Regional prediction method, device, equipment, computer-readable storage medium and computer program product

Abstract

The present application provides a regional prediction method, device, equipment, computer-readable storage medium and computer program product; the method includes: constructing a graph structure based on the hierarchical relationship between multiple power protection devices in the power distribution system and the electrical parameters of each power protection device; extracting features from the graph structure to obtain a first feature matrix; based on the first feature matrix, performing prediction processing on each power protection device to obtain the power outage probability of each power protection device; determining a power protection device with a power outage probability higher than a probability threshold as a target power protection device, and determining the control area corresponding to the target power protection device as a power outage area. Through the present application, the accuracy of regional prediction can be improved.

Description

Technical Field

[0001] The present application relates to the field of power control technology, and in particular to a regional prediction method, device, equipment, computer-readable storage medium and computer program product. Background Art

[0002] In the power distribution system, regional prediction refers to predicting the area where power outages may occur under specific conditions by analyzing the operating data, historical fault records, equipment status, and environmental factors of the power distribution system. Power outage area prediction is an important part of the risk management of the power distribution system, which helps to identify and assess potential failure risks. It can also replace or repair equipment that may fail in a timely manner based on the prediction results to reduce the failure rate. And by analyzing the prediction results, the weak links of the power distribution system can be discovered, providing data support for the upgrade and optimization of the power distribution system.

[0003] In related technologies, the probability of power outage is usually predicted based on data such as the current current value, current voltage value, working time and power outage records of the power equipment. Since the data used in the prediction only contains the most basic information and lacks data diversity and depth, it will lead to the problem of low prediction accuracy. Summary of the invention

[0004] The embodiments of the present application provide a region prediction method, apparatus, device, computer-readable storage medium, and computer program product, which can improve the accuracy of region prediction.

[0005] The technical solution of the embodiment of the present application is implemented as follows:

[0006] The present application provides a method for region prediction, the method comprising:

[0007] Constructing a graph structure based on a hierarchical relationship between a plurality of power protection devices in a power distribution system and electrical parameters of each of the power protection devices;

[0008] Performing feature extraction on the graph structure to obtain a first feature matrix;

[0009] Based on the first characteristic matrix, predicting each of the power protection devices is performed to obtain a power failure probability of each of the power protection devices;

[0010] A power protection device with a power outage probability higher than a probability threshold is determined as a target power protection device, and a control area corresponding to the target power protection device is determined as a power outage area.

[0011] The present application provides a region prediction device, the device comprising:

[0012] A construction module, configured to construct a graph structure based on a hierarchical relationship between a plurality of power protection devices in a power distribution system and electrical parameters of each of the power protection devices;

[0013] A feature extraction module, used to extract features from the graph structure to obtain a first feature matrix;

[0014] A prediction module, configured to perform prediction processing on each of the power protection devices based on the first characteristic matrix to obtain a power outage probability of each of the power protection devices;

[0015] The first determination module is used to determine a power protection device with a power outage probability higher than a probability threshold as a target power protection device, and determine a control area corresponding to the target power protection device as a power outage area.

[0016] An embodiment of the present application provides a computer device, including:

[0017] A memory for storing computer executable instructions;

[0018] The processor is used to implement the region prediction method provided in the embodiment of the present application when executing the computer executable instructions stored in the memory.

[0019] An embodiment of the present application provides a computer-readable storage medium storing computer-executable instructions for causing a processor to execute the region prediction method provided in the embodiment of the present application.

[0020] An embodiment of the present application provides a computer program product, including a computer program or computer executable instructions. When the computer program or computer executable instructions are executed by a processor, the region prediction method provided in the embodiment of the present application is implemented.

[0021] The embodiments of the present application have the following beneficial effects:

[0022] In the embodiment of the present application, first, a graph structure will be constructed based on the hierarchical relationship between the power protection devices and the electrical parameters of the power protection devices, that is, a graph structure will be constructed for the power protection devices; then, the graph structure is subjected to feature extraction to obtain a first feature matrix. The embodiment of the present application does not directly extract features from the real-time electrical parameters of the power protection devices. Since the graph structure can reflect the hierarchical relationship between the power protection devices, the embodiment of the present application can take into account the hierarchical relationship between the power protection devices when performing feature extraction. The first feature matrix obtained in this way can not only reflect the characteristics of the electrical parameters of the power protection device itself, but also reflect the influence of the neighboring power protection devices on the electrical parameters of a power protection device. That is, the first feature matrix obtained can more accurately reflect the actual working conditions of multiple power protection devices in the power distribution system. Based on this, when using the first feature matrix to predict each power protection device, a power outage probability with high accuracy can be obtained accordingly, and then based on the power outage probability, the power outage area can be accurately determined, thereby improving the accuracy of regional prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 1 is a schematic diagram of a network architecture of a regional prediction system 100 provided in an embodiment of the present application;

[0024] Figure 2 is a schematic diagram of the structure of the server 400 provided in an embodiment of the present application;

[0025] Figure 3 It is a first flow chart of the regional prediction method provided in an embodiment of the present application;

[0026] Figure 4 It is a schematic diagram of a process for constructing a graph structure provided in an embodiment of the present application;

[0027] Figure 5 It is a structural schematic diagram of a connection edge creation result provided in an embodiment of the present application;

[0028] Figure 6 It is a first flow chart of the feature extraction method provided in the embodiment of the present application;

[0029] Fig. 7A It is a schematic diagram of a first process of determining an adjacency matrix provided in an embodiment of the present application;

[0030] Figure 7B is a schematic diagram of a second process for determining an adjacency matrix provided in an embodiment of the present application;

[0031] Figure 8 It is a second flow chart of the feature extraction method provided in the embodiment of the present application;

[0032] Fig. 9 is a second flow chart of the region prediction method provided in an embodiment of the present application;

[0033] Fig.10 It is a network structure diagram of the graph structure provided in an embodiment of the present application.

[0034] It should be pointed out that the above-mentioned "first" and "second" are only used to distinguish different solutions, and do not represent the degree of superiority or inferiority of the solutions or the priority in the implementation process. DETAILED DESCRIPTION

[0035] 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.

[0036] 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.

[0037] In the following description, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" 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.

[0038] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program that has a predetermined function and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories), or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0039] Unless otherwise defined, all technical and scientific terms used in the embodiments of the present application have the same meanings as those commonly understood by those skilled in the art. The terms used in the embodiments of the present application are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.

[0040] The relevant data collection and processing in the embodiments of this application should be strictly in accordance with the requirements of relevant laws and regulations when applied in examples, obtain the informed consent or separate consent of the personal information subject, and carry out subsequent data use and processing within the scope of authorization of laws and regulations and the personal information subject.

[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) Circuit breaker is an important protective device in the power system. The main function of the circuit breaker is to automatically cut off the circuit when a circuit fault or abnormality occurs, so as to protect the circuit and electrical equipment from damage, prevent the fault from expanding, and ensure the safe and stable operation of the power system.

[0043] 2) Superior-subordinate relationship: Regarding the connection relationship of circuit breakers, the superior-subordinate relationship refers to the hierarchical position of the circuit breaker in the protection circuit and its corresponding protection and control functions.

[0044] 3) Graph structure, which is a set of nodes (also called vertices) and edges (also called connections or arcs). A node is the basic unit in a graph, usually representing an entity or object; an edge is a line segment connecting two vertices, representing the relationship between vertices.

[0045] 4) An undirected graph is a graph whose edges have no direction. The degree of a vertex refers to the number of edges connected to the vertex.

[0046] 5) Adjacency matrix is ​​a matrix used in graph theory to represent the adjacency relationship of a graph. It is a square matrix, in which each row element represents a vertex, each column element also represents a vertex, and the elements in the matrix indicate whether there is an edge or arc between the corresponding vertices. For undirected graphs, the adjacency matrix is ​​symmetric; for directed graphs, the adjacency matrix is ​​not necessarily symmetric.

