Electric power communication network frame weak point identification method and device, electronic equipment and medium
By obtaining multimodal data of the power communication network, processing node type data and optimizing the model, identifying the weaknesses of the power communication network, the problem of inefficient identification in the existing technology is solved, and more efficient and accurate weakness identification is achieved. Ensure the stability and security of the power communication system.
Patent Information
- Application Number
- CN202411119435.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-12
AI Technical Summary
The prior art is difficult to accurately identify weak points in the power communication network architecture, resulting in low identification efficiency and inaccurateness, especially in complex data environments.
By obtaining multimodal data of the power communication network, using the classification module and the identification module to process node type data, calculate the weak level, and optimize the initial model according to the loss function to obtain the target model, which is used to identify the weak points of the power communication network.
It improves the efficiency and accuracy of the identification of weak points of the power communication network architecture, provides a more solid data foundation, and ensures the safe and stable operation of the power communication system.
Smart Images

Figure CN118944945A_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present invention relate to the field of power communication technology, and in particular to a method, device, electronic device and medium for identifying weak points in a power communication grid. Background Art
[0002] The power communication network plays an important role in regional power transmission and communication. However, the power communication network often has architectural weaknesses, which have a negative impact on its stable operation. For example, during operation, architectural weaknesses may cause various abnormal situations in the power communication network, including security vulnerabilities, network failures, and insufficient capacity. Security vulnerabilities may lead to data leakage or tampering, network failures affect communication stability, and insufficient capacity may cause data transmission delays or losses, seriously threatening the safety and stable operation of the power communication system. The traditional manual inspection method to determine architectural weaknesses is inefficient, labor-intensive, and requires high investment costs, and there are safety risks in complex geographical environments.
[0003] In recent years, solutions for identifying weak points in the power communication network architecture have gradually emerged. For example, the architecture of the power communication network is comprehensively analyzed using public information models and ledger data to build a multi-dimensional reliability identification system. However, the public information model is difficult to handle complex data, and the power communication network has a complex structure, a large number of devices, and variable operation modes, so its operation data composition is relatively complex. This results in the method of using the public information model to identify weak points in the power communication network architecture is not accurate enough and inefficient. Summary of the invention
[0004] The embodiments of the present invention provide a method, device, electronic device and medium for identifying weak points in a power communication network architecture, which can obtain a target model that can accurately identify weak points in the power communication network architecture based on complex data of the power communication network, thereby using the target model to accurately identify weak points in the power communication network architecture, thereby improving the efficiency and accuracy of identifying weak points in the power communication network architecture.
[0005] In a first aspect, an embodiment of the present invention provides a method for identifying weak points in a power communication grid, comprising:
[0006] When the pre-established initial model does not meet the preset conditions, multimodal data of each network device of the power communication network is obtained; wherein the initial model includes a classification module and an identification module; the multimodal data includes a connection relationship and a topological structure of the network devices;
[0007] Processing the multimodal data according to the classification module to obtain type data of each node of the grid structure of the power communication network;
[0008] Performing weakness calculation on the type data of each node according to the identification module to obtain a weakness level of each node;
[0009] The initial model is optimized according to the weakness level of each node and a predetermined loss function until the initial model meets the preset conditions to obtain a target model; wherein the target model is used to identify weak points of the power communication network.
[0010] In a second aspect, an embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements a method for identifying weak points in a power communication grid as described in any one of the embodiments of the present invention.
[0011] In a third aspect, an embodiment of the present invention provides a device for identifying weak points in a power communication grid, the device comprising:
[0012] A data acquisition module, used for acquiring multimodal data of each network device of the power communication network when the pre-established initial model does not meet the preset conditions; wherein the initial model includes a classification module and an identification module; the multimodal data includes the connection relationship and topological structure of the network devices;
[0013] A data classification module, used for processing the multimodal data according to the classification module to obtain type data of each node of the grid structure of the power communication network;
[0014] A level determination module, used to calculate the weakness of the type data of each node according to the identification module to obtain the weakness level of each node;
[0015] A model optimization module is used to optimize the initial model according to the weakness level of each node and a predetermined loss function until the initial model meets the preset conditions to obtain a target model; wherein the target model is used to identify weak points of the power communication network.
[0016] In a fourth aspect, an embodiment of the present invention further provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a method for identifying weak points in a power communication grid as described in any one of the embodiments of the present invention is implemented.
[0017] In a fifth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for identifying weak points in a power communication grid as described in any one of the embodiments of the present invention.
