A processing method, device, electronic device, computer-readable storage medium and computer program product for a distribution network simulation model

Abstract

The present application provides a processing method, device, electronic device, computer-readable storage medium and computer program product for a distribution network simulation model; the method includes: constructing a graph structure based on the topological relationship between various devices in the distribution network simulation model, wherein each node in the graph structure corresponds to a device, and the device is a power device or a power-consuming device; determining the characteristics of the node based on the difference between the real circuit parameters and the simulated circuit parameters of the device corresponding to each node; aggregating the characteristics of each node included in the graph structure to obtain the aggregated characteristics of each node; determining the problem device in the distribution network simulation model based on the aggregated characteristics of each node. Through the present application, the verification efficiency of the equipment in the distribution network simulation model can be effectively improved, thereby improving the simulation efficiency.

Description

Technical Field

[0001] The present application relates to the field of system simulation technology, and in particular to a method, device, electronic device, computer-readable storage medium and computer program product for processing a distribution network simulation model. Background Art

[0002] In the related technology, operators obtain relevant information of various equipment such as switches, transformers, capacitors, etc. in the distribution network from the power supply bureau, including equipment model, equipment working principle, equipment parameters and other information. According to the equipment information, operators add these devices to the distribution network line model and set the equipment properties, thereby preliminarily building a distribution network simulation model, and then test whether the distribution network simulation model conforms to the actual distribution network situation. If there is a large difference between the distribution network simulation model and the actual distribution network situation, it is necessary to check the information of each device in the distribution network simulation model one by one, and even go to the location of the equipment for on-site verification. The verification efficiency is low, which in turn affects the simulation efficiency. Summary of the invention

[0003] The embodiments of the present application provide a method, device, electronic device, computer-readable storage medium and computer program product for processing a distribution network simulation model, which can effectively improve the efficiency of checking device information in the distribution network simulation model, thereby improving the simulation efficiency.

[0004] The technical solution of the embodiment of the present application is implemented as follows:

[0005] The present application provides a method for processing a distribution network simulation model, including:

[0006] Based on the topological relationship between the various devices in the distribution network simulation model, a graph structure is constructed, wherein each node in the graph structure corresponds to one of the devices, and the device is a power device or a power-consuming device;

[0007] Determine the characteristics of the node based on the difference between the real circuit parameters and the simulated circuit parameters of the device corresponding to each of the nodes;

[0008] Aggregating the features of each of the nodes included in the graph structure to obtain aggregated features of each of the nodes;

[0009] Based on the aggregated features of each of the nodes, problematic devices in the distribution network simulation model are determined.

[0010] The present application provides a processing device for a distribution network simulation model, including:

[0011] A construction module, used to construct a graph structure based on the topological relationship between various devices in the distribution network simulation model, wherein each node in the graph structure corresponds to one of the devices, and the device is a power device or a power-consuming device;

[0012] A determination module, configured to determine the characteristics of each node based on a difference between a real circuit parameter and a simulated circuit parameter of a device corresponding to each node;

[0013] An aggregation module, used to aggregate the features of each of the nodes included in the graph structure to obtain aggregated features of each of the nodes;

[0014] The determination module is further used to determine problematic devices in the distribution network simulation model based on the aggregated features of each of the nodes.

[0015] In the above scheme, the real circuit parameters include real voltage and real current, and the simulated circuit parameters include simulated voltage and simulated current. For each node included in the graph structure, the determination module is also used to obtain the real voltage and real current of the device corresponding to the node, and obtain the simulated voltage and simulated current of the device corresponding to the node; determine the voltage difference between the real voltage and the simulated voltage, and determine the current difference between the real current and the simulated current; splice the voltage difference and the current difference, and use the splicing result as the feature of the node.

[0016] In the above scheme, the determination module is also used to obtain the real data of the distribution network, wherein the real data of the distribution network includes the real export voltage and the real export current of the substation; the real export voltage and the real export current of the substation are input into the distribution network simulation model, and simulation processing is performed to obtain the simulated voltage and simulated current of the equipment corresponding to the node in the distribution network simulation model.

[0017] In the above scheme, the aggregation module is also used to determine the aggregation characteristics of the node corresponding to the power equipment in the graph structure based on the characteristics of the neighboring nodes of the node; for the node corresponding to the power equipment in the graph structure, the characteristics of the node within a set time length are aggregated to obtain the aggregation characteristics of the node.

[0018] In the above scheme, the aggregation module is also used to call the graph neural network model to perform the following processing: obtain the first-order neighbor nodes and second-order neighbor nodes of the node; aggregate the features of the second-order neighbor nodes to obtain aggregate features; add the aggregate features and the features of the first-order neighbor nodes to obtain the updated features of the first-order neighbor nodes; add the updated features of the first-order neighbor nodes and the features of the node to obtain the aggregate features of the node.

[0019] In the above scheme, for each of the nodes included in the graph structure, the determination module is also used to call the trained classifier for prediction based on the aggregation characteristics of the node to obtain the probability that the node is a problem node; when the probability is greater than the probability threshold, the device corresponding to the node in the distribution network simulation model is determined as a problem device.

[0020] In the above scheme, the device also includes a model training module, which is also used to obtain the aggregated features of the sample nodes, wherein the aggregated features of the sample nodes are obtained by calling the initialized graph neural network model to aggregate the features of the sample nodes in the sample graph structure; based on the aggregated features of the sample nodes, the initialized classifier is called to make a prediction to obtain the probability that the sample node belongs to the problem node; the difference between the probability output by the initialized classifier and the probability pre-marked for the sample node is determined; back propagation is performed based on the difference, and the parameters of the classifier and the graph neural network model are updated during the back propagation process.

[0021] In the above scheme, the device also includes a correction module, and the correction module is further used to obtain the real information of the problematic device when the original information of the problematic device is erroneous information, and to correct the original information of the problematic device based on the real information.

[0022] In the above scheme, when the differences between the real circuit parameters and the simulated circuit parameters of each device in the distribution network simulation model are less than the difference threshold, the determination module is also used to impose a fault on the distribution network simulation model according to the distribution network operation standard, and determine the operation data of each device in the distribution network simulation model; for each of the devices, determine the protection setting of the device based on the operation data of the device and the relay protection configuration logic.

[0023] An embodiment of the present application provides an electronic device, including:

[0024] A memory for storing computer executable instructions;

[0025] The processor is used to implement the processing method of the distribution network simulation model provided in the embodiment of the present application when executing the computer executable instructions stored in the memory.

[0026] An embodiment of the present application provides a computer-readable storage medium storing computer-executable instructions for implementing a processing method for a distribution network simulation model provided in an embodiment of the present application when executed by a processor.

[0027] An embodiment of the present application provides a computer program product, including a computer program or computer executable instructions, which are used to implement the processing method of the distribution network simulation model provided in the embodiment of the present application when executed by a processor.

