers, we adopt the topological structure of the edge layer to facilitate parameter transmission between neighbor servers. Enable information to flow between clients through gradient propagation at every K intervals of edge client communication. Specifically, first aggregates the model parameters of the aggregator neighbor servers, then averages the parameters and broadcasts them to the covered clients. This process can be described by the following formula: , where, is the connection situation between the edge server and the edge server ; is the client covered by the edge server ; represents the learnable weights of the node classifier in the client .

[0035] IV. Anomaly Detection After obtaining the optimal detection model according to the model update, use this model to perform anomaly detection on the data of power Internet of Things devices. First, it is necessary to collect real-time data from each power device and map these data into the previously constructed graph structure, where each device is used as a node in the graph. Subsequently, use the trained model to predict these nodes, and judge whether the working state of each device is normal through the anomaly score output by the model. If the score of a certain device exceeds the pre-set threshold, it is considered that the device may have an abnormal situation.

[0036] In summary, most of the existing federated learning frameworks cannot handle heterogeneous data, and most of the methods that use graph neural networks as detection models do not consider neighbor nodes across client subgraphs, and cannot extract more comprehensive features and associations. Therefore, a federated graph learning structure is used to detect anomalies in power IoT devices, and graph neural networks are used as the initial model to process heterogeneous data between devices. At the same time, in the process of constructing subgraphs, an adaptive graph interpolation generator is used to mine potential connections between subgraphs to fill the gap of lack of cross-client topology information.

[0037] And most federated learning systems use a single server as the central node. This architecture may be difficult to expand to a very large client group, because the central server needs to handle more and more communication requests, resulting in excessive consumption of central server resources. And a single server failure may cause the entire federated learning system to collapse, affecting model performance. To address this problem, we proposed multi-center federated learning. Multi-edge server federated learning uses multiple servers as central nodes, each of which is responsible for learning different subgraphs, and model parameters are shared between adjacent central nodes. While ensuring privacy, it can reduce resource consumption and reduce the risk caused by single point failures.

[0038] See also Figure 3 , which shows a structural block diagram of a power Internet of Things device anomaly detection system based on federated graph learning in the present application.

[0039] like Figure 3 As shown, the power Internet of Things device anomaly detection system 200 includes a construction module 210, a repair module 220, an upload module 230, an acquisition module 240, a training module 250 and an output module 260.

[0040] Among them, the construction module 210 is configured to construct graph structure data, where each node in the graph structure data represents an electric power Internet of Things device, and each edge represents the association relationship between two electric power Internet of Things devices; the repair module 220 is configured to use a graph neural network as a local node classifier for downstream tasks in the client layer, and use an adaptive graph imputation generator to repair subgraphs and lost cross-subgraph connections; the upload module 230 is configured to use multiple servers as central nodes, and after the client locally trains to convergence, upload the obtained model parameters to the central server in this region; the acquisition module 240 is configured to obtain the potential global features of the covered clients according to the autoencoder in the server layer, and allow adjacent servers to transmit model parameters, and use an aggregation algorithm to obtain the global model and distribute it to each client; the training module 250 is configured to continue the next round of training after each client updates the model until the optimal detection model is obtained; the output module 260 is configured to obtain the graph structure data to be measured, input the graph structure data to be measured into the optimal detection model, and the optimal detection model outputs the abnormal result of the electric power Internet of Things device.

[0041] It should be understood that Figure 3 The various modules described in Figure 1 correspond to the respective steps in the method described in the reference Figure 3 Therefore, the operations, features, and corresponding technical effects described above for the method also apply to Figure 3 the various modules herein, and will not be repeated here.

[0042] In some other embodiments, the embodiments of the present invention further provide a computer-readable storage medium, on which a computer program is stored. When the program instructions are executed by a processor, the processor is caused to execute the method for detecting anomalies of electric power Internet of Things devices based on federated graph learning in any of the above method embodiments; As an implementation manner, the computer-readable storage medium of the present invention stores computer-executable instructions, and the computer-executable instructions are set as: Construct graph structure data, where each node in the graph structure data represents an electric power Internet of Things device, and each edge represents the association relationship between two electric power Internet of Things devices; In the client layer, use a graph neural network as a local node classifier for downstream tasks, and use an adaptive graph imputation generator to repair subgraphs and lost cross-subgraph connections; Use multiple servers as central nodes, and after the client locally trains to convergence, upload the obtained model parameters to the central server in this region; In the server layer, obtain the potential global features of the covered clients according to the autoencoder, and allow adjacent servers to transmit model parameters, and use an aggregation algorithm to obtain the global model and distribute it to each client; After each client updates the model, continue the next round of training until the optimal detection model is obtained; Obtain the structure data of the graph to be measured, and input the structure data of the graph to be measured into the optimal detection model. The optimal detection model outputs the abnormal result of the power Internet of Things device.

[0043] The computer-readable storage medium may include a storage program area and a storage data area. Among them, the storage program area can store an operating system and application programs required for at least one function; the storage data area can store data created according to the use of the power Internet of Things device abnormal detection system based on federated graph learning, etc. In addition, the computer-readable storage medium may include high-speed random access memory, and may also include memories, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the computer-readable storage medium may optionally include a memory remotely provided with respect to the processor, and these remote memories may be connected to the power Internet of Things device abnormal detection system based on federated graph learning through a network. Examples of the above network include but are not limited to the Internet, enterprise intranet, local area network, mobile communication network, and combinations thereof.

