Power distribution network fault positioning method, device and equipment based on graph convolutional neural network

Abstract

The invention relates to the technical field of power distribution network fault positioning and the technical field of artificial intelligence application, and discloses a power distribution network fault positioning method, device and equipment based on a graph convolutional neural network, and the method comprises the steps: obtaining the total number of individuals and the maximum number of iterations from control parameters, and generating a plurality of cordyceps sinensis individuals according to the total number of individuals; updating the position vector of each cordyceps sinensis individual by adopting a cordyceps sinensis optimization algorithm, obtaining the current iteration times during the updating period, and when the current iteration times are the maximum iteration times, selecting a hyper-parameter combination represented by the position vector of the cordyceps sinensis individual with the highest fitness value as an optimal hyper-parameter combination, and importing the optimal hyper-parameter combination into the architecture of the graph convolutional neural network, generating a fault positioning model of the power distribution network, inputting the electrical data of the power distribution network at the current moment into the fault positioning model, and generating a fault positioning result of the power distribution network at the current moment by the fault positioning model. According to the invention, the acquisition efficiency of the fault positioning result is improved.

Description

Technical Field

[0001] This application relates to the fields of power distribution network fault location technology and artificial intelligence application technology, and in particular to a power distribution network fault location method, device and equipment based on graph convolutional neural network. Background Technology

[0002] With the large-scale integration of distributed power sources, energy storage devices, and charging piles, the coverage of distribution networks is constantly expanding, and the types of faults in distribution networks are becoming increasingly diversified. The difficulty of tracing fault propagation paths has significantly increased, which places higher demands on fault location in distribution networks.

[0003] However, traditional distribution network fault location technology has shown many limitations in current application scenarios. This is because the fault indicators used in traditional distribution network fault location rely on a certain installation density to achieve fault location. Once the number of fault indicators is insufficient, there will be monitoring blank areas in the distribution network. Since the blank areas cannot be detected by the fault indicators, there will be situations where faults cannot be located after they occur. Therefore, how to generate fault location results for distribution networks is a technical problem that urgently needs to be solved. Summary of the Invention

[0004] This application provides a method, apparatus, and device for fault location in power distribution networks based on graph convolutional neural networks, in order to solve the aforementioned technical problem of how to generate fault location results in power distribution networks.

[0005] In a first aspect, embodiments of this application provide a method for fault location in a power distribution network based on a graph convolutional neural network, applied to electronic equipment. The method for fault location in a power distribution network includes: Obtain electrical data of the distribution network at a preset time and fault location results of the distribution network at a preset time, and combine the electrical data of the distribution network at a preset time and fault location results of the distribution network at a preset time into a fault sample; Different fault samples are divided into training and validation sets to obtain the control parameters of the Cordyceps sinensis optimization algorithm. The total number of individuals and the maximum number of iterations are obtained from the control parameters. Based on the total number of individuals, multiple Cordyceps sinensis individuals are generated. Each Cordyceps sinensis individual is an independent search unit, and the position vector of each Cordyceps sinensis individual represents each hyperparameter combination. The hyperparameter combination corresponding to the position vector of each individual Cordyceps sinensis is imported into the architecture of a graph convolutional neural network to generate a graph convolutional neural network model corresponding to each individual Cordyceps sinensis. The graph convolutional neural network model corresponding to each individual Cordyceps sinensis is trained using the training set to obtain a trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis. The training graph convolutional neural network model corresponding to each individual Cordyceps sinensis was validated using a validation set. The accuracy of the training graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set was obtained, and the accuracy of the training graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set was selected as the fitness value of each individual Cordyceps sinensis. The position vector of each individual Cordyceps sinensis is updated using a Cordyceps sinensis optimization algorithm. The current iteration number is obtained during the update. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. The electrical data of the power distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the power distribution network at the current moment.

[0006] In one possible implementation of the first aspect, the position vector of each individual Cordyceps sinensis is updated using a Cordyceps sinensis optimization algorithm. The current iteration number is obtained during the update. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. The electrical data of the power distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the power distribution network at the current moment, including: The first or second operator in the Cordyceps sinensis optimization algorithm is used to update the position vector of each individual Cordyceps sinensis. The current iteration number during the update is obtained. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal combination of hyperparameters is imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. The system receives data uploaded by the acquisition equipment, obtains the electrical data of the power distribution network at the current moment from the uploaded data, and inputs the electrical data of the power distribution network at the current moment into the fault location model. The fault location model then generates the fault location result of the power distribution network at the current moment.