[0028] The embodiments of the present application have the following beneficial effects:

[0029] First, according to the topological relationship of each device in the distribution network simulation model, a graph structure is established, and the characteristics of each node are determined according to the difference between the real circuit parameters and the simulated circuit parameters of the device corresponding to each node, and then the characteristics of each node in the graph structure are aggregated to obtain the aggregated characteristics of each node. Finally, according to the aggregated characteristics of each node, the problem node in the distribution network simulation model is determined. In this way, by establishing the graph structure relationship of each device in the distribution network simulation model and obtaining the aggregated characteristics of the device corresponding to each node in the graph structure, the problem device in the distribution network simulation model can be identified through the aggregated characteristics, without manually checking the information of each device one by one, which can effectively improve the checking efficiency of the equipment in the distribution network simulation model, thereby improving the simulation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application;

[0031] Figure 2 It is a first flow chart of a method for processing a distribution network simulation model provided in an embodiment of the present application;

[0032] Figure 3 This is a schematic diagram of the result of constructing a graph structure based on topological relationships provided by an embodiment of the present application;

[0033] Figure 4 It is a second flow chart of the method for processing the distribution network simulation model provided in the embodiment of the present application;

[0034] Figure 5 It is a diagram of the structure of a processing method of a distribution network simulation model provided in an embodiment of the present application;

[0035] Figure 6 It is a third flow chart of the method for processing the distribution network simulation model provided in an embodiment of the present application. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0037] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0038] It is understandable that in the embodiments of the present application, related data such as user information is involved. When the embodiments of the present application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards.

[0039] In the following description, the terms "first\second\..." are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\..." can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0040] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0041] Before further describing the embodiments of the present application in detail, the nouns and terms involved in the embodiments of the present application are explained. The nouns and terms involved in the embodiments of the present application are subject to the following interpretations.

[0042] 1) Graph Sampling and Aggregating Network (GraphSAGE) model: The GraphSAGE model is a popular graph neural network (GNN) model, which is mainly used for graph node embedding learning. It is based on the basic concept of graph neural networks and learns the embedding representation of nodes by sampling and aggregating the features of neighboring nodes. GraphSAGE has a wide range of applications, including social network analysis, recommendation systems, knowledge graph completion, bioinformatics, etc. Its design makes it more feasible to train graph neural networks on large-scale graphs. It is a widely used model in the current field of graph neural networks.

[0043] 2) Topological relationship: Topological relationship is a concept that describes the positional relationship between spatial entities. It does not need to consider the specific shape and size of the spatial entities, but only needs to focus on the interconnection, adjacency, inclusion and connectivity between them. Topological relationship is widely used in many fields such as Geographic Information System (GIS), Computer Aided Design (CAD), and computer network design.

[0044] 3) Distribution network simulation model: The distribution network simulation model refers to a mathematical model for simulating and analyzing the distribution network (distribution network) in the power system. The distribution network is the intermediate link between the high-voltage transmission network and the user's electrical equipment, and is responsible for distributing electricity to end users. The distribution network simulation model is used to simulate the electrical performance of the distribution network under various operating conditions, including voltage, current, power, and fault conditions, etc., to facilitate grid planning, design, and operation and maintenance.

[0045] 4) Series compensation capacitor: The full name is series compensation capacitor, which is a power electronic device used to improve the performance of transmission lines in power systems. In AC transmission lines, series compensation capacitors are mainly realized by connecting capacitors of a certain capacitance in series in the transmission line.

[0046] 5) Circuit breaker: Circuit breaker is an important switch device, widely used in power systems, industrial control systems and household appliances, mainly used to connect or disconnect current in the circuit. Circuit breaker can not only control the on and off of the circuit under normal circumstances, but also automatically cut off the circuit when a circuit fails, protecting the circuit and electrical equipment from damage.

[0047] 6) Activation function: Activation function is a key component in neural networks. It is used to convert linear combinations of neurons into nonlinear outputs. In artificial neural networks, each neuron receives input from neurons in the previous layer, performs weighted summation on these inputs, and then processes them through the activation function to obtain the output of the neuron. The role of the activation function is to introduce nonlinear characteristics to the neural network, so that the neural network can learn and simulate complex nonlinear relationships.

[0048] 7) Fully Connected Layer (FCL): A fully connected layer is a type of layer in a neural network. It represents a structure of a neural network in which each neuron is connected to all neurons in the previous layer. In a fully connected layer, each neuron receives the output of all neurons in the previous layer as input and generates an output value.

[0049] Due to the complexity of the distribution network lines and the different installation locations of various devices, the relay protection settings need to be set after relevant calculations. In addition, the equipment is large and the fault types are complex. Professionals need to perform a lot of calculations, which is complicated and laborious, and prone to errors. However, different equipment has different calculation methods, so it is necessary to rely on the equipment manufacturer for calculations.

[0050] Related technologies usually use the method of simulating distribution networks and applying faults to determine protection settings for equipment in the distribution network. Specifically, first, the operator obtains relevant information about various equipment such as switches, transformers, capacitors, etc. in the distribution network from the power supply bureau, including equipment model, equipment working principle, equipment parameters and other parameter information. Then, the operator will add relevant equipment to the distribution network line model based on the obtained equipment information, and set the equipment parameters accordingly, thereby building a distribution network simulation model.

[0051] After building the distribution network simulation model, it is also necessary to test and tune the distribution network simulation model to make it as close to the actual distribution network situation as possible. The specific test and adjustment method is as follows: First, the operator obtains the real data of the distribution network at a certain moment, including the voltage and current at the outlet of the substation, and the voltage and current at each load end (i.e., power-consuming equipment); then, the operator inputs the real voltage and current at the outlet of the substation into the distribution network simulation model. Through the simulation software, the simulated voltage and current at each load end can be calculated. If the simulated voltage and current at a certain (or some) load end are significantly different from the real voltage and current, it means that there is a large difference between the distribution network simulation model and the real distribution network situation, and the distribution network simulation model needs to be adjusted; then, the operator needs to check with the personnel of the power supply bureau one by one whether the information of each device is provided accurately, and sometimes even needs to go to the equipment site to check the real information of the equipment one by one, resulting in very low simulation efficiency, which reduces the efficiency of building the distribution network simulation model and the fixed value simulation efficiency; finally, after finding the problematic device, it is necessary to re-determine the real device information of the device, so as to correct the parameter information of the device in the distribution network simulation model and optimize the distribution network simulation model. That is to say, in the related technology, when there is a large difference between the initially established distribution network simulation model and the actual distribution network situation, it is necessary to check the information of each device in the distribution network one by one, so the checking efficiency is low, which affects the simulation efficiency.