[0018] In an embodiment of the present invention, when the pre-established initial model does not meet the preset conditions, the multimodal data of each network device of the power communication network is obtained; the multimodal data is processed according to the classification module to obtain the type data of each node of the grid structure of the power communication network; the type data of each node is calculated according to the identification module to obtain the weakness level of each node; the initial model is optimized according to the weakness level of each node and the pre-determined loss function until the initial model meets the preset conditions to obtain the target model; wherein the target model is used to identify the weak points of the power communication network. That is, in an embodiment of the present invention, multimodal data collection can be performed on the power communication network to obtain the comprehensive information of each network device, and the classification module and the identification module are continuously optimized using the comprehensive information to obtain the target model that can accurately identify the weak points of the power communication network architecture. Therefore, the target model is used to accurately identify the weak points of the power communication network architecture, thereby improving the efficiency and accuracy of identifying the weak points of the power communication network architecture. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.
[0020] Figure 1 A first flow chart of a method for identifying weak points in a power communication grid provided by an embodiment of the present invention;
[0021] Figure 2 A second flow chart of a method for identifying weak points in a power communication grid provided by an embodiment of the present invention;
[0022] Figure 3 A flow chart of using a target model to identify weak points in a power communication grid provided by an embodiment of the present invention;
[0023] Figure 4 A schematic diagram of the structure of a device for identifying weak points in a power communication grid provided by an embodiment of the present invention;
[0024] Figure 5 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0025] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.
[0026] Figure 1 This is a first flow chart of a method for identifying weak points in a power communication grid provided in an embodiment of the present invention. The method in an embodiment of the present invention is suitable for obtaining a target model that can accurately identify weak points in the power communication network architecture, thereby using the target model to accurately identify weak points in the power communication network architecture, thereby improving the efficiency and accuracy of identifying weak points in the power communication network architecture. The method can be executed by a power communication grid weak point identification device provided in an embodiment of the present invention, and the device can be implemented in software and / or hardware. The following embodiments will be described by taking the device integrated in an electronic device as an example. The electronic device can be a computer device or a server, etc. Reference Figure 1 , the method may specifically include the following steps:
[0027] Step 101: When the pre-established initial model does not meet the preset conditions, multimodal data of each network device of the power communication network is obtained.
[0028] Among them, multimodal data includes the connection relationship and topological structure of network devices. Multimodal data also includes network traffic information, current information, voltage information and security information of each network device. The power communication network is an important infrastructure to ensure the safe and stable operation of the power system. The power communication network architecture is composed of various network devices. Network equipment includes transmission equipment, switching equipment, routing equipment and support equipment. Transmission equipment includes optical fiber transmission equipment, carrier transmission equipment and microwave transmission equipment; switching equipment includes circuit switches, soft switches and multimedia subsystem switches; routing equipment includes routers, three-layer switches and broadband remote access server switching equipment; support equipment includes network management equipment, synchronization equipment and security equipment.
[0029] The initial model is a model that has not been trained and optimized, and is used to identify the weak points of the power communication network architecture based on multimodal data. The initial model includes a classification module and an identification module, and the classification module is used to classify each node of the grid structure of the power communication network. Each node includes each network device and each connection point between the network devices, and the type of each node is a weak point or a non-weak point. The identification module is used to determine the degree of weakness of each weak point. In an optional embodiment, when the initial model does not meet the preset conditions, the server can collect the connection relationship and topological structure of the network device through a network management system such as a simple network protocol, a command line interface, or NetFlow network traffic technology to ensure the accuracy and real-time nature of the structural data. The network traffic is obtained through a traffic monitoring probe, and the traffic information of the network device is monitored in real time. The current of the network device is obtained through a current sensor, and the voltage of the network device is obtained through a voltage sensor, and the operating status and power load of the network device are monitored according to the current and voltage. Security data is obtained through a log management system and a network access control system, so as to record and analyze network attack logs and access control records. By collecting and organizing these multimodal data, the server can collect and analyze various types of data of the power communication network in real time and accurately, providing a solid data foundation for the security, stability and reliability of the power communication network, and laying the foundation for the subsequent use of multimodal data to obtain target models that can accurately identify weak points.
[0030] Step 102: Process the multimodal data according to the classification module to obtain type data of each node of the grid structure of the power communication network.