[0007] In one possible implementation of the first aspect, when the position vector of each individual Cordyceps sinensis is updated using the Cordyceps sinensis optimization algorithm, the current iteration number during the update is obtained. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. After the electrical data of the power distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the power distribution network at the current moment, the power distribution network fault location method includes: From the fault location results, obtain the line identifier of the faulty line in the distribution network, send the line identifier of the faulty line to the monitoring system of the distribution network, and receive the response data returned by the monitoring system based on the line identifier of the faulty line. Obtain the ground voltage at the first end of the faulty line from the response data. Based on the ground voltage at the first end of the faulty line, the path length from the fault point to the first end of the faulty line is generated according to the preset length model.

[0008] In one possible implementation of the first aspect, the step of obtaining the line identifier of the faulty line in the distribution network from the fault location result, sending the line identifier of the faulty line to the monitoring system of the distribution network, and receiving response data returned by the monitoring system based on the line identifier of the faulty line includes: The fault location results are analyzed to obtain the line identifier of the faulty line in the distribution network. The line identifier of the faulty line is sent to the monitoring system of the distribution network, and the response data returned by the monitoring system based on the line identifier of the faulty line is received.

[0009] In one possible implementation of the first aspect, after obtaining the ground voltage at the starting point of the faulty line and the ground voltage at the starting point of the faulty line from the response data, and generating the path length from the fault point to the starting point of the faulty line based on the ground voltage at the starting point of the faulty line, the ground voltage at the starting point of the faulty line, and a preset length model, the distribution network fault location method includes: Create a display window to show the line identifier of the faulty line and the path length from the fault point to the beginning and end points of the faulty line.

[0010] In one possible implementation of the first aspect, the length model is defined as follows: ; This indicates the path length from the fault point to the beginning of the faulty line. This indicates the voltage to ground at the starting point of the faulty line; This indicates the voltage to ground at the end of the faulty line. This indicates the path length from the first measurement point to the beginning of the faulty line; This indicates the path length from the second measurement point to the end point of the faulty line; Indicates the total path length of the faulty line; This represents the voltage to ground at the first measurement point; This represents the voltage to ground at the second measurement point.

[0011] In one possible implementation of the first aspect, the first operator is defined as follows: ; ; This represents the first search coefficient. The larger the first search coefficient, the larger the update magnitude of the first operator; the smaller the first search coefficient, the smaller the update magnitude of the first operator. This represents pi, with a value of 3.14. Represents the cosine function; Indicates the first The updated position vector of each individual Cordyceps sinensis. It was obtained by updating the first operator in the Cordyceps sinensis optimization algorithm; The serial number representing an individual Cordyceps sinensis; Indicates the first The position vector of each individual Cordyceps sinensis before updating; Indicates the first The position vector of the previous Cordyceps sinensis individual before the update; This represents the position vector of the Cordyceps sinensis individual with the highest fitness value. Indicates the column number; This indicates the generation of a string with 1 row and 1 column. A random vector.

[0012] In one possible implementation of the first aspect, the second operator is defined as follows: ; ; This represents the second search coefficient. The larger the second search coefficient, the larger the search range of the second operator; the smaller the second search coefficient, the smaller the search range of the second operator. Indicates the current iteration number; Indicates the maximum number of iterations; Indicates a A random number uniformly distributed within a range; It is a random integer, which can be either 1 or 2; Indicates the first The updated position vector of each individual Cordyceps sinensis. It was obtained by updating the second operator in the Cordyceps sinensis optimization algorithm; The serial number representing an individual Cordyceps sinensis; Indicates the first The position vector of each individual Cordyceps sinensis before updating; Indicates the first The position vector of the previous Cordyceps sinensis individual before the update; This represents the position vector of the Cordyceps sinensis individual with the highest fitness value. Indicates the column number; This indicates the generation of a string with 1 row and 1 column. A random vector.

[0013] Secondly, embodiments of this application provide a power distribution network fault location device based on a graph convolutional neural network, applied to electronic equipment, comprising: The first acquisition module is used to acquire the electrical data of the distribution network at a preset time and the fault location result of the distribution network at a preset time, and to combine the electrical data of the distribution network at a preset time and the fault location result of the distribution network at a preset time into a fault sample. The second acquisition module is used to divide different fault samples into training sets and validation sets, acquire the control parameters of the Cordyceps sinensis optimization algorithm, obtain the total number of individuals and the maximum number of iterations from the control parameters, generate multiple Cordyceps sinensis individuals based on the total number of individuals, each Cordyceps sinensis individual is an independent search unit, and the position vector of each Cordyceps sinensis individual represents each hyperparameter combination. The training module is used to combine the hyperparameters corresponding to the position vector of each individual Cordyceps sinensis, import them into the architecture of a graph convolutional neural network, generate a graph convolutional neural network model corresponding to each individual Cordyceps sinensis, and train the graph convolutional neural network model corresponding to each individual Cordyceps sinensis using the training set to obtain a trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis. The validation module is used to validate the trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis using a validation set, obtain the accuracy of the trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set, and select the accuracy of the trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set as the fitness value of each individual Cordyceps sinensis. The generation module is used to update the position vector of each individual Cordyceps sinensis using the Cordyceps sinensis optimization algorithm, obtain the current iteration number during the update, and when the current iteration number is the maximum iteration number, select the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the distribution network. The electrical data of the distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the distribution network at the current moment.