[0052] Based on this, the embodiments of the present application provide a method, device, electronic device, computer-readable storage medium and computer program product for processing a distribution network simulation model, which can effectively improve the verification efficiency of devices in the distribution network simulation model, thereby improving the simulation efficiency. The electronic device provided by the embodiments of the present application is described below. The electronic device provided by the embodiments of the present application can be implemented as various types of terminals such as laptops, tablet computers, desktop computers, etc., and can also be implemented as a server.

[0053] See also Figure 1 , Figure 1 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application, taking the electronic device as a terminal as an example, Figure 1 The electronic device 400 shown includes: at least one processor 410, a memory 450, at least one network interface 420 and a user interface 430. The various components in the electronic device 400 are coupled together via a bus system 440. It is understood that the bus system 440 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 440 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, the bus system 440 is not described in detail. Figure 1 Various buses are labeled as bus system 440 .

[0054] The processor 410 may be an integrated circuit chip having signal processing capabilities, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., wherein the general-purpose processor may be a microprocessor or any conventional processor, etc.

[0055] The user interface 430 includes one or more output devices 431 that enable presentation of media content, including one or more speakers and / or one or more visual display screens. The user interface 430 also includes one or more input devices 432, including user interface components that facilitate user input, such as a keyboard, mouse, microphone, touch screen display, camera, other input buttons and controls.

[0056] The memory 450 may be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid-state memory, hard disk drives, optical disk drives, etc. The memory 450 may optionally include one or more storage devices that are physically remote from the processor 410.

[0057] The memory 450 includes a volatile memory or a non-volatile memory, and may also include both volatile and non-volatile memories. The non-volatile memory may be a read-only memory (ROM), and the volatile memory may be a random access memory (RAM). The memory 450 described in the embodiment of the present application is intended to include any suitable type of memory.

[0058] In some embodiments, memory 450 can store data to support various operations, examples of which include programs, modules, and data structures, or a subset or superset thereof, as exemplarily described below.

[0059] Operating system 451, including system programs for processing various basic system services and performing hardware-related tasks, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and processing hardware-based tasks;

[0060] A network communication module 452, used to reach other electronic devices via one or more (wired or wireless) network interfaces 420, exemplary network interfaces 420 include: Bluetooth, Wireless Fidelity (WiFi), and Universal Serial Bus (USB);

[0061] a presentation module 453 for enabling presentation of information via one or more output devices 431 (e.g., display screen, speaker, etc.) associated with the user interface 430 (e.g., a user interface for operating peripherals and displaying content and information);

[0062] The input processing module 454 is used to detect one or more user inputs or interactions from one of the one or more input devices 432 and translate the detected inputs or interactions.

[0063] In some embodiments, the device provided in the embodiments of the present application can be implemented in software. Figure 2 The processing device 455 of the distribution network simulation model stored in the memory 450 is shown, which can be software in the form of programs and plug-ins, including the following software modules: construction module 4551, determination module 4552, aggregation module 4553, model training module 4554 and correction module 4555. These modules are logical, so they can be arbitrarily combined or further split according to the functions implemented. It should be pointed out that in Figure 1 For the sake of convenience, all the above modules are shown at one time, but it should not be considered that the processing device 455 of the distribution network simulation model excludes the implementation that can only include the construction module 4551, the determination module 4552 and the aggregation module 4553. The functions of each module will be explained below.

[0064] In other embodiments, the processing device of the distribution network simulation model provided in the embodiments of the present application can be implemented in hardware. As an example, the device provided in the embodiments of the present application can be a processor in the form of a hardware decoding processor, which is programmed to execute the processing method of the distribution network simulation model provided in the embodiments of the present application. For example, the processor in the form of a hardware decoding processor can adopt one or more application specific integrated circuits (Application Specific Integrated Circuit, ASIC), digital signal processors (Digital Signal Processor, DSP), programmable logic devices (Programmable Logic Device, PLD), complex programmable logic devices (Complex Programmable Logic Device, CPLD), field programmable gate arrays (Field-Programmable Gate Array, FPGA) or other electronic components.

[0065] The following describes the processing method of the distribution network simulation model provided by the embodiment of the present application. As mentioned above, the electronic device that implements the processing method of the distribution network simulation model of the embodiment of the present application can be a terminal or a server. Therefore, the execution subject of each step will not be repeatedly described below.

[0066] See also Figure 2 , Figure 2 is a first flow chart of a method for processing a distribution network simulation model provided in an embodiment of the present application, which will be combined with Figure 2 The steps shown are explained.

[0067] In step 101, a graph structure is constructed based on the topological relationship between various devices in the distribution network simulation model.

[0068] Here, each node in the graph structure may correspond to a device, wherein the device may be a power device or a power-consuming device.

[0069] In some embodiments, nodes and edges in a graph structure are first defined, where nodes are used to represent power equipment nodes (e.g., circuit breakers, transformers, capacitors, etc.) or power consumption nodes (e.g., each household electricity user can be regarded as a power consumption node) in a power distribution network, and edges are used to represent the connection relationship between nodes (e.g., connected through power lines); then, each device in the power distribution network is mapped to a node, where each node contains parameter information such as a node identifier (IDentifier, ID) (used to uniquely identify each device), a node type (e.g., a transformer, circuit breaker, etc.), and node attributes (e.g., voltage, current, capacity, etc.); then, based on the connection relationship between devices in a distribution network simulation model, edges between nodes are constructed, where each edge includes a starting node of the edge. Points (used to represent two connected device nodes), edge types (such as physical connections, control connections, etc.), edge attributes (such as line impedance, capacity, etc.) and other parameter information; then, the topological relationship of the distribution network is represented by a graph data structure. When the connection relationship of the distribution network is undirected, that is, the direction of current flow is not considered, an undirected graph can be used to represent the topological relationship of the distribution network. When the connection relationship of the distribution network is directed, that is, the current has a specific flow direction, a directed graph can be used to represent the topological relationship of the distribution network; finally, an empty graph structure is created, all devices in the distribution network simulation model are traversed, a node is created for each device, and it is added to the graph structure. According to the connection relationship between the devices, corresponding edges are created between the nodes in the graph to obtain the corresponding constructed graph structure.

[0070] For example, see Figure 3 , Figure 3 is a schematic diagram of the result of constructing a graph structure based on a topological relationship provided in an embodiment of the present application, such as Figure 3 As shown, a graph structure can be established based on the topological relationship between each power equipment node (circuit breaker, transformer, capacitor) and each power consumption node (each household can be regarded as a power consumption node) in the distribution network simulation model. Each node in the graph structure represents a power equipment node or a power consumption node. If two nodes are directly connected through a power line, a connection line is set between the two nodes, that is, Figure 3 The line between two nodes in the graph structure represents the existence of a power line between the two nodes.