[0047] 6) Degree matrix, which is a diagonal matrix used to represent the degree of each vertex in the graph. All non-diagonal elements of the degree matrix are 0, while the elements on the diagonal are non-negative integers. In an undirected graph, the degree of a vertex is the number of edges connected to the vertex; in a directed graph, the degree of a vertex is divided into out-degree (the number of edges starting from the vertex) and in-degree (the number of edges pointing to the vertex). The degree matrix can be represented as an out-degree matrix or an in-degree matrix.

[0048] In order to better understand the region prediction method provided in the embodiment of the present application, the region prediction method in the related art and its existing shortcomings are first described.

[0049] When predicting power outages, the relevant technologies simply collect data such as current, voltage, working hours, and power outage records of each power device, and analyze these data to predict the probability of power outages. However, this prediction method does not take into account the hierarchical relationship between pole-mounted circuit breakers, nor does it take into account the relationship between the real-time current and the set value of each pole-mounted circuit breaker, resulting in low accuracy in power outage prediction.

[0050] The embodiments of the present application provide a method, apparatus, computer device, computer-readable storage medium, and computer program product for regional prediction, which can improve the accuracy of regional prediction. The following describes an exemplary application of the computer device provided in the embodiments of the present application. The computer device provided in the embodiments of the present application can be implemented as various types of terminals such as laptop computers, tablet computers, desktop computers, set-top boxes, smart phones, smart speakers, smart watches, smart TVs, and vehicle-mounted terminals, and can also be implemented as servers. The following describes an exemplary application when the computer device is implemented as a server.

[0051] See also Figure 1 , Figure 1 This is a schematic diagram of the network architecture of the regional prediction system 100 provided in an embodiment of the present application. To support a target regional prediction application, the terminal 200 is connected to the server 400 via the network 300, and the server 400 is connected to the power distribution system 500 via the network 300. The network 300 can be a wide area network or a local area network, or a combination of the two.

[0052] The server 400 is used to obtain the hierarchical relationship between multiple power protection devices and the electrical parameters of each power protection device from the power distribution system 500, wherein the server 400 can periodically obtain the electrical parameters of each power protection device from the power distribution system 500; the server 400 is also used to construct a graph structure based on the hierarchical relationship between multiple power protection devices in the power distribution system and the electrical parameters of each of the power protection devices; and perform feature extraction on the graph structure to obtain a first feature matrix; based on the first feature matrix, perform prediction processing on each power protection device to obtain the power outage probability of each power protection device; determine the power protection device whose power outage probability is higher than the probability threshold as the target power protection device, and determine the control area corresponding to the target power protection device as the power outage area; based on this, the server 400 can also generate an alarm message based on the power outage area identifier, and send the alarm message to the terminal 200, so that the terminal 200 displays the alarm message to achieve the purpose of timely reminder.

[0053] In the embodiment of the present application, the server 400 will first construct a graph structure based on the hierarchical relationship between the power protection devices and the electrical parameters of the power protection devices, that is, it will construct a graph structure for the power protection devices; then perform feature extraction on the graph structure to obtain a first feature matrix. The embodiment of the present application does not directly perform feature extraction on the real-time electrical parameters of the power protection devices. Since the graph structure can reflect the hierarchical relationship between the power protection devices, the embodiment of the present application can take into account the hierarchical relationship between the power protection devices when performing feature extraction. The first feature matrix obtained in this way can not only reflect the characteristics of the electrical parameters of the power protection device itself, but also reflect the influence of the neighboring power protection devices on the electrical parameters of a power protection device. That is, the first feature matrix obtained can more accurately reflect the actual working conditions of multiple power protection devices in the power distribution system. Based on this, when using the first feature matrix to predict each power protection device, a power outage probability with high accuracy can be obtained accordingly, and then based on the power outage probability, the power outage area can be accurately determined, thereby improving the accuracy of regional prediction.

[0054] In some embodiments, the server 400 may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms. The terminal and the server may be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiments of the present application.

[0055] See also Figure 2 , Figure 2 is a schematic diagram of the structure of the server 400 provided in an embodiment of the present application, Figure 2 The server 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 server 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 2 Various buses are labeled as bus system 440 .

[0056] 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.

[0057] 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.

[0058] 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.

[0059] 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.

[0060] 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.

[0061] 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;

[0062] 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 LAN (WiFi), and Universal Serial Bus (USB), etc.;

[0063] 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);

[0064] 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.

[0065] In some embodiments, the device provided in the embodiments of the present application can be implemented in software. Figure 2 The region prediction device 455 stored in the memory 450 is shown, which can be software in the form of a program and a plug-in, etc., and includes the following software modules: a construction module 4551, a feature extraction module 4552, a prediction module 4553 and a first determination module 4554. These modules are logical, and therefore can be arbitrarily combined or further split according to the functions implemented. The functions of each module will be described below.

[0066] In other embodiments, the device 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 regional prediction method 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), DSP, programmable logic device (Programmable Logic Device, PLD), complex programmable logic device (Complex Programmable Logic Device, CPLD), field programmable gate array (Field-Programmable Gate Array, FPGA) or other electronic components.

[0067] In some embodiments, the server can implement the regional prediction method provided in the embodiments of the present application by running various computer executable instructions or computer programs. For example, computer executable instructions can be microprogram-level commands, machine instructions or software instructions. The computer program can be a native program or software module in the operating system; it can be a native application (APPlication, APP), that is, a program that needs to be installed in the operating system to run, such as a regional prediction APP, anomaly detection APP; it can also be a small program that can be embedded in any APP, that is, a program that can be run only by downloading it to a browser environment. In short, the above-mentioned computer executable instructions can be instructions in any form, and the above-mentioned computer program can be an application, module or plug-in in any form.

[0068] The regional prediction method provided in the embodiment of the present application will be explained in combination with the exemplary application and implementation of the server provided in the embodiment of the present application.

[0069] The following describes the region prediction method provided by the embodiment of the present application. As mentioned above, the computer device implementing the region prediction method of the embodiment of the present application can be a terminal, a server, or a combination of the two. Therefore, the execution subject of each step will not be repeatedly described below.

[0070] It should be noted that in the example of the regional prediction method below, the power protection device is described as a circuit breaker. Based on the understanding of the following text, those skilled in the art can apply the regional prediction method provided in the embodiment of the present application to regional prediction of other power protection devices such as fuses and circuit breakers.

[0071] See also Figure 3 , Figure 3 is a first flow chart of the regional prediction method provided in the embodiment of the present application, which will be combined with Figure 3 The steps shown are used to illustrate the region prediction method provided in the embodiment of the present application. Figure 3 The execution subject is the server.

[0072] In step S101 , a graph structure is constructed based on the hierarchical relationship between multiple power protection devices in the power distribution system and the electrical parameters of each power protection device.

[0073] In an embodiment of the present application, a plurality of power protection devices are included in the power distribution system, wherein the power protection device can quickly detect various faults in the power system, such as short circuit, overload, grounding, etc., and quickly cut off the fault circuit according to the set logic and conditions to isolate the fault point and prevent the fault from expanding. The power protection device can also protect various equipment in the power system, including generators, transformers, transmission lines, busbars, motors, etc., to avoid equipment damage caused by faults. The power protection device can help reduce the time of system interruption by isolating the fault in time, thereby improving the reliability of power supply. The power protection device can quickly respond to the fault and isolate it, which can reduce the power loss and equipment damage caused by the fault, and reduce the operating cost of the power system. When the power protection device detects an abnormal situation that may cause system instability, it can take timely measures to maintain the stable operation of the power system.

[0074] The hierarchical relationship between the power protection devices may include a superior-subordinate relationship between the power protection devices, and may also include at least one of a master-slave relationship, a regional protection relationship, a functional relationship, and a response time relationship between the power protection devices. Taking the power protection device as a pole-mounted circuit breaker as an example, the hierarchical relationship between multiple power protection devices may be a superior-subordinate relationship between multiple pole-mounted circuit breakers.

[0075] In some embodiments, the electrical parameters of the power protection device may include at least one of a current setting value, a real-time current, a current trend indicator, a voltage setting value, a real-time voltage, and a voltage trend indicator.