[0031] Among them, the classification module is used to obtain the type of each node according to the multimodal data. The type data of each node includes the type to which each node belongs. In this solution, the type of each node is divided into two categories, namely weak nodes and non-weak nodes. Each node includes each network device and the connection point between each network device.
[0032] Specifically, after obtaining the multimodal data, the server can preprocess the multimodal data, including data sorting, data cleaning, and feature extraction. After obtaining the preprocessed multimodal data, the preprocessed multimodal data is input into a classification module, which includes a predefined classification algorithm. The preprocessed multimodal data is classified using the classification algorithm to obtain the type data of each node of the grid structure. The classification algorithm can be a decision tree, a random forest, a K nearest neighbor algorithm, a logistic regression, and a deep learning algorithm. Exemplarily, after obtaining the multimodal data, the multimodal data is segmented so that each node of each grid structure represents a feature, and each branch represents a decision rule. According to the predefined classification rules, the classification result of each leaf node can be finally obtained, that is, the type data of each node of the grid structure.
[0033] In an optional embodiment, after obtaining the multimodal data, feature extraction is performed on the processed multimodal data to obtain feature information for processing each node, the processed feature information includes attribute information entropy features, inflow-outflow ratio features, average flow length features and unilateral link density features; the feature information for processing each node is normalized to obtain target features for processing each node; for the target feature of each node, the expected output corresponding to the target feature of the current node is determined based on the processed multimodal data; based on the predetermined Lagrange multiplier, the expected output of the processing and the sign function of the processing classification module, a classification operation is performed on the target feature of the current node to obtain type data for processing the current node.
[0034] Step 103: Calculate the weakness of the type data of each node according to the identification module to obtain the weakness level of each node.
[0035] Among them, the recognition module includes a predefined recognition method, which is used to determine the weakness of each node, that is, the weakness level. The larger the weakness level, the weaker the node. The recognition method can be a deep learning algorithm, a gradient boosting tree, or a gradient boosting machine. Exemplarily, the weakness level of each node is obtained using a gradient boosting tree. For each node, the server can build multiple decision trees based on the gradient boosting tree, and each tree learns the residual (i.e., error) of the previous tree. When each node is split, the feature and threshold that maximize the variance gain are selected to obtain a suitable threshold, and the weakness level of each node is determined according to the threshold and the loss function value corresponding to the gradient boosting tree of each node.
[0036] In an optional embodiment, after obtaining the type data of each node, a feature multiplier is determined based on a preset initial connection weight, a target feature, and a preset initial connection threshold; the feature multiplier is multiplied by the weakness probability of the current node to obtain the weakness of the current node, and the weakness of the current node is quantified to obtain the weakness level of the current node.
[0037] Step 104: Optimize the initial model according to the weakness level of each node and a predetermined loss function until the initial model meets the preset conditions to obtain the target model.
[0038] Among them, the target model is used to identify weak points of the power communication network. The initial model is a model that has not been trained and optimized, and is used to identify weak points of the power communication network architecture based on multimodal data. The target model is an initial model that has been trained and has reached the preset conditions. The preset conditions can be that the value of the loss function reaches convergence or the number of iterations reaches a preset number, etc.
[0039] In an optional implementation, after acquiring the multimodal data, the server performs data preprocessing on the multimodal data, and obtains a sample data set for training the initial model based on the multimodal data. The sample data set includes sample input data and sample labels. If the initial model has not yet met the preset conditions, a data is selected from the sample input data and input into the initial model to obtain the output value of the initial model output. The output value and the sample label corresponding to the data input into the initial model are brought into the loss function for calculation, and the model parameters of the initial model are adjusted according to the loss function value, and the initial model is optimized until the initial model meets the preset conditions to obtain the target model.
[0040] In this scheme, optionally, the initial model is optimized according to the weakness level of each node and a predetermined loss function, including: determining the target type data and target weakness level of each node based on the multimodal data, and obtaining the first loss value and the second loss value of each node based on the loss function, the type data and the weakness level of each node; for each node, adjusting the classification parameters of the classification module based on the first loss value of the current node, and adjusting the recognition parameters of the recognition module according to the second loss value of the current node.
[0041] Among them, the target weakness level is the weakness level to which each node belongs, which is predetermined based on the multimodal data. The type data of each node is the data output by the classification module, and the weakness level of each node is the data output by the recognition module. In an optional embodiment, after obtaining the type data and weakness level of the current node, the first loss value of the current node is calculated based on the type data, the target type data and the loss function, and the classification parameters are adjusted based on the first loss value. The second loss value of the current node is calculated based on the weakness level, the target weakness level and the loss function, and the recognition parameters of the recognition module are adjusted based on the second loss value, so as to optimize the initial model.