[0071] In step 102, the characteristics of each node are determined based on the difference between the real circuit parameters and the simulated circuit parameters of the device corresponding to each node.

[0072] It should be noted that circuit parameters may include resistance, inductance, current, voltage, capacitance, power factor, loss factor and other parameters, which are not specifically limited here.

[0073] In some embodiments, real circuit parameters and simulated circuit parameters of each node are obtained, wherein the real circuit parameters are obtained through on-site measurements, and the simulated parameters are calculated by a distribution network simulation model according to equipment characteristics and network configurations; then, a difference calculation is performed on the circuit parameters of each node to determine the parameter difference, where the parameter difference can be an absolute difference or a relative difference; finally, the node characteristics are determined based on the parameter difference, where the characteristic representation can be a difference value feature, a standardized feature, or a classified feature, which is not specifically limited here.

[0074] For example, assume that there is a node M, and the circuit parameters corresponding to node M include resistance, inductance and capacitance. The actual value of the resistance of node M is 10 ohms (Ω), and the simulation value is 9Ω. The absolute difference is 1Ω, and the relative difference is 0.1. The actual value of the inductance is 5 millihenry (mH), and the simulation value is 5.5mH. The absolute difference is 0.5mH, and the relative difference is 0.1. The actual value of the capacitance is 20 microfarads (μF), and the simulation value is 18μF. The absolute difference is 2μF, and the relative difference is 0.1. According to the circuit parameters of the above node M, a feature vector N of [1, 0.1, 0.5, 0.1, 2, 0.1] can be constructed, where each element represents the absolute difference or relative difference of a circuit parameter.

[0075] In some embodiments, the above-mentioned real circuit parameters may include real voltage and real current, and the simulated circuit parameters include simulated voltage and simulated current. Then the above-mentioned step 102 can be implemented in the following manner: for each node included in the graph structure, the following processing is performed respectively: obtain the real voltage and real current of the device corresponding to the node, and obtain the simulated voltage and simulated current of the device corresponding to the node; determine the voltage difference between the real voltage and the simulated voltage, and determine the current difference between the real current and the simulated current; splice the voltage difference and the current difference, and use the splicing result as the feature of the node.

[0076] In some embodiments, first, the real voltage and real current of the equipment corresponding to the node are obtained through sensors or detection systems installed on site, and the simulated voltage and simulated current of the equipment corresponding to the node are obtained through a distribution network simulation model; then, the collected real data and simulated data are mapped to the nodes to ensure that each device has a clear association with the node corresponding to it in the graph structure; then, the voltage difference between the real voltage and the simulated voltage and the current difference between the real current and the simulated current are calculated for each node; then, the calculated voltage difference and current difference are spliced ​​into a feature vector, and the spliced ​​feature vector is used as the node feature.

[0077] For example, first, obtain the real data of the distribution network at a certain moment, including the outlet voltage and outlet current of the substation at that moment, the voltage and current of each power equipment node at that moment, and the voltage and current of each power consumption node at that moment; then, input the outlet voltage and outlet current of the substation at that moment into the initially established distribution network simulation model, and run the simulation software to obtain the simulated voltage and simulated current at each node calculated by the simulation software; for each node, calculate the voltage difference percentage between the simulated voltage and the real voltage, and calculate the current difference percentage between the simulated current and the real current. If the voltage difference percentage of each node is lower than the threshold (for example, 2%), and the current difference percentage of each node is also lower than the threshold, it is considered that the distribution network simulation model is successfully built; otherwise, concatenate the voltage difference percentage and the current difference percentage to obtain the node feature vector of the node.

[0078] In some embodiments, the above-mentioned acquisition of the simulated voltage and simulated current of the equipment corresponding to the node can be implemented in the following way: obtaining the real data of the distribution network, wherein the real data of the distribution network includes the real export voltage and the real export current of the substation; inputting the real export voltage and the real export current of the substation into the distribution network simulation model, and performing simulation processing to obtain the simulated voltage and simulated current of the equipment corresponding to the node in the distribution network simulation model.

[0079] In some embodiments, first, the real export voltage and export current data are obtained from the substation through measuring equipment; then, the collected data is preprocessed (such as time synchronization, unit conversion, noise filtering, etc.); then, the preprocessed real export voltage and export current data are passed to the distribution network simulation model as input parameters; then, the distribution network simulation model is started, and the input real data is used as the initial condition for simulation calculation. The model will calculate the simulated voltage and simulated current of the equipment corresponding to each node in the network according to the topological structure, equipment parameters and operating status of the distribution network; then, after the simulation is completed, the results of the simulated voltage and simulated current of the node equipment corresponding to the substation outlet are extracted from the model.

[0080] For example, assume that there is a simple distribution network. The actual outlet voltage of the substation recorded by the measuring equipment is 230 volts (V), and the actual outlet current is 200 amperes (A). The collected data is preprocessed to ensure the uniformity of data units, and then the actual outlet voltage of 230V and the actual outlet current of 200A are input into the distribution network simulation model. After receiving the input, the distribution network simulation model starts to calculate and output the simulated voltage and simulated current of the node. For example, the simulated voltage is 220V and the simulated current is 210V.

[0081] In step 103, the features of each node included in the graph structure are aggregated to obtain the aggregated features of each node.

[0082] In some embodiments, see Figure 4 , Figure 4 is a second flow chart of the processing method of the distribution network simulation model provided in the embodiment of the present application, such as Figure 4 As shown, Figure 3 Step 103 shown may be performed by Figure 4 Steps 1031 to 1032 shown in the figure are implemented by combining Figure 4 The steps shown are explained.

[0083] In step 1031, for a node corresponding to an electric power device in a graph structure, an aggregation feature of the node is determined based on features of neighboring nodes of the node.

[0084] In some embodiments, the above-mentioned step 1031 can be implemented in the following manner: calling the graph neural network model to perform the following processing: obtaining the first-order neighbor nodes and second-order neighbor nodes of the node; aggregating the features of the second-order neighbor nodes to obtain aggregated features; adding the aggregated features and the features of the first-order neighbor nodes to obtain updated features of the first-order neighbor nodes; adding the updated features of the first-order neighbor nodes and the features of the node to obtain the aggregated features of the node.

[0085] For example, see Figure 5 , Figure 5 is a diagram of a processing method for a distribution network simulation model provided in an embodiment of the present application, such as Figure 5 As shown in FIG. 2 , node 201 is taken as the target node, node 202 and node 203 are the first-order neighbor nodes of target node 201, and node 204, node 205, node 206 and node 207 are the second-order neighbor nodes of target node 201. First, the first-order neighbor nodes (i.e., node 202 and node 203) and the second-order neighbor nodes (i.e., node 204, node 205, node 206 and node 207) of target node 201 are obtained. Then, the features of the second-order neighbor nodes 204 and 205 are aggregated to obtain aggregated features 1, and The features of the second-order neighbor nodes 206 and 207 are aggregated to obtain aggregate feature 2; then, the aggregate feature 1 is added to the feature of the first-order neighbor node 202 to obtain the updated feature of the first-order neighbor node 202, and the aggregate feature 2 is added to the feature of the first-order neighbor node 203 to obtain the updated feature of the first-order neighbor node 203; finally, the updated feature of the first-order neighbor node 202, the updated feature of the first-order neighbor node 203 and the feature of the target node 201 are added to obtain the aggregate feature of the target node 201.