[0076] In some embodiments, see Figure 4 The above step S101 can be implemented by the following steps S1011 to S1014, which are described in detail below.

[0077] In step S1011 , each power protection device is determined as a node.

[0078] In some embodiments, it is assumed that the power distribution system includes six power protection devices, namely, a first power protection device P1, a second power protection device P2, a third power protection device P3, a fourth power protection device P4, a fifth power protection device P5 and a sixth power protection device P6. Based on this, six nodes can be determined accordingly, namely, a first node N1, a second node N2, a third node N3, a fourth node N4, a fifth node N5 and a sixth node N6.

[0079] In step S1012, the electrical parameters of each power protection device are determined to correspond to the node information of the node corresponding to each power protection device.

[0080] In some embodiments, following the above example, the first power protection device P1 corresponds to the first node N1, the second power protection device P2 corresponds to the second node N2, the third power protection device P3 corresponds to the third node N3, the fourth power protection device P4 corresponds to the fourth node N4, the fifth power protection device P5 corresponds to the fifth node N5, and the sixth power protection device P6 corresponds to the sixth node N6. Based on this, assuming that the electrical parameters of the first power protection device P1 include current setting value, real-time current, current trend mark, voltage setting value, real-time voltage, voltage trend mark, and the current setting value, real-time current, current trend mark, voltage setting value, real-time voltage, and voltage trend mark of the first power protection device P1 are 100, 90, 0, 12, 8, and 1 respectively, then 100, 90, 0, 12, 8, and 1 are determined as the node information of the first node N1.

[0081] In step S1013, based on the hierarchical relationship between the plurality of power protection devices, a connection edge is created between corresponding nodes of two power protection devices having a superior-subordinate relationship.

[0082] In the embodiment of the present application, if the hierarchical relationship between two power protection devices indicates that there is a superior-subordinate relationship between the two power protection devices, a connection edge is created between the corresponding nodes of the two power protection devices. Correspondingly, if the hierarchical relationship between the two power protection devices indicates that there is no superior-subordinate relationship between the two power protection devices, a connection edge is not created between the corresponding nodes of the two power protection devices.

[0083] Continuing with the above example, assume that the hierarchical relationship is characterized as follows: there is a superior-subordinate relationship between the first power protection device P1 and the second power protection device P2, there is a superior-subordinate relationship between the first power protection device P1 and the third power protection device P3, there is a superior-subordinate relationship between the second power protection device P2 and the fourth power protection device P4, there is a superior-subordinate relationship between the second power protection device P2 and the fifth power protection device P5, and there is a superior-subordinate relationship between the third power protection device P3 and the sixth power protection device P6. Then create a connection edge between the first node N1 and the second node N2, a connection edge between the first node N1 and the third node N3, a connection edge between the second node N2 and the fourth node N4, a connection edge between the second node N2 and the fifth node N5, and a connection edge between the third node N3 and the sixth node N6, and obtain the following Figure 5 The structural diagram of the connection edge creation result is shown in Figure 2.

[0084] In step S1014, a graph structure is generated based on the nodes, node information and connection edges.

[0085] In some embodiments, the graph structure may be generated through the following four processes:

[0086] Process 1: Define data structure. Here, create classes or structures of nodes and edges to store node information of nodes and edge connection information between nodes. Process 2: Initialize graph. Here, create an empty graph container, which can usually be a dictionary or a specific graph data structure. Process 3: Add nodes. Here, based on the results determined in step S1011, create instances for each node and add them to the graph. Process 4: Add edges and connections. Here, based on the information of the connecting edges, add the connecting edges to the graph and update the adjacency list or matrix of the relevant nodes.

[0087] Through the above steps S1011 to S1014, a graph structure corresponding to the power protection device can be constructed, and the graph structure can reflect the hierarchical relationship between multiple power protection devices in the power distribution system and the electrical parameters of each power protection device.

[0088] Continue to see Figure 3 , and the description continues with step S101 above.

[0089] In step S102, feature extraction is performed on the graph structure to obtain a first feature matrix.

[0090] In the embodiment of the present application, a feature extraction method may be selected first; then the selected feature extraction method may be used to extract features from each node to obtain node features corresponding to each node; finally, the node features corresponding to each node may be combined to obtain a first feature matrix. The feature extraction method may include using deep learning methods such as graph convolutional networks (GCNs), centrality indicators such as degree, proximity, and betweenness based on graph theory, etc.; in the first feature matrix, each row represents a node feature corresponding to a node.

[0091] In some embodiments, see Figure 6 The above step S102 can be implemented by the following steps S1021 to S1023, which are described in detail below.

[0092] In step S1021, the adjacency matrix and degree matrix of the graph structure are determined.

[0093] In some embodiments, the adjacency matrix is ​​a binary matrix used to represent the connection relationship between nodes in the graph structure. For a graph with M nodes, the adjacency matrix is ​​an M*M matrix.

[0094] The degree matrix is ​​a diagonal matrix that represents the degree of each node in the graph structure. For a graph with M nodes, the degree matrix is ​​an M*M matrix. The matrix elements in the matrix represent the degree of the node, that is, the number of connected edges connected to the node. For an undirected graph, the degree matrix is ​​a matrix whose matrix elements on the diagonal are the node degrees; for a directed graph, the degree matrix can be an out-degree matrix or an in-degree matrix, representing the out-degree and in-degree of the node respectively.

[0095] In some embodiments, see Fig. 7A , the adjacency matrix of the graph structure can be determined by following steps S211A to S213A, which are described in detail below.

[0096] In step S211A, when there is a connecting edge between the i-th node and the j-th node, or when i is equal to j, the matrix element in the i-th row and j-th column is set to a first preset value.

[0097] In the embodiment of the present application, i and j are natural numbers less than or equal to M, where M is the total number of nodes.

[0098] In some embodiments, the first preset value is a value preset in advance. For example, the first preset value may be a value of 1. In this case, the value 1 is used to represent the existence of a connecting edge between the i-th node and the j-th node.

[0099] In some other embodiments, when i is equal to j, the matrix elements in the i-th row and j-th column are also set to the first preset value, that is, the matrix elements on the diagonal are also the first preset value.

[0100] In step S212A, when there is no connecting edge between the i-th node and the j-th node, and i is not equal to j, the matrix element in the i-th row and j-th column is set to a second preset value.

[0101] In the embodiment of the present application, the second preset value is different from the first preset value.

[0102] In some embodiments, the second preset value is also a pre-set value, but the second preset value is different from the first preset value. For example, the first preset value can be the value 0. In this case, the value 0 is used to represent that there is no connecting edge between the i-th node and the j-th node and i is not equal to j.

[0103] In other embodiments, since no self-loop is created when creating the graph structure, that is, no loop is created from one node to itself, two nodes without a connecting edge may refer to different nodes or the same node. Considering that in the above step S211A, when i is equal to j, the matrix element of the i-th row and j-th column is set to the first preset value, then in step S212A, when there is no connecting edge between the two nodes and the two nodes are different nodes, the matrix elements corresponding to the two nodes are set to the second preset value.

[0104] In step S213A, an adjacency matrix of the graph structure is constructed based on each matrix element.

[0105] In the embodiment of the present application, each matrix element is combined according to the row information and column information of each matrix element, so as to obtain an adjacency matrix of a graph structure.

[0106] In some embodiments, an initial adjacency matrix with empty initial elements may be obtained, and then the matrix elements determined in the above steps S211A and S212A may be used to update the initial elements in the initial adjacency matrix accordingly, thereby obtaining an adjacency matrix of a graph structure.

[0107] Through the above steps S211A to S213A, an adjacency matrix of a graph structure can be quickly constructed, thereby improving the efficiency of regional prediction.

[0108] In some embodiments, see Figure 7B , the adjacency matrix of the graph structure can be determined by following steps S211B to S214B, which are described in detail below.