[0014] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the power distribution network fault location method based on graph convolutional neural networks described in the first aspect above.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the power distribution network fault location method based on graph convolutional neural networks described in the first aspect above.

[0016] Fifthly, embodiments of this application provide a computer program product that, when run on an electronic device, causes the electronic device to execute the power distribution network fault location method based on graph convolutional neural networks described in the first aspect.

[0017] The beneficial effects of the embodiments of this application are as follows: Firstly, a Cordyceps sinensis optimization algorithm is used to update the position vector of each individual Cordyceps sinensis. The current iteration number is obtained during the update. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. The electrical data of the power distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the power distribution network at the current moment. This solves the technical problem of how to generate the fault location result of the power distribution network. Since no manual operation is required, the generation time of the fault location result is reduced, which is conducive to improving the efficiency of obtaining the fault location result. Secondly, since the fault location model generates the fault location results of the distribution network at the current moment, it is not affected by human intervention, which helps to improve the reliability of the fault location results of the distribution network at the current moment. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is an application scenario diagram of the power distribution network fault location method provided in the embodiments of this application; Figure 2 This is a flowchart illustrating the power distribution network fault location method provided in the embodiments of this application; Figure 3 A flowchart illustrating the implementation of S205 provided in this application embodiment; Figure 4 A schematic block diagram of a power distribution network fault location device provided in the embodiments of this application; Figure 5 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application; Figure 6 This is an example diagram of a faulty line in a power distribution network provided in an embodiment of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0021] The power distribution network fault location method provided in this application embodiment can be applied to electronic devices, including but not limited to servers, mobile phones, tablets, wearable devices, vehicle-mounted devices, and laptops. This application embodiment does not impose any restrictions on the specific type of electronic device.

[0022] Please see Figure 1 , Figure 1 The application scenario diagram of the power distribution network fault location method provided in the embodiments of this application is described in detail below: Electronic equipment accesses the database, retrieves the electrical data of the distribution network at a preset time and the fault location results of the distribution network at a preset time from the database, and combines the electrical data of the distribution network at a preset time and the fault location results of the distribution network at a preset time into a fault sample.

[0023] Among them, the database, as the core hub for data storage and management, can integrate massive electrical data of the power distribution network at a preset time and fault location results of the power distribution network at a preset time.

[0024] In this embodiment of the application, the electronic device obtains the electrical data of the distribution network at a preset time and the fault location result of the distribution network at a preset time from the database, which is beneficial to quickly improve the electrical data of the distribution network at a preset time and the fault location result of the distribution network at a preset time.

[0025] Please see Figure 2 , Figure 2 This is a flowchart illustrating the power distribution network fault location method provided in this application embodiment, which can be applied to electronic equipment.

[0026] like Figure 2 As shown in the figure, the power distribution network fault location method provided in this application includes the following steps, which are detailed below: S201, acquire the electrical data of the distribution network at a preset time and the fault location result of the distribution network at a preset time, and combine the electrical data of the distribution network at a preset time and the fault location result of the distribution network at a preset time into a fault sample. Among them, the electrical data of the distribution network at a preset time and the fault location results of the distribution network at a preset time are combined into a fault sample. This can construct a mapping relationship between the electrical data of the distribution network at a preset time and the fault location results of the distribution network at a preset time, which not only preserves the operating characteristics of the distribution network parameters such as voltage, current, and power at the preset time, but also anchors the specific location of the fault.

[0027] S202, different fault samples are divided into training set and validation set, the control parameters of the Cordyceps sinensis optimization algorithm are obtained, the total number of individuals and the maximum number of iterations are obtained from the control parameters, and multiple Cordyceps sinensis individuals are generated according to the total number of individuals. Each Cordyceps sinensis individual is an independent search unit, and the position vector of each Cordyceps sinensis individual represents each hyperparameter combination. S203, combine the hyperparameters corresponding to the position vector of each individual Cordyceps sinensis, import them into the architecture of a graph convolutional neural network, generate a graph convolutional neural network model corresponding to each individual Cordyceps sinensis, use the training set to train the graph convolutional neural network model corresponding to each individual Cordyceps sinensis, and obtain a trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis. Among them, the Caterpillar Fungus Optimizer (CFO) is a novel metaheuristic intelligent optimization algorithm.