[0042] In the above steps, the initial model can be continuously adaptively learned and optimized through multimodal data to improve the accuracy of identifying weak points of the target model.
[0043] The technical solution of this embodiment, when the pre-established initial model does not meet the preset conditions, obtain the multimodal data of each network device of the power communication network; wherein the initial model includes a classification module and an identification module; the multimodal data includes the connection relationship and topological structure of the network equipment; the multimodal data is processed according to the classification module to obtain the type data of each node of the grid structure of the power communication network; the type data of each node is calculated according to the identification module to obtain the weakness level of each node; the initial model is optimized according to the weakness level of each node and the predetermined loss function until the initial model meets the preset conditions to obtain the target model; wherein the target model is used to identify the weak points of the power communication network. The technical solution of this embodiment can collect multimodal data of the power communication network, so as to obtain the comprehensive information of each network device, and use the comprehensive information to continuously optimize the classification module and the identification module to obtain the target model that can accurately identify the weak points of the power communication network architecture. Therefore, the target model is used to accurately identify the weak points of the power communication network architecture, thereby improving the efficiency and accuracy of identifying the weak points of the power communication network architecture.
[0044] Figure 2 The second flow chart of a method for identifying weak points in a power communication grid provided by an embodiment of the present invention is a refinement of the above embodiment. The specific method can be as follows Figure 2 As shown, the method may include the following steps:
[0045] Step 201: When the pre-established initial model does not meet the preset conditions, multimodal data of each network device of the power communication network is obtained.
[0046] Among them, the initial model includes a classification module and a recognition module; the multimodal data includes the connection relationship and topological structure of network devices.
[0047] Step 202: extract features from the multimodal data to obtain feature information of each node.
[0048] Among them, the characteristic information includes attribute information entropy characteristics, inflow-outflow ratio characteristics, average flow length characteristics and unilateral connection density characteristics. Attribute information entropy is an indicator to measure the degree of data chaos, which can reflect the uncertainty of data. Inflow-outflow ratio is an indicator to describe the direction of network traffic, which is used to identify abnormal traffic patterns in the network. Average flow length refers to the average number of packets contained in the data flow, which is used to reflect the characteristics of the data flow. Unilateral connection density refers to the percentage of data packets that have not established normal communication connections in the Internet of Things to the total number of data packets, which is used to evaluate the stability and reliability of network connections.
[0049] In an optional implementation, after obtaining the multimodal data, four attributes of the source Internet Protocol (IP) address, destination IP address, source port, and destination port of each network device are obtained from the multimodal data, and information entropy is calculated based on the four attributes, as shown in the following formula:
[0050]
[0051] Among them, H(a i ) is the information entropy of the i-th feature attribute, a i (t) is the frequency of occurrence of the i-th feature attribute within a period of time; A(t) is the sum of the frequencies of occurrence of the i-th feature attribute within a period of time; n is the feature attribute quantity, t represents time, and T is the total time length.
[0052] Inflow-outflow ratio B j (t) is calculated as follows:
[0053]
[0054] Where b j (t) is the amount of data flowing into terminal channel j during a period of time; is the amount of data flowing out of terminal channel j within a period of time.
[0055] The average flow length is calculated as follows:
[0056]
[0057] In the formula, C j (t) is the average flow length of the data flow of terminal channel j over a period of time; c j (t) and They are the total number of packets flowing into terminal channel j and the total number of packets flowing out of terminal channel j within a period of time, respectively.
[0058] The unilateral connection density is the number of data packets m that do not establish a normal communication connection in the Internet of Things. j (t) accounts for the total number of packets M j The calculation formula is as follows:
[0059]
[0060] Among them, D j (t) is the one-sided connection density.
[0061] Step 203: normalize the feature information of each node to obtain the target feature of each node.
[0062] Specifically, after obtaining the feature information of each node, since the dimensions of different features are different, it is impossible to perform unified operations, so the extracted feature values need to be normalized. Normalization is a method of converting data to a uniform scale, as shown in the following formula:
[0063]
[0064] Among them, x is the original data, x min and x max are the minimum and maximum values corresponding to the original data, respectively, x n is the normalized data (target feature).
[0065] Step 204: Process the target features of each node according to the classification module to obtain type data of each node.