[0109] In step S211B, when there is a connection edge between the i-th node and the j-th node, system configuration information of the power distribution system is acquired, and the matrix element of the i-th row and the j-th column is determined based on the configuration information.

[0110] In the embodiment of the present application, i and j are natural numbers less than or equal to M, where M is the total number of nodes.

[0111] In some embodiments, the configuration information of the power distribution system may include network topology, system parameters, device parameters, protection configuration, etc. The network topology includes the numbering and identification of nodes and lines, and the redundancy and backup configuration of the network.

[0112] In some embodiments, when there is a connection edge between two nodes, the corresponding matrix element may also be determined by configuration information of the power distribution system.

[0113] In some embodiments, the power protection device is a pole-mounted circuit breaker, and the "determining the matrix element of the i-th row and the j-th column based on the configuration information" in the above step S211B can be implemented by the following steps S2111 and S2112, which are described in detail below.

[0114] In step S2111, when it is determined based on the configuration information that a transformer is provided between the ith pole breaker corresponding to the ith node and the jth pole breaker corresponding to the jth node, the matrix element in the ith row and jth column is set to a third value.

[0115] In an embodiment of the present application, the third value is determined during the training process of the feature extraction model, and the third value is between the first preset value and the second preset value. That is, the third value is similar to the model parameter and is determined during the training process of the feature extraction model. Thus, before the feature extraction model is trained, the third value can be represented by an unknown parameter, such as x, and it is agreed in advance that the third value is between the first preset value and the second preset value. Assuming that the first preset value is a value of 1 and the second preset value is a value of 0, the third value obtained through training is between 1 and 1.

[0116] In some embodiments, since the transformer has a certain isolation effect in the line, it can play a certain isolation role in isolating fault propagation. This isolation effect will weaken the superior-subordinate relationship between the pole-mounted circuit breakers to a certain extent. Based on this, if the configuration information indicates that a transformer is set between two nodes where a connection edge exists, the corresponding matrix element is set to a third value.

[0117] In step S2112, when it is determined based on the configuration information that no transformer is provided between the ith pole-mounted circuit breaker and the jth pole-mounted circuit breaker, the matrix element in the ith row and the jth column is set to a first preset value.

[0118] In some embodiments, if the configuration information indicates that there is no transformer between two nodes with a connection edge, similar to the processing process of step S211A above, the corresponding matrix element is set to a first preset value. The first preset value is a value preset in advance, and illustratively, the first preset value can be a value of 1.

[0119] Through step S2111 and step S2112, when there is a connecting edge between two nodes, the corresponding matrix elements are not set directly, but whether a transformer is set between the corresponding pole-mounted circuit breakers is considered, and the corresponding matrix elements are set according to different settings. The matrix elements obtained in this way can accurately reflect the actual configuration of the distribution system and improve the matching degree between the matrix elements and the actual working conditions of the pole-mounted circuit breakers.

[0120] Continue to see Figure 7B, and the description continues with step S211B above.

[0121] In step S212B, when there is no connecting edge between the i-th node and the j-th node, and i is not equal to j, the matrix element in the i-th row and j-th column is set to a second preset value.

[0122] In some embodiments, the implementation process of step S212B is similar to the implementation process of step S212A, and therefore, the implementation process of step S212B can refer to the implementation process of step S212A. The second preset value is also a value preset in advance, but the second preset value is different from the first preset value. For example, the first preset value can be 0.

[0123] In step S213B, when i is equal to j, the matrix element in the i-th row and j-th column is set to a first preset value.

[0124] In the embodiment of the present application, the first preset value is different from the second preset value.

[0125] In some embodiments, when i is equal to j, the matrix elements in the i-th row and j-th column are also set to the first preset value, that is, the matrix elements on the diagonal are also the first preset value.

[0126] In step S214B, an adjacency matrix of the graph structure is constructed based on each matrix element.

[0127] In the embodiment of the present application, the implementation process of step S214B is similar to the implementation process of the above-mentioned step S213A. Therefore, the implementation process of step S214B can refer to the implementation process of the above-mentioned step S213A.

[0128] Through the above steps S211B to S214B, in the process of constructing the adjacency matrix, it is considered whether there is a transformer between two upper and lower pole-mounted circuit breakers, that is, the actual configuration of the distribution network is fully considered, so that a highly accurate adjacency matrix can be obtained.

[0129] Continue to see Figure 6 , and the description continues with step S1021 above.

[0130] In step S1022, an initial feature matrix of the graph structure is constructed based on the node information.

[0131] In the embodiment of the present application, the initial feature matrix of the graph structure can be considered as the matrix before feature extraction.

[0132] In some embodiments, the node information includes current setting value, real-time current, current trend mark, voltage setting value, real-time voltage and voltage trend mark, wherein the current setting value can be set in advance according to the rated current of the production equipment and the expected maximum load current, and the current setting value ensures that the circuit breaker can cut off the circuit in time when the current rises abnormally; in implementation, when the real-time current exceeds the current setting value, the power protection device automatically cuts off the circuit, thereby protecting the production equipment, and the power protection device can be a pole-mounted circuit breaker. The voltage setting value can be set according to the rated voltage of the production equipment and the expected operating voltage range. By setting the voltage setting value, the production equipment can be protected from damage by excessively high or low voltage; when the real-time voltage exceeds or is lower than the voltage setting value, the power protection device automatically cuts off the circuit to prevent damage to the production equipment.

[0133] Based on this, the above step S1022 can be implemented through the following steps S221 and S222, which are described in detail below.

[0134] In step S221, the current setting value, real-time current, current trend mark, voltage setting value, real-time voltage and voltage trend mark of the i-th node are determined as the i-th row vector.

[0135] In the embodiment of the present application, i is a natural number less than or equal to M, and M is the total number of nodes.

[0136] Exemplarily, assuming that the current setting value, real-time current, current trend identifier, voltage setting value, real-time voltage and voltage trend identifier of the first node are: 100, 90, 0, 12, 8, 1, then (100 90 0 12 8 1) is determined as the first row vector, and so on, the corresponding row vector of each node can be determined.

[0137] In step S222, an initial feature matrix of the graph structure is constructed based on the first row vector to the Mth row vector.

[0138] Continuing with the above example, assume that the first row vector is (100 90 0 12 8 1), the second row vector is (100 901 10 9 1), the third row vector is (100 90 1 10 7 0), the fourth row vector is (100 90 1 5 5 0), the fifth row vector is (100 90 1 5 4 0), and the sixth row vector is (100 65 0 5 3 1). Then the first row vector is determined as the matrix element of the first row in the initial feature matrix. Similarly, the second to sixth row vectors are determined as the matrix elements of the second to sixth rows in the initial feature matrix, thereby obtaining the initial feature matrix of the graph structure. .

[0139] Through the above-mentioned steps S221 and S222, an initial characteristic matrix can be constructed based on the current setting value, real-time current, current trend identification, voltage setting value, real-time voltage and voltage trend identification. During the construction process, not only the current trend and voltage trend are considered, but also the current setting value, real-time current, voltage setting value and real-time voltage are fully considered, thereby improving the integrity and diversity of the initial characteristic matrix.

[0140] Continue to see Figure 6 , and the description continues with step S1022 above.

[0141] In step S1023, a first feature matrix is ​​determined based on the adjacency matrix, the degree matrix and the initial feature matrix using the feature extraction model.

[0142] In an embodiment of the present application, the feature extraction model may be a model corresponding to a deep learning method, and exemplarily, the feature extraction model may be a graph convolutional network. Based on this, the process of determining the first feature matrix includes:

[0143] Process 1, normalized adjacency matrix:

[0144] The adjacency matrix is ​​normalized by the degree matrix to obtain the normalized adjacency matrix B. The commonly used normalization method is shown in formula (1):

[0145] Formula (1);

[0146] In formula (1), represents the degree matrix, represents the inverse square root matrix of the degree matrix, Represents the adjacency matrix.