[0028] S204. The training graph convolutional neural network model corresponding to each individual Cordyceps sinensis is validated using a validation set to obtain the accuracy of the training graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set. The accuracy of the training graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set is selected as the fitness value of each individual Cordyceps sinensis. The higher the fitness value of an individual Cordyceps sinensis, the better the performance of the hyperparameter combination corresponding to the position vector of the individual Cordyceps sinensis on the validation set; the lower the fitness value of an individual Cordyceps sinensis, the worse the performance of the hyperparameter combination corresponding to the position vector of the individual Cordyceps sinensis on the validation set.

[0029] For ease of explanation, the following example is provided: For example, there are C individual Cordyceps sinensis, namely individual Cordyceps sinensis A, individual Cordyceps sinensis B, and individual Cordyceps sinensis C; The graph convolutional neural network model corresponding to individual Cordyceps sinensis A after training is graph convolutional neural network model A; The graph convolutional neural network model corresponding to individual Cordyceps sinensis B after training is graph convolutional neural network model B; The graph convolutional neural network model corresponding to the individual Cordyceps sinensis C after training is graph convolutional neural network model C; The accuracy of graph convolutional neural network model A on the validation set is selected as the fitness value of individual Cordyceps sinensis A; the accuracy of graph convolutional neural network model B on the validation set is selected as the fitness value of individual Cordyceps sinensis B; and the accuracy of graph convolutional neural network model C on the validation set is selected as the fitness value of individual Cordyceps sinensis C.

[0030] For example, among the fitness values ​​of individual Cordyceps sinensis A, B, and C, the highest fitness value for individual Cordyceps sinensis A indicates that the hyperparameter combination corresponding to individual Cordyceps sinensis A performs better on the validation set. Conversely, the lowest fitness value for individual Cordyceps sinensis B indicates that the hyperparameter combination corresponding to individual Cordyceps sinensis B performs worse on the validation set.

[0031] For example, among the fitness values ​​of individual Cordyceps sinensis A, B, and C, the highest fitness value for individual Cordyceps sinensis C indicates that the hyperparameter combination corresponding to individual Cordyceps sinensis C performs better on the validation set. Conversely, the lowest fitness value for individual Cordyceps sinensis B indicates that the hyperparameter combination corresponding to individual Cordyceps sinensis B performs worse on the validation set.

[0032] S205. The position vector of each individual Cordyceps sinensis is updated using the Cordyceps sinensis optimization algorithm. The current iteration number is obtained during the update. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal hyperparameter combination is imported into the architecture of a graph convolutional neural network to generate a fault location model for the distribution network. The electrical data of the distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the distribution network at the current moment.

[0033] The method for locating a distribution network fault includes the following steps: First, the position vector of each individual Cordyceps sinensis is updated using a Cordyceps sinensis optimization algorithm. The current iteration number is obtained during the update process. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. This optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the distribution network. After the electrical data of the distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result for the distribution network at the current moment, the method further includes: From the fault location results, obtain the line identifier of the faulty line in the distribution network, send the line identifier of the faulty line to the monitoring system of the distribution network, and receive the response data returned by the monitoring system based on the line identifier of the faulty line. Obtain the ground voltage at the first end of the faulty line from the response data. Based on the ground voltage at the first end of the faulty line, the path length from the fault point to the first end of the faulty line is generated according to the preset length model.

[0034] The steps of obtaining the line identifier of the faulty line in the distribution network from the fault location results, sending the line identifier of the faulty line to the monitoring system of the distribution network, and receiving response data returned by the monitoring system based on the line identifier of the faulty line include: The fault location results are analyzed to obtain the line identifier of the faulty line in the distribution network. The line identifier of the faulty line is sent to the monitoring system of the distribution network, and the response data returned by the monitoring system based on the line identifier of the faulty line is received.

[0035] The distribution network fault location method, after obtaining the ground voltage at the starting point of the faulty line from the response data and generating the path length from the fault point to the starting point of the faulty line based on the ground voltage at the starting point of the faulty line, the ground voltage at the starting point of the faulty line, and a preset length model, includes: Create a display window to show the line identifier of the faulty line and the path length from the fault point to the beginning and end points of the faulty line.