[0086] In some other embodiments, the above-mentioned step 1031 can also be implemented in the following way: calling the graph neural network model to perform the following processing: obtaining the first-order neighbor nodes and second-order neighbor nodes of the node; aggregating the features of the second-order neighbor nodes to obtain aggregated features; based on pre-trained connection weights, adding the aggregated features and the features of the first-order neighbor nodes to obtain updated features of the first-order neighbor nodes; based on pre-trained connection weights, adding the updated features of the first-order neighbor nodes and the features of the node to obtain the aggregated features of the node.

[0087] For example, based on a pre-trained aggregation function, the node features of the second-order neighbor nodes (circuit breakers, transformers) are aggregated to obtain aggregate feature 1, and based on the pre-trained connection weights, the aggregate feature 1 is added to the node features of the first-order neighbor nodes (such as series compensation capacitors) to obtain updated node features of the first-order neighbor nodes (such as series compensation capacitors); based on the pre-trained aggregation function, the node features of the second-order neighbor nodes (such as two power nodes) are aggregated to obtain aggregate feature 2, and based on the pre-trained connection weights, the aggregate feature 2 is added to the node features of the first-order neighbor nodes (such as transformers) to obtain updated node features of the first-order neighbor nodes (such as transformers); based on the pre-trained aggregation function, the updated node features of the first-order neighbor nodes (such as series compensation capacitors, transformers) are aggregated to obtain aggregate feature 3, and based on the pre-trained connection weights, the aggregate feature 3 is added to the node features of the target node (such as circuit breaker) to obtain the aggregate feature of the target node (such as circuit breaker).

[0088] In step 1032, for the nodes corresponding to the electric devices in the graph structure, the features of the nodes within a set time period are aggregated to obtain aggregated features of the nodes.

[0089] In some embodiments, the above step 1032 can be implemented in the following manner: obtaining a set duration; collecting feature data of the node within the set duration; and aggregating the feature data to obtain aggregate features of the node.

[0090] In some embodiments, a set duration is first defined, where the set duration can be hours, days, weeks, etc., depending on the specific application scenario and data collection frequency; then, feature data of the node within the set duration is collected, such as physical quantities such as voltage, current, power, etc. of the node at each time point, or status information of the node (such as switch status, fault indication, etc.); then, the collected feature data is preprocessed, including time synchronization, data cleaning, data standardization, etc.; then, the preprocessed feature data is aggregated (including simple aggregation, weighted aggregation, dimensionality reduction, etc.) to obtain aggregated features of the node.

[0091] For example, suppose there is a node P, and you want to analyze the characteristics of node P in the past 24 hours. First, collect the power consumption of node P in the past 24 hours; then, synchronize and clean the acquired data to ensure the quality of the acquired feature data; finally, calculate the average, maximum and minimum power consumption in the past 24 hours as the aggregated features of node P.

[0092] In step 104, based on the aggregated features of each node, the problematic devices in the distribution network simulation model are determined.

[0093] In some embodiments, see Figure 6 , Figure 6 is a third flow chart of the processing method of the distribution network simulation model provided in the embodiment of the present application, such as Figure 6 As shown, for each node included in the graph structure, Figure 3 Step 104 shown may be performed by Figure 6 Steps 1041 to 1042 shown in the figure are implemented by combining Figure 6 The steps shown are explained.

[0094] In step 1041, based on the aggregated features of the node, the trained classifier is called to perform prediction to obtain the probability that the node is a problem node.

[0095] In some embodiments, the aggregate features of the node are first preprocessed to ensure that the input features meet the input requirements of the classifier; then, the model parameters are loaded from the trained classifier model, that is, the corresponding model is imported from the model file or model library; then, the trained classifier model is used to predict the aggregate features of the node as the model input, and the probability of each node belonging to the problem node is output.

[0096] For example, suppose there is a node A whose aggregate feature is [0.5, 0.3, 0.2]. Load the trained classifier model from the model library, and input the aggregate feature of node A into the classifier for prediction. The output probability that node A is a problem node is 0.65, so that we can judge whether node A is a problem node based on the probability later.

[0097] In some embodiments, before calling the trained classifier for prediction, the following processing can also be performed: obtaining the aggregated features of the sample nodes, wherein the aggregated features of the sample nodes are obtained by calling the initialized graph neural network model to aggregate the features of the sample nodes in the sample graph structure; based on the aggregated features of the sample nodes, calling the initialized classifier for prediction to obtain the probability that the sample node belongs to the problem node; determining the difference between the probability output by the initialized classifier and the probability pre-annotated for the sample node; performing back propagation based on the difference, and updating the parameters of the classifier and the graph neural network model during the back propagation process.

[0098] For example, the graph neural network model and classifier are initialized, the sample graph structure is used as the input of the graph neural network model, the graph neural network model is called to perform feature aggregation on the sample nodes, and the aggregated features of each sample node are output; then, the aggregated features of the sample nodes are used as input, and the initialized classifier is called for prediction to obtain the predicted probability that the sample node belongs to the problem node; then, the difference between the output prediction probability of the initialized classifier and the pre-labeled probability is calculated, and back propagation is performed based on the difference. During the back propagation process, the gradients of the classifier and the graph neural network model are calculated, and the parameters of the graph neural network model and the classifier are updated through the gradients.

[0099] Here we explain back propagation. The training sample data is input into the input layer of the model, passes through the hidden layer, and finally reaches the output layer and outputs the result. This is the forward propagation process of the model. Since there is an error between the output result of the model and the actual result, the error between the result and the actual value is calculated, and the error is back propagated from the output layer to the hidden layer until it propagates to the input layer. During the back propagation process, the value of the model parameter is adjusted according to the error; the above process is continuously iterated until convergence.

[0100] Taking the difference value obtained by the classifier model in the embodiment of the present application as an example, the server starts backpropagation from the output layer of the classifier model based on the difference value, and backpropagates layer by layer. When the difference value reaches each layer, the gradient of the current layer is solved in combination with the conducted difference value, and the parameters in each layer are updated with the corresponding gradient values.

