Low-current grounding system single-phase grounding fault line selection method and system

Abstract

The invention belongs to the field of single-phase earth fault line selection, and provides a small-current grounding system single-phase earth fault line selection method and system. The method comprises the following steps: acquiring fault current information of a small current grounding system; extracting fault features in the fault current information; the fault characteristics at least comprise a zero-sequence current amplitude, a zero-sequence current phase angle, a fifth harmonic current amplitude, a fifth harmonic current phase angle and a transient zero-sequence current amplitude; the extracted fault features are parallelly and correspondingly input to a first neural network model, and single-phase earth fault line selection initial results corresponding to the fault features are obtained; the number of parallel fault classifiers contained in the first neural network model is the same as the number of fault features; and inputting all the initial line selection results into the second neural network model to obtain a final single-phase earth fault line selection result.

Description

Technical Field

[0001] The invention belongs to the field of single-phase grounding fault line selection, and in particular relates to a single-phase grounding fault line selection method and system for a small current grounding system. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] In the medium voltage distribution network of 6kV-35kV, the low current grounding method of ungrounded neutral point and grounded neutral point through arc suppression coil is mostly used. In the low current grounding system, single-phase grounding fault is a common power fault, which occurs frequently and has a wide impact range, posing a serious threat to the safe and stable operation of the power system. Unlike the neutral point direct grounding system, the low current grounding system can continue to operate with power for 1-2 hours when a single-phase grounding fault occurs. However, it is now required that the handling principle of the small current grounding system when a single-phase grounding fault occurs is changed from "operating with fault" to "quickly isolating nearby". This change in the handling principle has put forward higher requirements on the accuracy and reliability of the single-phase grounding fault line selection of the small current system.

[0004] Because the line selection problem of small current system grounding fault has problems such as unclear fault characteristics, influence of three-phase unbalanced current, compensation of arc suppression coil, intermittent arc in fault grounding, and complex distribution network fault conditions, the traditional small current grounding fault line selection method using single fault characteristics has certain limitations. Summary of the invention

[0005] In order to solve the problem that the fault line selection method using a single fault feature is difficult to complete the correct line selection under complex conditions, the present invention provides a single-phase grounding fault line selection method and system for a small current grounding system, which can fully explore the value of fault data of the small current grounding system, obtain simulation data by building a fault simulation model, and apply the simulation data to the training of multiple groups of neural network models to complete the fault line selection.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A first aspect of the present invention provides a single-phase grounding fault line selection method for a small current grounding system.

[0008] In one or more embodiments, a method for selecting a single-phase grounding fault line in a small current grounding system is provided, comprising:

[0009] Obtain fault current information of small current grounding system;

[0010] Extracting fault features from the fault current information; the fault features at least include zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, fifth harmonic current phase angle and transient zero-sequence current amplitude;

[0011] The extracted fault features are inputted into the first neural network model in parallel to obtain the initial single-phase grounding fault line selection results corresponding to the fault features; the first neural network model includes a plurality of parallel fault classifiers, and the number of the fault classifiers is the same as the number of the fault features;

[0012] All initial line selection results are input into the second neural network model to obtain the final single-phase grounding fault line selection result.

[0013] A second aspect of the present invention provides a single-phase grounding fault line selection system for a small current grounding system.

[0014] In one or more embodiments, a single-phase grounding fault line selection system for a small current grounding system includes:

[0015] A fault current acquisition module, which is used to obtain fault current information of a small current grounding system;

[0016] A fault feature extraction module, which is used to extract fault features from fault current information; the fault features at least include zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, fifth harmonic current phase angle and transient zero-sequence current amplitude;

[0017] The line selection initial result prediction module is used to extract each fault feature and input it into the first neural network model in parallel to obtain the initial line selection result of the single-phase grounding fault corresponding to each fault feature; the first neural network model includes a plurality of parallel fault classifiers, and the number of fault classifiers is the same as the number of fault features;

[0018] The line selection result finalization module is used to input all the initial line selection results into the second neural network model to obtain the final single-phase grounding fault line selection result.