[0147] Process 2: Define graph convolution operation:

[0148] The graph convolution operation can be implemented in many ways, one of which is shown in formula (2):

[0149] Formula (2);

[0150] In formula (2), represents the output feature matrix, represents the initial feature matrix (the input feature matrix), Represents the weight matrix of the graph convolutional layer, which is also the parameter matrix of the feature extraction model.

[0151] Process three, apply activation function:

[0152] After the graph convolution operation, a nonlinear activation function (such as Sigmoid, Tanh, ReLU) is used to increase the nonlinear ability of the model. In this way, the above process 1 to process 3 are iterated to obtain the first feature matrix.

[0153] In some embodiments, the feature extraction model includes N feature extraction layers, where N is an integer greater than or equal to 2. Figure 8 The above step S1023 can be implemented by the following steps S231 to S233, which are described in detail below.

[0154] When the feature extraction model can be a graph convolutional network, the feature extraction layer corresponds to the graph convolutional layer. In the graph convolutional network, the graph convolutional layer is responsible for processing the structural information of the graph. The graph convolutional layer combines the adjacency relationship and node features of the nodes, and extracts and transforms features through graph convolution operations. Each graph convolutional layer can be regarded as a feature extraction layer. The graph convolutional layer performs the following main functions:

[0155] Function 1: Aggregate neighbor information: The graph convolution layer combines the features of each node with the features of its neighbor nodes through a weight matrix to capture the relationship between nodes.

[0156] Function 2: Feature transformation: By applying a weight matrix, the graph convolution layer transforms the aggregated features to learn new feature representations of nodes in the graph.

[0157] Function three, activation function: A nonlinear activation function is applied after the graph convolution layer to increase the expressive power of the model.

[0158] Function 4: Output the first feature matrix: The output of the graph convolution layer is a new feature matrix, which contains the high-level feature representation of the nodes. The first feature matrix can be used for downstream tasks such as classification and prediction.

[0159] Therefore, in the graph convolutional network, each graph convolution layer can be regarded as a feature extraction layer, which can gradually extract more and more abstract first feature matrices from the initial feature matrix.

[0160] In step S231, based on the adjacency matrix, the degree matrix and the parameter matrix of the feature extraction model, the initial feature matrix is ​​linearly processed to obtain a second feature matrix; and the second feature matrix is ​​nonlinearly processed to obtain a first output matrix of the first feature extraction layer.

[0161] In some embodiments, during the linear processing, the adjacency matrix may be linearly processed based on the degree matrix, referring to the above formula (1) to obtain a normalized adjacency matrix; then, the initial feature matrix may be linearly processed based on the normalized adjacency matrix and the parameter matrix, referring to the above formula (2) to obtain a second feature matrix; then, the second feature matrix may be nonlinearly processed by a nonlinear activation function to obtain a first output matrix.

[0162] In step S232, based on the adjacency matrix, the degree matrix and the parameter matrix, the kth output matrix is ​​linearly processed to obtain a third feature matrix; and the third feature matrix is ​​nonlinearly processed to obtain a k+1th output matrix of the k+1th feature extraction layer.

[0163] In the embodiment of the present application, k is a natural number greater than 0 and less than N-1.

[0164] In some embodiments, after obtaining the first output matrix, the first output matrix is ​​used as the input matrix of the second feature extraction layer, and the above step S231 is continued to be performed to obtain the second output matrix of the second feature extraction layer. Similarly, the k+1th output matrix of the k+1th feature extraction layer is obtained.

[0165] In step S233, the Nth output matrix is ​​determined as the first characteristic matrix.

[0166] In the embodiment of the present application, the output matrix of the last feature extraction layer is determined as the first feature matrix, that is, the Nth output matrix is ​​determined as the first feature matrix.

[0167] Through step S231 to step S233, a first feature matrix with high abstractness can be extracted from the initial feature matrix through feature extraction, so as to be used for the execution of downstream tasks.

[0168] Through the above steps S1021 to S1023, the adjacency matrix, degree matrix and initial feature matrix of the graph structure are first determined, and then the feature extraction model is used to perform feature extraction, so as to determine a high-level and highly generalized first feature matrix.

[0169] Continue to see Figure 3 , and the description continues with step S102 above.

[0170] In step S103, prediction processing is performed on each power protection device based on the first characteristic matrix to obtain the power failure probability of each power protection device.

[0171] In the embodiment of the present application, the first characteristic matrix may be input into a classifier to implement prediction processing for each power protection device, and the classifier may output the power-off probability of each power protection device.

[0172] In some embodiments, the classifier can be at least one of logistic regression, support vector machine, random forest, gradient boosting decision tree, neural network, recurrent neural network, ensemble method, K nearest neighbor, and multilayer perceptron.

[0173] In step S104, a power protection device with a power outage probability higher than a probability threshold is determined as a target power protection device, and a control area corresponding to the target power protection device is determined as a power outage area.

[0174] In the embodiment of the present application, the probability threshold is a value set in advance. For example, the probability threshold may be 0.5, 0.55, 0.6, etc.

[0175] In some embodiments, the power outage probability of each power protection device is compared with the probability threshold to obtain a target power outage probability whose power outage probability is higher than the probability threshold, and then the power protection device corresponding to the target power outage probability is determined as the target power protection device, and finally the control area corresponding to the target power protection device is determined as the power outage area.

[0176] In some embodiments, if the power failure probability of each power protection device is less than or equal to the probability threshold, it indicates that there is no power failure area. If there are multiple target power failure probabilities higher than the probability threshold, it indicates that there are multiple power failure areas.

[0177] Through the above steps S101 to S104, a graph structure will be constructed based on the hierarchical relationship between the power protection devices and the electrical parameters of the power protection devices, that is, a graph structure will be constructed for the power protection devices, and the graph structure can reflect the hierarchical relationship between multiple power protection devices in the power distribution system and the electrical parameters of each power protection device; then feature extraction is performed based on the adjacency matrix, degree matrix and initial feature matrix of the graph structure to obtain the first feature matrix. The embodiment of the present application does not directly extract features from the real-time electrical parameters of the power protection device. On the one hand, since the graph structure can reflect the hierarchical relationship between the power protection devices, the embodiment of the present application can take into account the hierarchical relationship between the power protection devices when performing feature extraction. The first feature matrix obtained in this way can not only reflect the characteristics of the electrical parameters of the power protection device itself, but also reflect the influence of the neighboring power protection device on the electrical parameters of a power protection device; on the other hand. In the process of constructing the initial feature matrix, not only the current trend and voltage trend are considered, but also the current setting value, real-time current, voltage setting value and real-time voltage are fully considered, thereby improving the integrity and diversity of the initial feature matrix. That is, the obtained first characteristic matrix can more accurately reflect the actual working conditions of multiple power protection devices in the power distribution system. Based on this, when using the first characteristic matrix to predict each power protection device, a high-accuracy power outage probability can be obtained accordingly, and then based on the power outage probability, the power outage area can be accurately determined, thereby improving the accuracy of regional prediction.

[0178] In some embodiments, Fig. 9 is a second flow chart of the regional prediction method provided in the embodiment of the present application, see Fig. 9 After the above step S104, the following steps S105 and S106 may also be performed, which are described in detail below.

[0179] In step S105, an alarm message is generated based on the area identifier of the power outage area; and a target object is determined based on the area identifier.

[0180] In some embodiments, each area controlled by the power protection device has an area identifier. For example, the area controlled by the first power protection device is the first area, and the area identifier of the first area is R1; the area controlled by the second power protection device is the second area, and the identifier of the second area is R2. Assuming that the power outage area is the second area, an alarm message carrying R2 can be generated. The type of the alarm message can be a character type or an audio and video type. When the type of the alarm message is a character type, the alarm message can be "R2 is about to have a power outage, please handle it."

[0181] In some embodiments, the target object corresponding to the area identifier can also be obtained from a preset relationship table, wherein the preset relationship table stores the area identifier of each area and the corresponding object. Continuing with the above example, it can be determined from the preset relationship table that the object corresponding to R2 is the maintenance master M1 or the maintenance team T1, and then the maintenance master M1 or the maintenance team T1 is determined as the target object. That is, in the embodiment of the present application, the target object can be an individual or a group.