[0101] In some embodiments, before calling the trained classifier for prediction, the following processing can also be performed: based on the sample graph structure, calling the initialized graph neural network model for aggregation to obtain the aggregated features of the sample nodes; determining the difference between the aggregated features output by the initialized graph neural network model and the aggregated features pre-labeled for the sample nodes; performing back propagation based on the difference, and updating the parameters of the graph neural network model during the back propagation process; based on the aggregated features of the sample nodes, calling the initialized classifier for prediction to obtain the probability that the sample node belongs to the problem node; determining the difference between the probability output by the initialized classifier and the probability pre-labeled for the sample nodes; performing back propagation based on the difference, and updating the parameters of the classifier during the back propagation process.

[0102] In step 1042, when the probability is greater than the probability threshold, the device corresponding to the node in the distribution network simulation model is determined as a problematic device.

[0103] It should be noted that the probability threshold may be a preset value or may be dynamically set according to different application scenarios and requirements.

[0104] Continuing with the above example, assuming that the probability that node A is a problem node is 0.65 and the probability threshold is 0.6, the probability that node A is a problem node is greater than the probability threshold. Therefore, the device corresponding to node A in the distribution network simulation model is determined as a problem device.

[0105] In some embodiments, after executing step 1042, the following processing may also be performed on the problem node: when the original information of the problem device is erroneous information, the real information of the problem device is obtained, and the original information of the problem device is corrected based on the real information.

[0106] In some embodiments, after executing step 104, when the differences between the real circuit parameters and the simulated circuit parameters of each device in the distribution network simulation model are less than the difference threshold, the following processing can also be performed: according to the distribution network operation standard, a fault is applied to the distribution network simulation model, and the operation data of each device in the distribution network simulation model is determined; for each device, the protection setting of the device is determined based on the device's operation data and the relay protection configuration logic.

[0107] It should be noted that the simulated fault should be set according to the local power grid operation requirements and local operation characteristics, and the maximum value should be taken to ensure that the actual faults are included in the simulation calculation process (for example, the grounding resistance of the grounding fault should be calculated as 0Ω and the maximum resistance that may occur locally (assuming that there are many trees locally, it can be calculated according to the resistance of the tree trunks), and the short-circuit current should be calculated according to the short circuit that occurs directly behind the equipment); the setting confirmation should be set according to the relay protection configuration logic and the equipment characteristics (assuming that the calculated fault current is 500A, but the maximum current that the equipment can withstand is 400A, then the upper switch protection setting should be less than 400A to ensure that the equipment will not be damaged).

[0108] The following is an explanation of an exemplary application of the embodiments of the present application in a practical application scenario.

[0109] In the related technology, operators need to obtain relevant information of various devices such as switches, transformers, capacitors, etc. in the distribution network from the power supply bureau, including equipment model, equipment working principle, equipment parameters, etc. Based on this equipment information, operators add these devices to the distribution network line model and set the equipment properties to build a distribution network simulation model. However, when there is a large difference between the initially established distribution network simulation model and the actual distribution network situation, it is necessary to check the information of each device in the distribution network one by one, and the checking efficiency is low, which affects the simulation efficiency.

[0110] Based on this, an embodiment of the present application provides a method for processing a distribution network simulation model. When there is a large difference between the initially established distribution network simulation model and the actual distribution network situation, in order to help operators quickly locate the problem equipment, the problem equipment is predicted based on the GraphSAGE model; first, according to the topological relationship between each power equipment and power-consuming equipment in the initially established distribution network simulation model, a graph structure is established, and the characteristics of each node are defined according to the real voltage, real current, simulated voltage, and simulated current of each device, and then the GraphSAGE model is used to aggregate the characteristics of each node to obtain the aggregated characteristics of each node. Finally, according to the aggregated characteristics of each node, the problem node is predicted, that is, the power equipment that provides incorrect equipment information, thereby improving the verification efficiency of equipment in the distribution network.

[0111] The specific implementation scheme of the embodiment of the present application can be implemented by the following steps:

[0112] Phase 1: Establish a graph structure based on the topological relationship of each node in the distribution network simulation model.

[0113] For example, Figure 3As shown, a graph structure is established based on the topological relationship between each power equipment node (circuit breaker, transformer, capacitor) and each power consumption node (each household can be regarded as a power consumption node) in the distribution network. Each node in the graph structure represents a power equipment node or a power consumption node. If two nodes are directly connected by a power line, a connection line is set between the two nodes, that is, the connection line between the two nodes in the graph structure represents that there is a power line between the two nodes.

[0114] Phase 2: Assign node features to each node in the graph structure.

[0115] First, the real data of the distribution network at a certain moment is obtained, including the export voltage and export current of the substation at that moment, the voltage and current of each power equipment node at that moment, and the voltage and current of each power consumption node at that moment.

[0116] Then, the output voltage and output current of the substation at that moment are input into the initially established distribution network simulation model (i.e. Figure 3 The corresponding distribution network simulation model, the connection relationship between each node in the distribution network simulation model is Figure 3 The connection relationship is consistent with that shown in the figure), and by running the simulation software, the simulation voltage and simulation current at each node calculated by the simulation software are obtained.

[0117] Subsequently, for each node, the voltage difference percentage between the simulated voltage and the real voltage is calculated, and the current difference percentage between the simulated current and the real current is calculated. If the voltage difference percentage of each node is lower than the threshold (for example, 2%), and the current difference percentage of each node is also lower than the threshold, then the distribution network simulation model is considered to be successfully built; otherwise, the voltage difference percentage and the current difference percentage are concatenated to obtain the node characteristic vector of the node; among them, the real outlet voltage and the real outlet current of the substation are equal to the simulated outlet voltage and the simulated outlet current, so the voltage difference percentage and the current difference percentage of the substation are both equal to 0.

[0118] Phase 3: Use the pre-trained GraphSAGE model to aggregate the features of each node in the graph structure to obtain the aggregated features of each node.

[0119] First, based on the pre-set sampling strategy, first-order neighbor nodes and second-order neighbor nodes are randomly selected for each node, where the total number of first-order neighbor nodes is 2 and the total number of second-order neighbor nodes is 4. If the number of first-order neighbor nodes of a node is less than 2, or the number of second-order neighbor nodes is less than 4, you can choose to repeatedly sample these neighbor nodes to reach the required number.

[0120] Next, based on the pre-trained aggregation function, the node features of the second-order neighbor nodes (circuit breakers, transformers) are aggregated to obtain aggregate feature 1, and based on the pre-trained connection weights, the aggregate feature 1 is added to the node features of the first-order neighbor nodes (series compensation capacitors) to obtain updated node features of the first-order neighbor nodes (series compensation capacitors); based on the pre-trained aggregation function, the node features of the second-order neighbor nodes (two power nodes) are aggregated to obtain aggregate feature 2, and based on the pre-trained connection weights, the aggregate feature 2 is added to the node features of the first-order neighbor nodes (transformers) to obtain updated node features of the first-order neighbor nodes (transformers); based on the pre-trained aggregation function, the updated node features of the first-order neighbor nodes (series compensation capacitors, transformers) are aggregated to obtain aggregate feature 3, and based on the pre-trained connection weights, the aggregate feature 3 is added to the node features of the target node (circuit breaker) to obtain the aggregate features of the target node (circuit breaker).