[0036] The length model is defined as follows: ; This indicates the path length from the fault point to the beginning of the faulty line. This indicates the voltage to ground at the starting point of the faulty line; This indicates the voltage to ground at the end of the faulty line. This indicates the path length from the first measurement point to the beginning of the faulty line; This indicates the path length from the second measurement point to the end point of the faulty line; Indicates the total path length of the faulty line; This represents the voltage to ground at the first measurement point; This represents the voltage to ground at the second measurement point.

[0037] Figure 6 Example diagrams of faulty lines in a distribution network provided in this application embodiment are detailed below: exist Figure 6 In the diagram, the black lines represent faulty lines in the distribution network, the gray circles represent the connecting components of the faulty lines, and the black circles represent the fault points of the faulty lines. Indicates the starting and ending points of the faulty line. Indicates the end point of the faulty line. Indicate the fault point; This indicates the path length from the fault point to the beginning of the faulty line. This indicates the path length from the first measurement point to the beginning of the faulty line; This indicates the path length from the second measurement point to the end point of the faulty line; This represents the voltage to ground at the first measurement point; This represents the voltage to ground at the second measurement point.

[0038] The total path length of the faulty line is crucial. As the terminal link of the power system directly facing users, the impact range of a fault in the distribution network is closely related to the distance from the fault point to the starting point. By determining the path length from the fault point to the starting point of the faulty line, the distribution network monitoring system can quickly identify the power supply areas, number of users, and critical loads affected by the fault.

[0039] For ease of explanation, the following example is provided: For example, a 10kV distribution network line in an industrial park connects to the park's substation at its first end, with a total length of 8km, passing five precision manufacturing enterprises and one employee dormitory area. When the path length from the fault point to the starting point of the faulty line is determined to be 3.5km, the distribution network monitoring system can quickly determine that the fault point affects two manufacturing enterprises and the dormitory area. It can immediately transfer this part of the load to the park's backup line through the distribution network automation system, and at the same time dispatch a repair team to the fault point to handle it. This not only ensures the continuous power supply to the critical production load, but also reduces the power outage inconvenience to the residents of the dormitory area.

[0040] The beneficial effects of the embodiments of this application are as follows: Firstly, a Cordyceps sinensis optimization algorithm is used to update the position vector of each individual Cordyceps sinensis. The current iteration number is obtained during the update. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. The electrical data of the power distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the power distribution network at the current moment. This solves the technical problem of how to generate the fault location result of the power distribution network. Since no manual operation is required, the generation time of the fault location result is reduced, which is conducive to improving the efficiency of obtaining the fault location result. Secondly, since the fault location model generates the fault location results of the distribution network at the current moment, it is not affected by human intervention, which helps to improve the reliability of the fault location results of the distribution network at the current moment.

[0041] Please see Figure 3 , Figure 3 The implementation flowchart of S205 provided in the embodiments of this application is described in detail below: S301, using the first or second operator in the Cordyceps sinensis optimization algorithm, update the position vector of each individual Cordyceps sinensis, obtain the current iteration number during the update period, and when the current iteration number is the maximum iteration number, select the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value as the optimal hyperparameter combination. S302, the optimal hyperparameter combination is imported into the architecture of the graph convolutional neural network to generate the fault location model of the distribution network. The data uploaded by the acquisition device is received, the electrical data of the distribution network at the current moment is obtained from the uploaded data, and the electrical data of the distribution network at the current moment is input into the fault location model. The fault location model generates the fault location result of the distribution network at the current moment.

[0042] The first operator is defined as follows: ; ; This represents the first search coefficient. The larger the first search coefficient, the larger the update magnitude of the first operator; the smaller the first search coefficient, the smaller the update magnitude of the first operator. This represents pi, with a value of 3.14. Represents the cosine function; Indicates the first The updated position vector of each individual Cordyceps sinensis. It was obtained by updating the first operator in the Cordyceps sinensis optimization algorithm; The serial number representing an individual Cordyceps sinensis; Indicates the first The position vector of each individual Cordyceps sinensis before updating; Indicates the first The position vector of the previous Cordyceps sinensis individual before the update; This represents the position vector of the Cordyceps sinensis individual with the highest fitness value. Indicates the column number; This indicates the generation of a string with 1 row and 1 column. A random vector.

[0043] The second operator is defined as follows: ; ; This represents the second search coefficient. The larger the second search coefficient, the larger the search range of the second operator; the smaller the second search coefficient, the smaller the search range of the second operator. Indicates the current iteration number; Indicates the maximum number of iterations; Indicates a A random number uniformly distributed within a range; It is a random integer, which can be either 1 or 2; Indicates the first The updated position vector of each individual Cordyceps sinensis. It was obtained by updating the second operator in the Cordyceps sinensis optimization algorithm; The serial number representing an individual Cordyceps sinensis; Indicates the first The position vector of each individual Cordyceps sinensis before updating; Indicates the first The position vector of the previous Cordyceps sinensis individual before the update; This represents the position vector of the Cordyceps sinensis individual with the highest fitness value. Indicates the column number; This indicates the generation of a string with 1 row and 1 column. A random vector.