[0182] In step S106, an alarm message is sent to the terminal of the target object.

[0183] In some embodiments, the preset relationship table also stores the contact information of the target object, which may be an instant messaging number. Then, an alarm message may be sent to the target object's terminal based on the contact information. For example, a message may be sent to the target object's terminal, an audio may be sent to the target object's terminal, or a video may be sent to the target object's terminal.

[0184] Through the above steps S105 and S106, an alarm message can be automatically generated and sent, thereby improving the degree of automation and intelligence and achieving the purpose of timely reminder, thereby minimizing the losses caused by power outages.

[0185] The following is an explanation of an exemplary application of the embodiments of the present application in a practical application scenario.

[0186] The pole-mounted circuit breaker is a key device in the distribution network, which is used to control and protect the safety of the distribution system. The pole-mounted circuit breaker can automatically disconnect the circuit when a circuit fault occurs, preventing the fault from spreading and equipment damage, thereby improving the reliability and safety of power supply. The main functions of the pole-mounted circuit breaker include overload protection, short-circuit protection and isolation disconnection. The pole-mounted circuit breaker can also respond quickly to fault conditions, disconnect the power supply, help locate the fault point, and facilitate maintenance and restoration of power supply. Among them, the pole-mounted circuit breaker corresponds to the power protection device in other embodiments.

[0187] In the distribution network, pole-mounted circuit breakers are directly related to power outage events. When abnormal current or voltage fluctuations are detected, the pole-mounted circuit breaker automatically disconnects to prevent further equipment damage or instability in the power system. This automatic power-off function is an important safety mechanism in system design, ensuring that the power supply system can react quickly when faced with electrical faults, minimizing the impact on users and the power grid. Therefore, pole-mounted circuit breakers play a vital role in preventing and managing power outages and are key equipment to ensure the continuity and stability of power supply.

[0188] According to the position of the pole-mounted circuit breaker in the power grid, the pole-mounted circuit breaker can be divided into an upper circuit breaker and a lower circuit breaker. One upper circuit breaker can usually correspond to multiple lower circuit breakers. The setting of the upper circuit breaker is usually higher than the setting of the lower circuit breaker. When a circuit abnormality occurs in the area controlled by a lower circuit breaker, the lower circuit breaker will be disconnected first because the setting of the lower circuit breaker is smaller, thereby protecting the area controlled by the lower circuit breaker and preventing the power outage from expanding to other areas.

[0189] In order to improve the accuracy of power outage prediction, the embodiment of the present application predicts the power outage area based on the graph convolutional network (GCN). The prediction method includes: first, establishing a graph structure based on the superior-subordinate relationship of the pole-mounted circuit breaker; then extracting features from the graph structure based on GCN to obtain a feature matrix; finally, inputting the extracted feature matrix into the classifier to output the tripping probability of each pole-mounted circuit breaker node, and taking the area controlled by the pole-mounted circuit breaker with a tripping probability higher than the threshold as the predicted power outage area. Among them, the graph convolutional network corresponds to the feature extraction model in other embodiments, the feature matrix corresponds to the first feature matrix in other embodiments, the tripping probability corresponds to the power outage probability in other embodiments, and the threshold corresponds to the probability threshold in other embodiments.

[0190] Fig.10 This is a schematic diagram of a network structure provided by the embodiment of the present application, see Fig.10 , is to establish a graph structure based on the hierarchical relationship of the pole-mounted circuit breakers. The graph structure is a tree structure, each node in the graph structure represents a pole-mounted circuit breaker, and each edge represents the hierarchical relationship between pole-mounted short-circuit breakers. Fig.10 For example, the pole-mounted circuit breaker 1 governs the entire industrial park. When the pole-mounted circuit breaker 1 is disconnected, the entire industrial park will be out of power. The pole-mounted circuit breaker 1 belongs to the first-level pole-mounted circuit breaker. The industrial park includes factories A and B. The pole-mounted circuit breaker 2 governs factory A. When the pole-mounted circuit breaker 2 is disconnected, factory A will be out of power. The pole-mounted circuit breaker 3 governs factory B. When the pole-mounted circuit breaker 3 is disconnected, factory B will be out of power. Therefore, the pole-mounted circuit breakers 2 and 3 are the subordinate nodes of the pole-mounted circuit breaker 1, and the pole-mounted circuit breakers 2 and 3 belong to the second-level pole-mounted circuit breakers. Factory A includes workshops a1 and a2. The pole-mounted circuit breaker 4 governs workshop a1, and the pole-mounted circuit breaker 5 governs workshop a2. Therefore, the pole-mounted circuit breakers 4 and 5 are the subordinate nodes of the pole-mounted circuit breaker 2, and the pole-mounted circuit breakers 4 and 5 belong to the third-level pole-mounted circuit breakers. Factory B includes workshop b1, workshop b2 and workshop b3. Pole-mounted circuit breaker 6 governs workshop b1, pole-mounted circuit breaker 7 governs workshop b2, and pole-mounted circuit breaker 8 governs workshop b3. Therefore, pole-mounted circuit breakers 6, 7 and 8 are lower-level nodes of pole-mounted circuit breaker 3, and pole-mounted circuit breakers 6, 7 and 8 belong to the third-level pole-mounted circuit breakers.

[0191] In the embodiment of the present application, an undirected adjacency matrix is ​​also established for the graph structure, and the undirected adjacency matrix can be recorded as A, wherein the undirected adjacency matrix corresponds to the adjacency matrix in other embodiments. When establishing the undirected adjacency matrix A, there are two ways of establishing it. The first way of establishing it is: not considering the case where a transformer is set between the upper and lower pole-mounted circuit breakers; the second way of establishing it is: considering the case where a transformer is set between the upper and lower pole-mounted circuit breakers.

[0192] For the first construction method, no transformer is installed between the upper and lower column circuit breakers. Fig.10 Taking the graph structure in as an example, based on the first establishment method, the undirected adjacency matrix A1 shown in Table 1 can be obtained. Among them, the matrix element a_ij indicates whether the i-th pole-mounted circuit breaker and the j-th pole-mounted circuit breaker have a superior-subordinate relationship. If the i-th pole-mounted circuit breaker and the j-th pole-mounted circuit breaker have a superior-subordinate relationship, a_ij is equal to 1, so a_12, a_13, a_21, a_24, a_25, a_31, a_36, a_37, a_38, a_42, a_52, a_63, a_73 and a_83 are all equal to 1. If i is equal to j, a_ij is also equal to 1, so a_11, a_22, a_33, a_44, a_55, a_66, a_77 and a_88 are also equal to 1. For other cases, a_ij is equal to 0.

[0193] Table 1 Undirected adjacency matrix A1

[0194]

[0195] For the second establishment method, consider setting a transformer between the upper and lower pole-mounted circuit breakers. Since the transformer has a certain isolation effect in the line, it has a certain isolation effect in isolating fault propagation. This isolation effect can weaken the upper and lower relationship between the pole-mounted circuit breakers. For the situation where there is a transformer between the pole-mounted circuit breakers, an undirected adjacency matrix A2 as shown in Table 2 can be established. Among them, the matrix element a_ij indicates whether the i-th pole-mounted circuit breaker and the j-th pole-mounted circuit breaker have a superior-subordinate relationship, and whether a transformer is set between the two. Still with Fig.10 For example, if the i-th pole-mounted circuit breaker and the j-th pole-mounted circuit breaker have a superior-subordinate relationship and a transformer is set between them, then a_ij is equal to x, where x is a decimal between 0 and 1, and x is determined by learning during the model training process. If the i-th pole-mounted circuit breaker and the j-th pole-mounted circuit breaker have a superior-subordinate relationship, but no transformer is set between them, then a_ij is equal to 1. If i is equal to j, then a_ij is also equal to 1. For other cases, a_ij is equal to 0.