[0066] Among them, the types of each node are divided into two categories, namely weak nodes and non-weak nodes. The type data includes the weakness probability of each node, and the type of each node can be determined according to the weakness probability of each node. After obtaining the target features of each node, the pre-processed multimodal data is input into the classification module, which includes a pre-defined classification algorithm. The classification algorithm is used to classify the pre-processed multimodal data to obtain the type data of each node of the grid structure.
[0067] In this scheme, optionally, the target features of each node are processed according to the classification module to obtain the type data of each node, including: for the target features of each node, determining the expected output corresponding to the target features of the current node based on the multimodal data; performing classification operations on the target features of the current node based on the predetermined Lagrange multiplier, the expected output and the symbolic function of the classification module to obtain the type data of the current node.
[0068] Among them, the expected output is the result of predicting or classifying the input data (target feature), and the expected output is used to guide the learning process of the initial model. After obtaining the multimodal data, the expected output corresponding to each feature can be obtained by manual annotation and other methods. After obtaining the expected output, the target feature of the current node is classified according to the predetermined Lagrange multiplier, the expected output and the symbolic function of the classification module to obtain the probability of the current node belonging to each node type (i.e., weak probability), Q i (x) represents the type probability of node i, as shown in the following formula:
[0069] Q i (x) = sgn(v i ·x i ·H(x i )+b i H(xi ))
[0070] Where H() is the dot product operation; x i is the characteristic value of the target feature, b is the expected output; v i is the Lagrange multiplier. The types of nodes are divided into two categories, and the probability values of the two categories are described by (Q1, Q2). Q1 and Q2 represent the probability value of non-weak points (1, 0) and the probability value of weak points (0, 1), respectively. That is, the closer the probability of a category is to 1, the more likely the node belongs to that category.
[0071] In the above steps, each node can be accurately classified through the binary classification algorithm to obtain the type of each node, which lays the foundation for the subsequent determination of the weakness level of each node.
[0072] Step 205: Calculate the weakness of the type data of each node according to the identification module to obtain the weakness level of each node.
[0073] Among them, the recognition module includes a predefined recognition method, and the recognition method is used to determine the weakness of each node, that is, the weakness level. The larger the weakness level, the weaker the node. The recognition method can be a deep learning algorithm, a gradient boosting tree or a gradient boosting machine. In this scheme, optionally, the weakness of the type data of each node is calculated according to the recognition module to obtain the weakness level of each node, including: for each node, determining a feature multiplier based on a preset initial connection weight, a target feature and a preset initial connection threshold; multiplying the feature multiplier and the weakness probability of the current node to obtain the weakness of the current node, and quantifying the weakness of the current node to obtain the weakness level of the current node.
[0074] Among them, the initial connection weight and the initial connection threshold are predetermined based on domain big data, etc. According to the preset initial connection weight, target feature and the preset initial connection threshold, the feature multiplier can be determined as: Where n is the total number of features, w i represents the connection weight; x i represents the target feature; r represents the connection threshold.
[0075] Furthermore, the feature multiplier is multiplied by the weakness probability of the current node to obtain the weakness of the current node:
[0076]
[0077] Among them, F i (x) is the degree of weakness, Q i(x) is the weakness probability of the node. Specifically, the weakness (weakness) of a node is a description of the state of a node. The result is quantified by the probability value, and the weakness is divided into three levels, namely relatively weak (0, 0, 1), obviously weak (0, 1, 0) and severely weak (1, 0, 0). The weakness level corresponding to the probability of the weak point with the largest category probability (the closer to 1) among the three categories is determined as the weakness level to which the weak point belongs.
[0078] In the above steps, the complex patterns and associated data that may exist in the power communication network can be identified through the recognition algorithm, so as to obtain the weakness of each node. The recognition algorithm of this solution can process multiple data points at the same time, which helps to improve the efficiency of processing multi-modal data of the power communication grid structure.
[0079] Step 206: Optimize the initial model according to the weakness level of each node and a predetermined loss function until the initial model meets the preset conditions to obtain the target model.
[0080] Among them, the target model is used to identify weak points in the power communication network.