[0044] Figure 4 It is a schematic structural diagram of the electronic device provided by an embodiment of the present invention. As Figure 4 shown, the device includes: a processor 310 and a memory 320. The electronic device may further include: an input device 330 and an output device 340. The processor 310, the memory 320, the input device 330, and the output device 340 may be connected through a bus or other means, Figure 4 taking the connection through the bus as an example. The memory 320 is the above-mentioned computer-readable storage medium. The processor 310 executes various functional applications and data processing of the server by running non-volatile software programs, instructions, and modules stored in the memory 320, that is, implements the power Internet of Things device abnormal detection method based on federated graph learning in the above method embodiment. The input device 330 can receive input digital or character information, and generate key signal inputs related to user settings and function controls of the power Internet of Things device abnormal detection system. The output device 340 may include display devices such as a display screen.

[0045] The above electronic device can execute the method provided by the embodiment of the present invention, and has corresponding functional modules and beneficial effects for executing the method. For technical details not described in detail in this embodiment, reference may be made to the method provided by the embodiment of the present invention.

[0046] As an implementation manner, the above electronic device is applied to an abnormal detection system for power Internet of Things devices based on federated graph learning and is used for a client, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to: Construct graph structure data, where each node in the graph structure data represents a power Internet of Things device, and each edge represents the association relationship between two power Internet of Things devices; At the client layer, use a graph neural network as a local node classifier for downstream tasks, and use an adaptive graph imputation generator to repair subgraphs and missing cross-subgraph connections; Adopt multiple servers as central nodes. After local training at the client converges, upload the obtained model parameters to the central server in this region; At the server layer, obtain the potential global features of the covered clients according to the autoencoder, and allow the transmission of model parameters between adjacent servers. Adopt an aggregation algorithm to obtain the global model and distribute it to each client; After each client updates the model, continue the next round of training until the optimal detection model is obtained; Obtain the graph structure data to be measured, input the graph structure data to be measured into the optimal detection model, and the optimal detection model outputs the abnormal result of the power Internet of Things device.

[0047] Through the description of the above implementation manners, those skilled in the art can clearly understand that each implementation manner can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disc, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiments.

[0048] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A method for detecting abnormality of power Internet of Things devices based on federated graph learning, characterized in that: include: Construct graph structure data, where each node represents a power IoT device, and each edge represents the association between two power IoT devices; At the client level, we use graph neural networks as local node classifiers for downstream tasks and an adaptive graph interpolation generator to repair subgraphs and lost cross-subgraph connections. Using multiple servers as central nodes, the client will upload the model parameters obtained after local training to convergence to the central server in the region; At the server level, the potential global features of the covered clients are obtained based on the autoencoder, and the model parameters are allowed to be transmitted between adjacent servers. The global model is obtained by using the aggregation algorithm and sent to each client. After each client updates the model, it continues the next round of training until the optimal detection model is obtained; The structure data of the graph to be tested is obtained, and the structure data of the graph to be tested is input into the optimal detection model. The optimal detection model outputs the abnormal results of the power Internet of Things equipment.

2. According to claim 1, a method for detecting abnormality of power Internet of Things devices based on federated graph learning is characterized in that: Building graph structure data includes: Obtain relevant data of power IoT devices, and extract data generated by power IoT devices from the relevant data as a node set , extract the association relationship between power IoT devices as the edge set , extract the spatiotemporal features in the relevant data as the feature matrix ; Building a graph network As a global graph, where , Indicates the number of nodes, Is a storage node and nodes The edge set of the connection relationship between , For Node and nodes There are edges connecting the represents the node feature matrix, is the field of real numbers, is the node feature vector of the ith node, is the dimension of the feature vector; Assume that Edge servers and Clients, Edge Servers cover Local Clients Conduct federated graph learning training, where , Client Have a local subgraph of the global graph, with To express, represents the set of nodes of the local subgraph, represents the edge set of the local subgraph, Represents the node feature matrix of the local subgraph.

3. According to claim 1, a method for detecting abnormality of power Internet of Things devices based on federated graph learning is characterized in that: The adaptive graph interpolation generator includes: input layer, encoding layer, multi-granularity pooling layer, graph matching layer, decoding layer and topology regularization layer.

4. According to claim 1, a method for detecting abnormality of power Internet of Things devices based on federated graph learning is characterized in that: The loss function of the autoencoder is: ， In the formula, is the loss function of the autoencoder, for The expectation of the variables in is the conditional probability distribution, For Node The characteristic representation of Indicates from Sampling from the distribution of , is the global information evaluator; Among them, the loss function of the global information evaluator is: ， In the formula, is the loss function of the global information estimator.

5. According to the method for detecting abnormality of power Internet of Things devices based on federated graph learning in claim 1, it is characterized in that: in, The expression that allows model parameters to be transmitted between adjacent servers is: ， In the formula, is a global parameter, is the number of edge servers, For edge servers With edge servers The connection status, For edge servers Clients covered, are the learnable weights of the node classifier in the client.

6. A power Internet of Things equipment anomaly detection system based on federated graph learning, characterized in that: include: A construction module is configured to construct graph structure data, where each node in the graph structure data represents an electric power Internet of Things device, and each edge represents an association relationship between two electric power Internet of Things devices; The repair module is configured to use the graph neural network as a local node classifier for downstream tasks at the client layer and an adaptive graph interpolation generator to repair subgraphs and lost cross-subgraph connections; The upload module is configured to use multiple servers as central nodes. After the client trains locally until convergence, the obtained model parameters are uploaded to the central server in the region; The acquisition module is configured to obtain the potential global features of the covered clients according to the autoencoder at the server layer, and allow the model parameters to be transmitted between adjacent servers, and adopt the aggregation algorithm to obtain the global model and send it to each client; The training module is configured to continue the next round of training after each client updates the model until the optimal detection model is obtained; The output module is configured to obtain the graph structure data to be tested, input the graph structure data to be tested into the optimal detection model, and the optimal detection model outputs the abnormal result of the power Internet of Things equipment.

7. An electronic device, characterized in that: include: At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the method of any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images