[0044] In this embodiment of the application, the fault location model generates the fault location result of the distribution network at the current moment. The fault location result helps to improve the operation and maintenance efficiency of the distribution network.

[0045] For the distribution network fault location method described in the above embodiments, please refer to [link / reference]. Figure 4 , Figure 4 This is a schematic block diagram of the power distribution network fault location device provided in the embodiments of this application. Figure 4 The distribution network fault location device 400 shown can be applied to, for example... Figure 1 The application scenario diagram shows electronic devices. The following section uses electronic devices as an example to illustrate this. Figure 4 The distribution network fault location device 400 shown is described in detail. The distribution network fault location device 400 may include a first acquisition module 401, a second acquisition module 402, a training module 403, a verification module 404, and a generation module 405.

[0046] The first acquisition module 401 is used to acquire electrical data of the distribution network at a preset time and fault location results of the distribution network at a preset time, and to combine the electrical data of the distribution network at a preset time and fault location results of the distribution network at a preset time into a fault sample. The second acquisition module 402 is used to divide different fault samples into training sets and validation sets, acquire the control parameters of the Cordyceps sinensis optimization algorithm, obtain the total number of individuals and the maximum number of iterations from the control parameters, generate multiple Cordyceps sinensis individuals based on the total number of individuals, each Cordyceps sinensis individual is an independent search unit, and the position vector of each Cordyceps sinensis individual represents each hyperparameter combination. Training module 403 is used to combine the hyperparameters corresponding to the position vector of each individual Cordyceps sinensis, import them into the architecture of a graph convolutional neural network, generate a graph convolutional neural network model corresponding to each individual Cordyceps sinensis, and train the graph convolutional neural network model corresponding to each individual Cordyceps sinensis using the training set to obtain a trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis. The validation module 404 is used to validate the trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis using a validation set, to obtain the accuracy of the trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set, and to select the accuracy of the trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set as the fitness value of each individual Cordyceps sinensis. The generation module 405 is used to update the position vector of each individual Cordyceps sinensis using the Cordyceps sinensis optimization algorithm, obtain the current iteration number during the update, and when the current iteration number is the maximum iteration number, select the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the distribution network. The electrical data of the distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the distribution network at the current moment.

[0047] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0048] The beneficial effects of the embodiments of this application are as follows: Firstly, a Cordyceps sinensis optimization algorithm is used to update the position vector of each individual Cordyceps sinensis. The current iteration number is obtained during the update. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. The electrical data of the power distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the power distribution network at the current moment. This solves the technical problem of how to generate the fault location result of the power distribution network. Since no manual operation is required, the generation time of the fault location result is reduced, which is conducive to improving the efficiency of obtaining the fault location result. Secondly, since the fault location model generates the fault location results of the distribution network at the current moment, it is not affected by human intervention, which helps to improve the reliability of the fault location results of the distribution network at the current moment.

[0049] Please see Figure 5 , Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0050] like Figure 5 As shown, Figure 5 The electronic device includes: at least one processor 20, a memory 21, and a computer program 22 stored in the memory 21 and executable on the at least one processor 20, wherein the processor 20 executes the computer program 22 to implement the steps in any of the above method embodiments.

[0051] The electronic device may include, but is not limited to, processor 20 and memory 21. Those skilled in the art will understand that... Figure 5 This is merely an example of an electronic device and does not constitute a limitation on electronic devices. It may include more or fewer components than shown in the illustration, or combinations of certain components, or different components. For example, it may also include input / output devices, network access devices, etc.

[0052] The processor 20 is used to run a computer program 22 stored in the memory 21, and performs the following steps when executing the computer program 22: Obtain electrical data of the distribution network at a preset time and fault location results of the distribution network at a preset time, and combine the electrical data of the distribution network at a preset time and fault location results of the distribution network at a preset time into a fault sample; Different fault samples are divided into training and validation sets to obtain the control parameters of the Cordyceps sinensis optimization algorithm. The total number of individuals and the maximum number of iterations are obtained from the control parameters. Based on the total number of individuals, multiple Cordyceps sinensis individuals are generated. Each Cordyceps sinensis individual is an independent search unit, and the position vector of each Cordyceps sinensis individual represents each hyperparameter combination. The hyperparameter combination corresponding to the position vector of each individual Cordyceps sinensis is imported into the architecture of a graph convolutional neural network to generate a graph convolutional neural network model corresponding to each individual Cordyceps sinensis. The graph convolutional neural network model corresponding to each individual Cordyceps sinensis is trained using the training set to obtain a trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis. The training graph convolutional neural network model corresponding to each individual Cordyceps sinensis was validated using a validation set. The accuracy of the training graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set was obtained, and the accuracy of the training graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set was selected as the fitness value of each individual Cordyceps sinensis. The position vector of each individual Cordyceps sinensis is updated using a Cordyceps sinensis optimization algorithm. The current iteration number is obtained during the update. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. The electrical data of the power distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the power distribution network at the current moment.