[0081] In the technical solution of this embodiment, when the pre-established initial model does not meet the preset conditions, the multimodal data of each network device of the power communication network is obtained. Among them, the initial model includes a classification module and an identification module; the multimodal data includes the connection relationship and topological structure of the network equipment; the multimodal data is feature extracted to obtain the feature information of each node. The feature information includes attribute information entropy features, inflow and outflow ratio features, average flow length features and unilateral link density features. The feature information of each node is normalized to obtain the target feature of each node. The target features of each node are processed according to the classification module to obtain the type data of each node. The type data of each node is weakly calculated according to the identification module to obtain the weakness level of each node. The initial model is optimized according to the weakness level of each node and a predetermined loss function until the initial model meets the preset conditions to obtain the target model; wherein the target model is used to identify the weak points of the power communication network. The technical solution of this embodiment can extract features from multimodal data, providing a basis for subsequent improvement of model performance. Through the binary classification algorithm, each node is accurately classified to obtain the type to which each node belongs. Through the recognition algorithm, the complex patterns and associated data that may exist in the power communication network are identified, thereby obtaining the weakness of each node. By optimizing the initial model based on the type of each node and the weakness of each node, the accuracy and work efficiency of identifying weak points of the target model can be improved.
[0082] Figure 3The flowchart of using the target model to identify the weak points of the power communication grid provided by the embodiment of the present invention. The specific method can be as follows Figure 3 As shown, the method may include the following steps:
[0083] Step 301: Obtain the data to be analyzed from the power communication network at the current moment.
[0084] The data to be analyzed is data directly obtained by the server or uploaded by the user, and needs to identify weaknesses. The server can obtain the data to be analyzed of the power communication network in real time through a pre-set interface. Alternatively, the user can directly upload the data to be analyzed to the server so that the server can obtain the data to be analyzed.
[0085] Step 302: input the data to be analyzed into the target model to obtain the weakness level of each node of the grid structure of the power communication network output by the target model.
[0086] Among them, the target model is used to identify weak points of the power communication network. After obtaining the data to be analyzed, the server can input the data to be analyzed into the target model. The classification module of the target model is used to classify each node to obtain the type data of each node. The identification module is used to calculate the weakness of the type data of each node to obtain the weakness of each node.
[0087] Step 303: Perform visual analysis on the weakness level of each node, and display the visual analysis results to the user.
[0088] Specifically, after obtaining the weakness level of each node, the weakness level of each node is visualized and analyzed according to the predefined analysis rules to obtain the visualization analysis results of multimodal data. The visualization analysis results include the weakness level of each node, anomaly detection information, and recommendations and action steps for predictive maintenance. Combined with multimodal data, the visualization analysis results are presented to users (operation and maintenance personnel) through visualization and interactive interfaces. The display methods include data charts, images, sound and vibration waveforms, etc., so that operation and maintenance personnel can intuitively view the operating status and abnormal conditions of network equipment, so as to quickly discover abnormal modes of network equipment, identify possible problems with network equipment, and take corresponding measures to repair or maintain them, so as to effectively manage the operation and maintenance of network equipment.
[0089] The technical solution of this embodiment obtains the data to be analyzed of the power communication network at the current moment. The data to be analyzed is input into the target model to obtain the weak nodes and the degree of weakness of the weak nodes in all nodes of the grid structure of the power communication network output by the target model. The weak nodes and the degree of weakness of the weak nodes are visually analyzed, and the visual analysis results are displayed to the user. The technical solution of this embodiment can use the target model to accurately identify the weak points of the power communication network architecture, thereby improving the efficiency and accuracy of identifying the weak points of the power communication network architecture. The visual analysis results are displayed to the user, so that the user can intuitively view the operating status and abnormal conditions of the network equipment, thereby quickly discovering the abnormal mode of the network equipment, identifying possible problems with the network equipment, and taking corresponding measures to repair or maintain it, and effectively managing the operation and maintenance of the network equipment.
[0090] An embodiment of the present invention also provides a computer program product.
[0091] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer program products, which can include one or more computer programs, which can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor, which can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0092] Figure 4 The present invention provides a schematic diagram of a structure of a power communication grid weak point identification device, which is suitable for executing a power communication grid weak point identification method provided by an embodiment of the present invention. Figure 4 As shown, the device may specifically include:
[0093] The data acquisition module 401 is used to acquire multimodal data of each network device of the power communication network when the pre-established initial model does not meet the preset conditions; wherein the initial model includes a classification module and an identification module; the multimodal data includes the connection relationship and topological structure of the network devices;
[0094] A data classification module 402, configured to process the multimodal data according to the classification module to obtain type data of each node of the grid structure of the power communication network;
[0095] A level determination module 403, configured to calculate the weakness of the type data of each node according to the identification module to obtain the weakness level of each node;
[0096] The model optimization module 404 is used to optimize the initial model according to the weakness level of each node and a predetermined loss function until the initial model meets the preset conditions to obtain a target model; wherein the target model is used to identify weak points of the power communication network.