[0121] Stage 4: The aggregate node features of each node are input into a pre-trained classifier, and the classifier is used to predict nodes that provide incorrect information.

[0122] Here, the classifier includes a fully connected layer and an activation function (such as softmax). The classifier outputs the probability that each node is a problem node, and regards nodes with probabilities higher than a threshold as nodes that provide incorrect information.

[0123] It should be noted that in this application, supervised learning or unsupervised learning strategies can be used to jointly train the GraphSAGE model and the classifier. When supervised learning is used, the loss value can be calculated based on the probability of each node output by the classifier and the true result, so as to update the parameters of the GraphSAGE model and the classifier using the loss value.

[0124] Phase 5: Troubleshoot problem nodes.

[0125] Here, if the information of the problem node is really provided incorrectly, the real information of the node is obtained, and the properties of the node in the distribution network simulation model are corrected using the real information.

[0126] Stage 6: Repeat stages 2-5 until the distribution network simulation model is successfully built.

[0127] Here, if the true information of a node is "the node has been cancelled", the node needs to be deleted from the graph structure before repeating steps 2-5.

[0128] In some embodiments, various types of faults are applied to the distribution network simulation model according to the local distribution network operation standard, the operation data of each device node is calculated, and the protection setting of the device is set according to the relay protection calculation logic.

[0129] Among them, the simulated fault should be set according to the local power grid operation requirements and local operation characteristics, and the maximum value should be taken to ensure that the actual faults are included in the simulation calculation process (such as the grounding resistance of the grounding fault should be calculated as 0Ω and the maximum resistance that may occur locally (assuming that there are many trees locally, it can be calculated according to the resistance of the tree trunks), and the short-circuit current should be calculated according to the short circuit that occurs directly behind the equipment); the set value confirmation should be set according to the relay protection configuration logic combined with the equipment characteristics (assuming that the calculated fault current is 500A, but the maximum current that the equipment can withstand is 400A, then the upper switch protection set value should be less than 400A to ensure that the equipment will not be damaged).

[0130] The following further describes an exemplary structure of the processing device 455 of the distribution network simulation model provided in the embodiment of the present application implemented as a software module. In some embodiments, for example Figure 1 As shown, the software modules stored in the processing device 455 of the distribution network simulation model in the memory 450 may include: a construction module 4551 , a determination module 4552 and an aggregation module 4553 .

[0131] A construction module 4551 is used to construct a graph structure based on the topological relationship between each device in the distribution network simulation model, wherein each node in the graph structure corresponds to a device, which is a power device or a power-consuming device; a determination module 4552 is used to determine the characteristics of the node based on the difference between the real circuit parameters and the simulated circuit parameters of the device corresponding to each node; an aggregation module 4553 is used to aggregate the characteristics of each node included in the graph structure to obtain the aggregated characteristics of each node; the determination module 4552 is also used to determine the problem device in the distribution network simulation model based on the aggregated characteristics of each node.

[0132] In some embodiments, real circuit parameters include real voltage and real current, and simulated circuit parameters include simulated voltage and simulated current. For each node included in the graph structure, the determination module 4552 is also used to obtain the real voltage and real current of the device corresponding to the node, and obtain the simulated voltage and simulated current of the device corresponding to the node; determine the voltage difference between the real voltage and the simulated voltage, and determine the current difference between the real current and the simulated current; splice the voltage difference and the current difference, and use the splicing result as the feature of the node.

[0133] In some embodiments, the determination module 4552 is also used to obtain real data of the distribution network, wherein the real data of the distribution network includes the real export voltage and the real export current of the substation; the real export voltage and the real export current of the substation are input into the distribution network simulation model, and simulation processing is performed to obtain the simulated voltage and simulated current of the equipment corresponding to the node in the distribution network simulation model.

[0134] In some embodiments, the aggregation module 4553 is also used to determine the aggregation characteristics of the nodes corresponding to the power equipment in the graph structure based on the characteristics of the node's neighbor nodes; for the nodes corresponding to the power equipment in the graph structure, the characteristics of the nodes within a set time period are aggregated to obtain the aggregation characteristics of the nodes.

[0135] In some embodiments, the aggregation module 4553 is also used to call the graph neural network model to perform the following processing: obtain the first-order neighbor nodes and second-order neighbor nodes of the node; aggregate the features of the second-order neighbor nodes to obtain aggregate features; add the aggregate features and the features of the first-order neighbor nodes to obtain updated features of the first-order neighbor nodes; add the updated features of the first-order neighbor nodes and the features of the node to obtain the aggregate features of the node.

[0136] In some embodiments, for each node included in the graph structure, the determination module 4552 is also used to call the trained classifier for prediction based on the aggregated features of the node to obtain the probability that the node is a problem node; when the probability is greater than the probability threshold, the device corresponding to the node in the distribution network simulation model is determined as a problem device.

[0137] In some embodiments, the processing device 455 of the distribution network simulation model mentioned above also includes a model training module 4554, which is used to obtain the aggregated features of the sample nodes, wherein the aggregated features of the sample nodes are obtained by calling the initialized graph neural network model to aggregate the features of the sample nodes in the sample graph structure; based on the aggregated features of the sample nodes, calling the initialized classifier for prediction to obtain the probability that the sample node belongs to the problem node; determining the difference between the probability output by the initialized classifier and the probability pre-marked for the sample node; performing back propagation based on the difference, and updating the parameters of the classifier and the graph neural network model during the back propagation process.

[0138] In some embodiments, the processing device 455 of the distribution network simulation model further includes a correction module 4555, which is used to obtain the real information of the problem device when the original information of the problem device is erroneous information, and to correct the original information of the problem device based on the real information.

[0139] In some embodiments, when the differences between the real circuit parameters and the simulated circuit parameters of each device in the distribution network simulation model are less than the difference threshold, the determination module 4552 is also used to apply a fault to the distribution network simulation model according to the distribution network operation standard, and determine the operation data of each device in the distribution network simulation model; for each device, the protection setting of the device is determined based on the device's operation data and the relay protection configuration logic.

[0140] It should be noted that the description of the device in the embodiment of the present application is similar to the description of the method embodiment above, and has similar beneficial effects as the method embodiment, so it will not be repeated. Figure 2 , Figure 4 ,or Figure 6 The present invention should be understood by referring to the description of any one of the accompanying drawings.