[0053] The processor 20 may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors, field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0054] In some embodiments, the memory 21 may be an internal storage unit of the electronic device, such as a hard disk or memory of the electronic device.

[0055] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for fault location in a distribution network based on a graph convolutional neural network, characterized in that, The power distribution network fault location method, applied to electronic equipment, includes: Obtain electrical data of the distribution network at a preset time and fault location results of the distribution network at a preset time, and combine the electrical data of the distribution network at a preset time and fault location results of the distribution network at a preset time into a fault sample; Different fault samples are divided into training and validation sets to obtain the control parameters of the Cordyceps sinensis optimization algorithm. The total number of individuals and the maximum number of iterations are obtained from the control parameters. Based on the total number of individuals, multiple Cordyceps sinensis individuals are generated. Each Cordyceps sinensis individual is an independent search unit, and the position vector of each Cordyceps sinensis individual represents each hyperparameter combination. The hyperparameter combination corresponding to the position vector of each individual Cordyceps sinensis is imported into the architecture of a graph convolutional neural network to generate a graph convolutional neural network model corresponding to each individual Cordyceps sinensis. The graph convolutional neural network model corresponding to each individual Cordyceps sinensis is trained using the training set to obtain a trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis. The training graph convolutional neural network model corresponding to each individual Cordyceps sinensis was validated using a validation set. The accuracy of the training graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set was obtained, and the accuracy of the training graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set was selected as the fitness value of each individual Cordyceps sinensis. The position vector of each individual Cordyceps sinensis is updated using a Cordyceps sinensis optimization algorithm. The current iteration number is obtained during the update. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. The electrical data of the power distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the power distribution network at the current moment.

2. The method for locating faults in a power distribution network according to claim 1, characterized in that, The method employs a Cordyceps sinensis optimization algorithm to update the position vector of each individual Cordyceps sinensis, obtaining the current iteration number during the update process. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. The electrical data of the power distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the power distribution network at the current moment, including: The first or second operator in the Cordyceps sinensis optimization algorithm is used to update the position vector of each individual Cordyceps sinensis. The current iteration number during the update is obtained. When the current iteration number is the maximum iteration number, the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value is selected as the optimal hyperparameter combination. The optimal combination of hyperparameters is imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. The system receives data uploaded by the acquisition equipment, obtains the electrical data of the power distribution network at the current moment from the uploaded data, and inputs the electrical data of the power distribution network at the current moment into the fault location model. The fault location model then generates the fault location result of the power distribution network at the current moment.

3. The method for locating faults in a power distribution network according to claim 1, characterized in that, The method involves updating the position vector of each individual Cordyceps sinensis using a Cordyceps sinensis optimization algorithm, obtaining the current iteration number during the update, and selecting the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value as the optimal hyperparameter combination when the current iteration number is the maximum iteration number. This optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the power distribution network. After inputting the electrical data of the power distribution network at the current moment into the fault location model, and the fault location model generating the fault location result for the power distribution network at the current moment, the method includes: From the fault location results, obtain the line identifier of the faulty line in the distribution network, send the line identifier of the faulty line to the monitoring system of the distribution network, and receive the response data returned by the monitoring system based on the line identifier of the faulty line. Obtain the ground voltage at the first end of the faulty line from the response data. Based on the ground voltage at the first end of the faulty line, the path length from the fault point to the first end of the faulty line is generated according to the preset length model.

4. The method for locating faults in a power distribution network according to claim 3, characterized in that, The process of obtaining the line identifier of the faulty line in the distribution network from the fault location results, sending the line identifier of the faulty line to the monitoring system of the distribution network, and receiving response data returned by the monitoring system based on the line identifier of the faulty line includes: The fault location results are analyzed to obtain the line identifier of the faulty line in the distribution network. The line identifier of the faulty line is sent to the monitoring system of the distribution network, and the response data returned by the monitoring system based on the line identifier of the faulty line is received.