[0097] Optionally, the data classification module 402 is specifically used to: extract features from the multimodal data to obtain feature information of each node, wherein the feature information includes attribute information entropy features, inflow-outflow ratio features, average flow length features, and unilateral link density features;
[0098] Normalizing the feature information of each node to obtain a target feature of each node;
[0099] The target features of the nodes are processed according to the classification module to obtain type data of the nodes.
[0100] Optionally, the data classification module 402 is further used to: determine, for each target feature of each node, an expected output corresponding to the target feature of the current node based on the multimodal data;
[0101] Based on the predetermined Lagrange multiplier, the expected output and the symbolic function of the classification module, a classification operation is performed on the target feature of the current node to obtain the type data of the current node.
[0102] Optionally, the level determination module 403 is specifically configured to: determine a feature multiplier based on a preset initial connection weight, the target feature and a preset initial connection threshold;
[0103] The characteristic multiplier and the weakness probability of the current node are multiplied to obtain the weakness degree of the current node, and the weakness degree of the current node is quantified to obtain the weakness grade of the current node.
[0104] Optionally, the model optimization module 404 is specifically configured to: determine the target type data and the target weakness level of each node according to the multimodal data, and obtain a first loss value and a second loss value of each node according to the loss function, the type data of each node and the weakness level of each node;
[0105] For each node, the classification parameters of the classification module are adjusted based on the first loss value of the current node, and the recognition parameters of the recognition module are adjusted based on the second loss value of the current node.
[0106] Optionally, the model optimization module 404 is further used to: obtain the data to be analyzed of the power communication network at the current moment;
[0107] Inputting the data to be analyzed into the target model to obtain the weakness level of each node of the grid structure of the power communication network output by the target model;
[0108] The weakness level of each node is visually analyzed, and the visual analysis result is displayed to the user.
[0109] The power communication grid weak point identification device provided in the embodiment of the present invention can execute the power communication grid weak point identification method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method. The contents not described in detail in this embodiment can refer to the description in any method embodiment of the present invention.
[0110] Figure 5 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, referring to Figure 5 , Figure 5 The electronic device 12 shown is only an example and should not limit the functions and scope of use of the embodiments of the present application. Figure 5 As shown, the electronic device 12 is in the form of a general purpose computing device. The components of the electronic device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 that connects various system components (including the system memory 28 and the processing unit 16).
[0111] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.
[0112] The electronic device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 12, including volatile and non-volatile media, removable and non-removable media.
[0113] The system memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Figure 5 not shown, usually called a "hard drive"). Although Figure 5 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical medium) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data medium interfaces. The memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present application.
[0114] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in the memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. The program modules 42 generally perform the functions and / or methods of the embodiments described herein.
[0115] The electronic device 12 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), may communicate with one or more devices that enable a user to interact with the electronic device 12, and / or may communicate with any device that enables the electronic device 12 to communicate with one or more other computing devices (e.g., a network card, a modem, etc.). Such communication may be performed via an input / output (I / O) interface 22. Furthermore, the electronic device 12 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 20. As shown, the network adapter 20 communicates with other modules of the electronic device 12 via the bus 18. It should be understood that although Figure 5 Not shown, other hardware and / or software modules may be used in conjunction with the electronic device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
[0116] The processing unit 16 executes various functional applications and data processing by running the program stored in the system memory 28, for example, to implement a method for identifying weak points in a power communication grid provided in an embodiment of the present invention: when a pre-established initial model does not meet the preset conditions, multimodal data of each network device of the power communication network is obtained; wherein the initial model includes a classification module and an identification module; the multimodal data includes the connection relationship and topological structure of the network devices; the multimodal data is processed according to the classification module to obtain type data of each node of the grid structure of the power communication network; the weakness of the type data of each node is calculated according to the identification module to obtain the weakness level of each node; the initial model is optimized according to the weakness level of each node and a predetermined loss function until the initial model meets the preset conditions to obtain a target model; wherein the target model is used to identify weak points in the power communication network.