[0196] In some embodiments, assuming that the line managed by pole breaker 1 is a 10 kilovolt (kv) line, and the lines managed by pole breaker 2 and pole breaker 3 are 380 volts (volt, v), 380 v is the industrial power voltage, then a transformer is set between pole breaker 1 and pole breaker 2, and a transformer is also set between pole breaker 1 and pole breaker 3. Based on this, a_12, a_13, a_21 and a_31 in Table 1 are set to x, so as to obtain the undirected adjacency matrix A2 shown in Table 2.

[0197] Table 2 Undirected adjacency matrix A2

[0198]

[0199] In an embodiment of the present application, a feature matrix is ​​established for the graph structure, and the feature matrix can be denoted as X. Each row of the feature matrix X corresponds to a pole-mounted circuit breaker, and the elements in each row include the current constant, real-time current, current trend, voltage constant, real-time voltage, and voltage trend of the pole-mounted circuit breaker. Among them, the current trend and the voltage trend can be predicted by fitting or by a prediction model, and the current trend and the voltage trend need to be normalized to a value between 0 and 1. The feature matrix corresponds to the initial feature matrix in other embodiments, the current constant corresponds to the current setting value in other embodiments, and the voltage constant corresponds to the voltage setting value in other embodiments.

[0200] Continue with Fig.10 Taking the graph structure in as an example, the feature matrix established for the graph structure is shown in Table 3.

[0201] Table 3 Feature matrix

[0202]

[0203] In the embodiment of the present application, after obtaining the undirected adjacency matrix and the feature matrix, GCN is used to extract features. The GCN includes N layers, N is greater than or equal to 2, and the i-th layer performs feature extraction according to the following formula (3):

[0204] Formula (3);

[0205] in, represents a nonlinear activation function, express The degree matrix of represents an undirected adjacency matrix, Indicates the input The feature matrix of the layer, Indicates The parameter matrix of the layer, parameter matrix is determined through learning during model training, Indicates The feature matrix of the layer output, Also as input In this way, after the feature extraction of the Nth layer, the feature matrix extracted by GCN is obtained. Among them, the feature matrix corresponds to the first feature matrix in other embodiments.

[0206] In the embodiment of the present application, the extracted feature matrix is ​​input into a classifier, which includes a fully connected layer and an activation function, to obtain the probability of disconnection of each node, that is, the probability of disconnection of each pole-mounted circuit breaker, and the area governed by the pole-mounted circuit breaker with a disconnection probability higher than a threshold is regarded as an area where power outage may occur. Exemplarily, the activation function may be softmax.

[0207] When training GCN and classifiers, the broken nodes in the sample data can be marked in advance, and then the loss value can be determined based on the broken probability of each node output by the classifier and the marked broken nodes, thereby updating the GCN and classifier.

[0208] Through the regional prediction method provided in the embodiment of the present application, a graph structure is constructed according to the upper and lower level relationships of the pole-mounted circuit breakers, and an adjacency matrix is ​​established for the graph structure, and a feature matrix is ​​established according to the constant value, real-time current or voltage, and current or voltage trend; then feature extraction is performed through GCN, so that each node can aggregate its own features and the constant values, real-time data, change trends and other features of the upper and lower nodes, so as to make predictions based on the extracted features, which can improve the prediction accuracy.

[0209] It can be understood that in the embodiments of the present application, data related to hierarchical relationships, electrical parameters, system configuration information, area identification, etc. are involved, and the collection, use and processing of such relevant data need to comply with relevant laws, regulations and standards.

[0210] The following is a description of an exemplary structure of the region prediction device 455 provided in the embodiment of the present application implemented as a software module. In some embodiments, Figure 2 As shown, the software modules stored in the region prediction device 455 of the memory 450 may include:

[0211] A construction module 4551 is used to construct a graph structure based on the hierarchical relationship between multiple power protection devices in the power distribution system and the electrical parameters of each of the power protection devices; a feature extraction module 4552 is used to extract features from the graph structure to obtain a first feature matrix; a prediction module 4553 is used to perform prediction processing on each of the power protection devices based on the first feature matrix to obtain the power outage probability of each of the power protection devices; a first determination module 4554 is used to determine a power protection device whose power outage probability is higher than a probability threshold as a target power protection device, and determine a control area corresponding to the target power protection device as a power outage area.

[0212] In some embodiments, the construction module 4551 is also used to: determine each of the power protection devices as a node; determine the electrical parameters of each of the power protection devices as the node information of the corresponding node of each of the power protection devices; based on the hierarchical relationship between the multiple power protection devices, create a connecting edge between the corresponding nodes of two power protection devices with a superior-subordinate relationship; and generate the graph structure based on the nodes, the node information and the connecting edges.

[0213] In some embodiments, the feature extraction module 4552 is also used to: determine the adjacency matrix and degree matrix of the graph structure; construct an initial feature matrix of the graph structure based on the node information; and determine a first feature matrix based on the adjacency matrix, the degree matrix and the initial feature matrix using a feature extraction model.

[0214] In some embodiments, the feature extraction module 4552 is also used for: when there is a connecting edge between the i-th node and the j-th node, or i is equal to j, setting the matrix element in the i-th row and j-th column to a first preset value, i and j are natural numbers less than or equal to M, and M is the total number of the nodes; when there is no connecting edge between the i-th node and the j-th node, and i is not equal to j, setting the matrix element in the i-th row and j-th column to a second preset value, which is different from the first preset value; and constructing an adjacency matrix of the graph structure based on each of the matrix elements.

[0215] In some embodiments, the feature extraction module 4552 is also used to: when there is a connecting edge between the i-th node and the j-th node, obtain the system configuration information of the distribution system, and determine the matrix element of the i-th row and j-th column based on the configuration information, i and j are natural numbers less than or equal to M, and M is the total number of the nodes; when there is no connecting edge between the i-th node and the j-th node, and i is not equal to j, set the matrix element of the i-th row and j-th column to a second preset value; when i is equal to j, set the matrix element of the i-th row and j-th column to a first preset value, and the first preset value is different from the second preset value; construct an adjacency matrix of the graph structure based on each of the matrix elements.

[0216] In some embodiments, the power protection device is a pole-mounted circuit breaker, and the feature extraction module 4552 is also used for: when it is determined based on the configuration information that a transformer is provided between the i-th pole-mounted circuit breaker corresponding to the i-th node and the j-th pole-mounted circuit breaker corresponding to the j-th node, setting the matrix element of the i-th row and the j-th column to a third value, wherein the third value is determined during the training process of the feature extraction model, and the third value is between the first preset value and the second preset value; when it is determined based on the configuration information that no transformer is provided between the i-th pole-mounted circuit breaker and the j-th pole-mounted circuit breaker, setting the matrix element of the i-th row and the j-th column to the first preset value.

[0217] In some embodiments, the node information includes current setting value, real-time current, current trend mark, voltage setting value, real-time voltage and voltage trend mark, and the feature extraction module 4552 is also used to: determine the current setting value, real-time current, current trend mark, voltage setting value, real-time voltage and voltage trend mark of the i-th node as the i-th row vector, i is a natural number less than or equal to M, and M is the total number of the nodes; based on the first row vector to the M-th row vector, construct the initial feature matrix of the graph structure.

[0218] In some embodiments, the feature extraction model includes N feature extraction layers, and the feature extraction module 4552 is also used to: based on the adjacency matrix, the degree matrix and the parameter matrix of the feature extraction model, linearly process the initial feature matrix to obtain a second feature matrix; and perform nonlinear processing on the second feature matrix to obtain a first output matrix of the first feature extraction layer; based on the adjacency matrix, the degree matrix and the parameter matrix, linearly process the kth output matrix to obtain a third feature matrix; and perform nonlinear processing on the third feature matrix to obtain the k+1th output matrix of the k+1th feature extraction layer; wherein k is a natural number greater than 0 and less than N-1; and determine the Nth output matrix as the first feature matrix.

[0219] In some embodiments, the software modules stored in the regional prediction device 455 of the memory 450 may include: a second determination module, used to generate an alarm message based on the regional identification of the power outage area; and determine the target object based on the regional identification; and a sending module, used to send the alarm message to the terminal of the target object.