[0141] The embodiment of the present application provides a computer program product, which includes a computer program or a computer executable instruction, and the computer program or the computer executable instruction is stored in a computer-readable storage medium. The processor of the computer device reads the computer executable instruction from the computer-readable storage medium, and the processor executes the computer executable instruction, so that the computer device executes the processing method of the distribution network simulation model described in the embodiment of the present application.

[0142] The embodiment of the present application provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, the processor will execute the processing method of the distribution network simulation model provided by the embodiment of the present application, for example, Figure 2 , Figure 4 ,or Figure 6 The processing method of the distribution network simulation model is shown.

[0143] In some embodiments, the computer-readable storage medium may be a ferroelectric random access memory (FRAM), ROM, programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory, magnetic surface memory, optical disk, or compact disc read-only memory (CD-ROM) and other memories; it may also be various devices including one or any combination of the above memories.

[0144] In some embodiments, executable instructions may be in the form of a program, software, software module, script or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine or other unit suitable for use in a computing environment.

[0145] As an example, the executable instructions may be deployed to be executed on one electronic device, or on multiple electronic devices located at one site, or on multiple electronic devices distributed at multiple sites and interconnected by a communication network.

[0146] In summary, the embodiments of the present application can effectively improve the verification efficiency of devices in the distribution network simulation model, thereby improving the simulation efficiency.

[0147] The above is only an embodiment of the present application and is not intended to limit the protection scope of the present application. Any modifications, equivalent substitutions and improvements made within the spirit and scope of the present application are included in the protection scope of the present application.

Claims

1. A method for processing a distribution network simulation model, characterized in that: The method comprises: Based on the topological relationship between the devices in the distribution network simulation model, a graph structure is constructed, wherein each node in the graph structure corresponds to one device, and the device is a power device or a power-consuming device; Determine the characteristics of the node based on the difference between the real circuit parameters and the simulated circuit parameters of the device corresponding to each of the nodes; For the node corresponding to the electric power equipment in the graph structure, the graph neural network model is called to perform the following processing: obtaining the first-order neighbor node and the second-order neighbor node of the node; aggregating the features of the second-order neighbor node to obtain the aggregated features; adding the aggregated features to the features of the first-order neighbor node to obtain the updated features of the first-order neighbor node; adding the updated features of the first-order neighbor node to the features of the node to obtain the aggregated features of the node; For the node corresponding to the electric device in the graph structure, aggregating the features of the node within a set time period to obtain aggregated features of the node; Based on the aggregated features of each of the nodes, problematic devices in the distribution network simulation model are determined.

2. The method according to claim 1, characterized in that The real circuit parameters include real voltage and real current, and the simulated circuit parameters include simulated voltage and simulated current; The determining the characteristics of the node based on the difference between the real circuit parameters and the simulated circuit parameters of the device corresponding to each node includes: For each node included in the graph structure, the following processing is performed: Acquire the real voltage and real current of the device corresponding to the node, and acquire the simulated voltage and simulated current of the device corresponding to the node; determining a voltage difference between the real voltage and the simulated voltage, and determining a current difference between the real current and the simulated current; The voltage difference and the current difference are concatenated, and a concatenated result is used as a feature of the node.

3. The method according to claim 2, characterized in that The obtaining of the simulated voltage and the simulated current of the device corresponding to the node includes: Acquire real data of the distribution network, wherein the real data of the distribution network includes the real outlet voltage and the real outlet current of the substation; The real outlet voltage and the real outlet current of the substation are input into the distribution network simulation model, and simulation processing is performed to obtain the simulated voltage and the simulated current of the equipment corresponding to the node in the distribution network simulation model.

4. The method according to claim 1, characterized in that: The determining, based on the aggregated features of each of the nodes, problematic devices in the distribution network simulation model comprises: For each of the nodes included in the graph structure, the following processing is performed: Based on the aggregated features of the node, the trained classifier is called to perform prediction to obtain the probability that the node is a problem node; When the probability is greater than a probability threshold, a device corresponding to the node in the distribution network simulation model is determined as a problematic device.

5. The method according to claim 4, characterized in that Before calling the trained classifier for prediction, the method further includes: Obtaining aggregate features of sample nodes, wherein the aggregate features of the sample nodes are obtained by aggregating features of the sample nodes in the sample graph structure by calling an initialized graph neural network model; Based on the aggregated features of the sample node, calling the initialized classifier to perform prediction, and obtaining the probability that the sample node belongs to the problem node; Determine the difference between the probability output by the initialized classifier and the probability pre-labeled for the sample node; Back propagation is performed based on the difference, and parameters of the classifier and the graph neural network model are updated during the back propagation process.

6. The method according to claim 1, characterized in that The method further comprises: When the original information of the problematic device is erroneous information, the real information of the problematic device is obtained, and the original information of the problematic device is corrected based on the real information.

7. The method according to claim 1, characterized in that When the differences between the real circuit parameters and the simulated circuit parameters of each device in the distribution network simulation model are less than the difference threshold, the method further includes: According to the distribution network operation standard, applying a fault to the distribution network simulation model and determining the operation data of each device in the distribution network simulation model; For each of the devices, the protection setting of the device is determined based on the operating data of the device and the relay protection configuration logic.

8. A processing device for a distribution network simulation model, characterized in that: The device comprises: A construction module, used to construct a graph structure based on the topological relationship between various devices in the distribution network simulation model, wherein each node in the graph structure corresponds to one of the devices, and the device is a power device or a power-consuming device; A determination module, configured to determine the characteristics of each node based on a difference between a real circuit parameter and a simulated circuit parameter of a device corresponding to each node; an aggregation module, for calling a graph neural network model to perform the following processing on a node corresponding to the electric power equipment in the graph structure: obtaining a first-order neighbor node and a second-order neighbor node of the node; aggregating features of the second-order neighbor node to obtain an aggregate feature; adding the aggregate feature to the features of the first-order neighbor node to obtain an updated feature of the first-order neighbor node; adding the updated feature of the first-order neighbor node to the feature of the node to obtain an aggregate feature of the node; The aggregation module is further configured to aggregate the features of the nodes corresponding to the electrical devices in the graph structure within a set time period to obtain aggregate features of the nodes; The determination module is further used to determine problematic devices in the distribution network simulation model based on the aggregated features of each of the nodes.

9. An electronic device, characterized in that: include: A memory for storing computer executable instructions; A processor, used to implement the processing method of the distribution network simulation model according to any one of claims 1 to 7 when executing the computer executable instructions stored in the memory.

10. A computer-readable storage medium storing computer-executable instructions, characterized in that: When the computer executable instructions are executed by a processor, the processing method of the distribution network simulation model described in any one of claims 1 to 7 is implemented.

11. A computer program product comprising a computer program or computer executable instructions, characterized in that: When the computer program or computer executable instruction is executed by a processor, the processing method of the distribution network simulation model described in any one of claims 1 to 7 is implemented.

Images