[0117] The embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, a method for identifying weak points in a power communication grid provided by all the embodiments of the present invention is implemented: when the pre-established initial model does not meet the preset conditions, multimodal data of each network device of the power communication network is obtained; wherein the initial model includes a classification module and an identification module; the multimodal data includes the connection relationship and topological structure of the network device; the multimodal data is processed according to the classification module to obtain the type data of each node of the grid structure of the power communication network; the type data of each node is calculated according to the identification module to obtain the weakness level of each node; the initial model is optimized according to the weakness level of each node and a predetermined loss function until the initial model meets the preset conditions to obtain a target model; wherein the target model is used to identify weak points in the power communication network. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electronic device, device or device of electrical, magnetic, optical, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection with one or more conductors, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction-executing electronic device, apparatus, or device.
[0118] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction-executing electronic device, apparatus, or device.
[0119] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
[0120] Computer program code for performing the operations of the present invention may be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0121] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims
1. A method for identifying weak points in power communication grids. It is characterized in that The method comprises: When the pre-established initial model does not meet the preset conditions, multimodal data of each network device of the power communication network is obtained; wherein the initial model includes a classification module and an identification module; the multimodal data includes a connection relationship and a topological structure of the network devices; Processing the multimodal data according to the classification module to obtain type data of each node of the grid structure of the power communication network; Performing weakness calculation on the type data of each node according to the identification module to obtain a weakness level of each node; The initial model is optimized according to the weakness level of each node and a predetermined loss function until the initial model meets the preset conditions to obtain a target model; wherein the target model is used to identify weak points of the power communication network.
2. The method according to claim 1, It is characterized in that The multimodal data is processed according to the classification module to obtain type data of each node of the grid structure of the power communication network, including: Performing feature extraction on the multimodal data to obtain feature information of each node, wherein the feature information includes attribute information entropy feature, inflow-outflow ratio feature, average flow length feature, and unilateral link density feature; Normalizing the feature information of each node to obtain a target feature of each node; The target features of the nodes are processed according to the classification module to obtain type data of the nodes.
3. The method according to claim 2, It is characterized in that Processing the target features of each node according to the classification module to obtain type data of each node includes: For each target feature of each node, determining an expected output corresponding to the target feature of the current node based on the multimodal data; Based on the predetermined Lagrange multiplier, the expected output and the sign function of the classification module, a classification operation is performed on the target feature of the current node to obtain the type data of the current node.
4. The method according to claim 1, It is characterized in that The type data includes a weakness probability, and the type data of each node is subjected to weakness calculation according to the identification module to obtain a weakness level of each node, including: Determine a feature multiplier based on a preset initial connection weight, the target feature and a preset initial connection threshold; The characteristic multiplier and the weakness probability of the current node are multiplied to obtain the weakness degree of the current node, and the weakness degree of the current node is quantified to obtain the weakness grade of the current node.
5. The method according to claim 1, It is characterized in that Optimizing the initial model according to the weakness level of each node and a predetermined loss function, including: Determine the target type data and the target weakness level of each node according to the multimodal data, and obtain a first loss value and a second loss value of each node according to the loss function, the type data of each node and the weakness level of each node; For each node, the classification parameters of the classification module are adjusted based on the first loss value of the current node, and the recognition parameters of the recognition module are adjusted based on the second loss value of the current node.
6. The method according to claim 1, It is characterized in that The method further comprises: Acquire the data to be analyzed of the power communication network at the current moment; Inputting the data to be analyzed into the target model to obtain the weakness level of each node of the grid structure of the power communication network output by the target model; The weakness level of each node is visually analyzed, and the visual analysis result is displayed to the user.
7. A computer program product comprising a computer program, It is characterized in that When the computer program is executed by a processor, the computer program implements a power communication grid weak point method according to any one of claims 1 to 6.
8. A device for identifying weak points in a power communication grid. It is characterized in that include: A data acquisition module, used for acquiring multimodal data of each network device of the power communication network when the pre-established initial model does not meet the preset conditions; wherein the initial model includes a classification module and an identification module; the multimodal data includes the connection relationship and topological structure of the network devices; A data classification module, used for processing the multimodal data according to the classification module to obtain type data of each node of the grid structure of the power communication network; A level determination module, used to calculate the weakness of the type data of each node according to the identification module to obtain the weakness level of each node; A model optimization module is used to optimize the initial model according to the weakness level of each node and a predetermined loss function until the initial model meets the preset conditions to obtain a target model; wherein the target model is used to identify weak points of the power communication network.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the program, the method for identifying weak points in a power communication grid as described in any one of claims 1 to 6 is implemented.
10. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the program is executed by a processor, a method for identifying weak points in a power communication grid as described in any one of claims 1 to 6 is implemented.