[0220] The embodiment of the present application provides a computer program product or a computer program, which includes computer executable instructions, and the computer executable instructions are stored in a computer readable storage medium. The processor of the computer device reads the computer executable instructions from the computer readable storage medium, and the processor executes the computer executable instructions, so that the computer device performs the above-mentioned regional prediction method of the embodiment of the present application.

[0221] 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 be caused to execute the region prediction method provided by the embodiment of the present application, for example, Figure 3 , 8 The regional prediction method is shown.

[0222] In some embodiments, the computer-readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, flash memory, magnetic surface storage, optical disk, or CD-ROM; or it may be various devices including one or any combination of the above memories.

[0223] In some embodiments, computer 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.

[0224] As an example, computer-executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file storing other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files storing one or more modules, subroutines, or code portions).

[0225] By way of example, computer executable instructions may be deployed to be executed on one computing device or on multiple computing devices located at one site or on multiple computing devices distributed across multiple sites and interconnected by a communication network.

[0226] 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 regional prediction method, characterized in that: The method comprises: Determine each power protection device in the power distribution system as a node; determine the electrical parameters of each power protection device as the node information of each node corresponding to the power protection device; create a connection edge between two nodes corresponding to the power protection devices having a superior-subordinate relationship based on the hierarchical relationship between the plurality of power protection devices; generate a graph structure based on the nodes, the node information and the connection edge; Determine an adjacency matrix and a degree matrix of the graph structure; construct an initial feature matrix of the graph structure based on the node information; determine a first feature matrix based on the adjacency matrix, the degree matrix and the initial feature matrix using a feature extraction model; Wherein, the power protection device is a pole-mounted circuit breaker, and determining the adjacency matrix of the graph structure includes: when there is a connecting edge between the i-th node and the j-th node, obtaining system configuration information of the power distribution system, and when it is determined based on the configuration information that a transformer is provided between the i-th pole-mounted circuit breaker corresponding to the i-th node and the j-th pole-mounted circuit breaker corresponding to the j-th node, setting the matrix element of the i-th row and the j-th column to a third value, wherein the third value is determined during the training process of the feature extraction model, and the third value is between a first preset value and a second preset value; when it is determined based on the configuration information that a transformer is provided between the i-th pole-mounted circuit breaker corresponding to the i-th node and the j-th pole-mounted circuit breaker corresponding to the j-th node, setting the matrix element of the i-th row and the j-th column to a third value. When no transformer is provided between the upper circuit breaker and the j-th pole upper circuit breaker, the matrix element in the i-th row and j-th column is set to the first preset value, i and j are natural numbers less than or equal to M, and M is the total number of the nodes; when there is no connecting edge between the i-th node and the j-th node, and i is not equal to j, the matrix element in the i-th row and j-th column is set to the second preset value; when i is equal to j, the matrix element in the i-th row and j-th column is set to the first preset value, and the first preset value is different from the second preset value; an adjacency matrix of the graph structure is constructed based on each of the matrix elements; Based on the first characteristic matrix, predicting each of the power protection devices is performed to obtain a power failure probability of each of the power protection devices; A power protection device with a power outage probability higher than a probability threshold is determined as a target power protection device, and a control area corresponding to the target power protection device is determined as a power outage area.

2. The method according to claim 1, characterized in that: The determining of the adjacency matrix of the graph structure further includes: When there is a connecting edge between the i-th node and the j-th node, or i is equal to j, the matrix element in the i-th row and j-th column is set to a first preset value, where i and j are natural numbers less than or equal to M, and M is the total number of the nodes; When there is no connecting edge between the i-th node and the j-th node, and i is not equal to j, setting the matrix element in the i-th row and j-th column to a second preset value, where the second preset value is different from the first preset value; An adjacency matrix of the graph structure is constructed based on each of the matrix elements.

3. The method according to claim 1, characterized in that: The node information includes current setting value, real-time current, current trend mark, voltage setting value, real-time voltage and voltage trend mark. Based on the node information, the initial characteristic matrix of the graph structure is constructed, including The current setting value, real-time current, current trend mark, voltage setting value, real-time voltage and voltage trend mark of the i-th node are determined as the i-th row vector, where i is a natural number less than or equal to M, and M is the total number of the nodes; Based on the first row vector to the Mth row vector, an initial feature matrix of the graph structure is constructed.

4. The method according to claim 1, characterized in that The feature extraction model includes N feature extraction layers, and the use of the feature extraction model to determine a first feature matrix based on the adjacency matrix, the degree matrix and the initial feature matrix includes: Based on the adjacency matrix, the degree matrix and the parameter matrix of the feature extraction model, linearly process the initial feature matrix to obtain a second feature matrix; and nonlinearly process the second feature matrix to obtain a first output matrix of the first feature extraction layer; Based on the adjacency matrix, the degree matrix and the parameter matrix, linearly process the kth output matrix to obtain a third feature matrix; and nonlinearly process the third feature matrix to obtain the k+1th output matrix of the k+1th feature extraction layer; wherein k is a natural number greater than 0 and less than N-1; The Nth output matrix is ​​determined as the first characteristic matrix.

5. The method according to any one of claims 1 to 4, characterized in that: After determining the control area corresponding to the target power protection device as a power outage area, the method further includes: Based on the area identification of the power outage area, an alarm message is generated; and based on the area identification, a target object is determined; Send the warning message to the terminal of the target object.

6. A region prediction device, characterized in that: The device comprises: A construction module is used to determine each power protection device in the power distribution system as a node; determine the electrical parameters of each of the power protection devices as the node information of the node corresponding to each of the power protection devices; based on the hierarchical relationship between the plurality of power protection devices, create a connection edge between the nodes corresponding to two power protection devices with a superior-subordinate relationship; and generate a graph structure based on the nodes, the node information and the connection edge; A feature extraction module, configured to obtain system configuration information of the power distribution system when there is a connection edge between the i-th node and the j-th node, and when it is determined based on the configuration information that a transformer is provided between the i-th pole-mounted circuit breaker corresponding to the i-th node and the j-th pole-mounted circuit breaker corresponding to the j-th node, set the matrix element of the i-th row and the j-th column to a third value, wherein the third value is determined during the training process of the feature extraction model, and the third value is between a first preset value and a second preset value; when it is determined based on the configuration information that a transformer is not provided between the i-th pole-mounted circuit breaker and the j-th pole-mounted circuit breaker, set the matrix element of the i-th row and the j-th column to the first preset value, where i and j are less than or equal to M is a natural number, where M is the total number of the nodes; when there is no connecting edge between the i-th node and the j-th node, and i is not equal to j, the matrix element of the i-th row and j-th column is set to the second preset value; when i is equal to j, the matrix element of the i-th row and j-th column is set to the first preset value, and the first preset value is different from the second preset value; construct an adjacency matrix of the graph structure based on each of the matrix elements; determine the degree matrix of the graph structure; construct an initial feature matrix of the graph structure based on the node information; use a feature extraction model to determine a first feature matrix based on the adjacency matrix, the degree matrix and the initial feature matrix; the power protection device is a pole-mounted circuit breaker, A prediction module, configured to perform prediction processing on each of the power protection devices based on the first characteristic matrix to obtain a power outage probability of each of the power protection devices; The first determination module is used to determine a power protection device with a power outage probability higher than a probability threshold as a target power protection device, and determine a control area corresponding to the target power protection device as a power outage area.

7. A computer device, characterized in that: The computer device comprises: A memory for storing computer executable instructions; A processor, configured to implement the region prediction method according to any one of claims 1 to 5 when executing the computer executable instructions stored in the memory.

8. A computer-readable storage medium storing computer-executable instructions, characterized in that: When the computer executable instructions are executed by a processor, the region prediction method according to any one of claims 1 to 5 is implemented.

9. A computer program product comprising a computer program or computer executable instructions, characterized in that: When the computer program or computer executable instructions are executed by a processor, the region prediction method according to any one of claims 1 to 5 is implemented.

Images