5. The method for locating faults in a power distribution network according to claim 3, characterized in that, After obtaining the ground voltage at the starting point of the faulty line and the ground voltage at the starting point of the faulty line from the response data, and generating the path length from the fault point to the starting point of the faulty line based on the ground voltage at the starting point of the faulty line, the ground voltage at the starting point of the faulty line, and a preset length model, the distribution network fault location method includes: Create a display window to show the line identifier of the faulty line and the path length from the fault point to the beginning and end points of the faulty line.

6. The method for locating faults in a power distribution network according to claim 3, characterized in that, The length model is defined as follows: ； This indicates the path length from the fault point to the beginning of the faulty line. This indicates the voltage to ground at the starting point of the faulty line; This indicates the voltage to ground at the end of the faulty line. This indicates the path length from the first measurement point to the beginning of the faulty line; This indicates the path length from the second measurement point to the end point of the faulty line; Indicates the total path length of the faulty line; This represents the voltage to ground at the first measurement point; This represents the voltage to ground at the second measurement point.

7. The method for locating faults in a power distribution network according to claim 2, characterized in that, The first operator is defined as follows: ； ； This represents the first search coefficient. The larger the first search coefficient, the larger the update magnitude of the first operator; the smaller the first search coefficient, the smaller the update magnitude of the first operator. This represents pi, with a value of 3.

14. Represents the cosine function; Indicates the first The updated position vector of each individual Cordyceps sinensis. It was obtained by updating the first operator in the Cordyceps sinensis optimization algorithm; The serial number representing an individual Cordyceps sinensis; Indicates the first The position vector of each individual Cordyceps sinensis before updating; Indicates the first The position vector of the previous Cordyceps sinensis individual before the update; This represents the position vector of the Cordyceps sinensis individual with the highest fitness value. Indicates the column number; This indicates the generation of a string with 1 row and 1 column. A random vector.

8. The method for locating faults in a power distribution network according to claim 2, characterized in that, The second operator is defined as follows: ； ； This represents the second search coefficient. The larger the second search coefficient, the larger the search range of the second operator; the smaller the second search coefficient, the smaller the search range of the second operator. Indicates the current iteration number; Indicates the maximum number of iterations; Indicates a A random number uniformly distributed within a range; It is a random integer, which can be either 1 or 2; Indicates the first The updated position vector of each individual Cordyceps sinensis. It was obtained by updating the second operator in the Cordyceps sinensis optimization algorithm; The serial number representing an individual Cordyceps sinensis; Indicates the first The position vector of each individual Cordyceps sinensis before updating; Indicates the first The position vector of the previous Cordyceps sinensis individual before the update; This represents the position vector of the Cordyceps sinensis individual with the highest fitness value. Indicates the column number; This indicates the generation of a string with 1 row and 1 column. A random vector.

9. A power distribution network fault location device based on graph convolutional neural networks, characterized in that, Applied to electronic devices, including: The first acquisition module is used to acquire the electrical data of the distribution network at a preset time and the fault location result of the distribution network at a preset time, and to combine the electrical data of the distribution network at a preset time and the fault location result of the distribution network at a preset time into a fault sample. The second acquisition module is used to divide different fault samples into training sets and validation sets, acquire the control parameters of the Cordyceps sinensis optimization algorithm, obtain the total number of individuals and the maximum number of iterations from the control parameters, generate multiple Cordyceps sinensis individuals based on the total number of individuals, each Cordyceps sinensis individual is an independent search unit, and the position vector of each Cordyceps sinensis individual represents each hyperparameter combination. The training module is used to combine the hyperparameters corresponding to the position vector of each individual Cordyceps sinensis, import them into the architecture of a graph convolutional neural network, generate a graph convolutional neural network model corresponding to each individual Cordyceps sinensis, and train the graph convolutional neural network model corresponding to each individual Cordyceps sinensis using the training set to obtain a trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis. The validation module is used to validate the trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis using a validation set, obtain the accuracy of the trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set, and select the accuracy of the trained graph convolutional neural network model corresponding to each individual Cordyceps sinensis on the validation set as the fitness value of each individual Cordyceps sinensis. The generation module is used to update the position vector of each individual Cordyceps sinensis using the Cordyceps sinensis optimization algorithm, obtain the current iteration number during the update, and when the current iteration number is the maximum iteration number, select the hyperparameter combination represented by the position vector of the individual Cordyceps sinensis with the highest fitness value as the optimal hyperparameter combination. The optimal hyperparameter combination is then imported into the architecture of a graph convolutional neural network to generate a fault location model for the distribution network. The electrical data of the distribution network at the current moment is input into the fault location model, and the fault location model generates the fault location result of the distribution network at the current moment.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the power distribution network fault location method as described in any one of claims 1 to 7.

Images