[0019] A third aspect of the present invention provides a computer-readable storage medium.

[0020] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in the single-phase grounding fault line selection method for a small current grounding system as described above.

[0021] A fourth aspect of the present invention provides a computer program product.

[0022] A computer program product includes a computer program / instruction, which, when executed by a processor, implements the steps in the above-mentioned single-phase grounding fault line selection method for a small current grounding system.

[0023] A fifth aspect of the present invention provides an electronic device.

[0024] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the single-phase grounding fault line selection method for a small current grounding system as described above are implemented.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] The present invention constructs a first neural network model and a second neural network model, and based on each fault feature and the first neural network model, obtains the initial results of single-phase grounding fault line selection corresponding to each fault feature; then completes the fault line selection under the comprehensive model based on the second neural network model. The present invention fully explores the value of small current grounding system fault data, solves the problem that the fault line selection method with a single fault feature is difficult to complete the correct line selection under complex conditions, and can give full play to the fitting advantages of neural networks and the complementarity between different fault line selection methods, thereby improving the accuracy and reliability of fault line selection. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0028] Figure 1 It is a flow chart of a method for selecting a single-phase grounding fault line in a small current grounding system according to an embodiment of the present invention;

[0029] Figure 2 This is an example diagram of a single-phase grounding fault line selection process in a small current grounding system according to an embodiment of the present invention;

[0030] Figure 3 1 is a schematic diagram of the structure of a single-phase grounding fault line selection system for a small current grounding system according to an embodiment of the present invention;

[0031] Figure 4 is a schematic diagram of an electronic device according to an embodiment of the present invention;

[0032] Figure 5 is a schematic diagram of a GA-BP neural network model according to an embodiment of the present invention;

[0033] Figure 6The present invention uses MATLAB / Simulink to build a simulation model of a neutral point grounding system through an arc suppression coil;

[0034] Figure 7 It is a simulation model of a neutral point ungrounded system built by the present invention using MATLAB / Simulink;

[0035] Figure 8 It is an FFT data measurement module built by the present invention using MATLAB / Simulink. DETAILED DESCRIPTION

[0036] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0037] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0038] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0039] Figure 1 is a flow chart of a method for selecting a single-phase ground fault line in a small current grounding system according to an embodiment of the present invention, combined with Figure 1 and Figure 2 The single-phase grounding fault line selection method of a small current grounding system in this embodiment may include:

[0040] S101, obtaining fault current information of a low-current grounding system.

[0041] In the specific implementation process, a current collection device such as a current transformer may be used to obtain the fault current information of the small current grounding system, which will not be described in detail here.

[0042] S102, extracting fault features from the fault current information; the fault features at least include zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, fifth harmonic current phase angle and transient zero-sequence current amplitude.

[0043] The number of fault features here can be set specifically according to actual conditions.

[0044] S103, extracting each fault feature and inputting them into the first neural network model in parallel to obtain the initial single-phase grounding fault line selection results corresponding to each fault feature; the first neural network model includes a number of parallel fault classifiers, and the number of fault classifiers is the same as the number of fault features.

[0045] The following is a detailed explanation of the five characteristics of zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, fifth harmonic current phase angle and transient zero-sequence current amplitude.

[0046] The number of fault classifiers in the first neural network model corresponds to five. In this embodiment, BP neural network models are all used.

[0047] It can be understood here that, in other embodiments, the type of fault classifier can be selected from existing classifiers according to actual conditions, such as deep neural networks such as CNN or RNN, which will not be described in detail here.

[0048] BP neural network is a common multi-layer feedforward neural network. It adjusts weights through the back-propagation algorithm and has powerful nonlinear mapping capabilities. Compared with deep neural networks such as CNN and RNN, BP neural network has relatively low demand for data volume. For some problems with less data but more obvious features, BP neural network may perform better, while deep neural network may fall into the dilemma of overfitting. For a BP neural network model, its input data is the fault feature quantity of different lines under this feature. For a certain input fault feature quantity, the difference between the fault line and the non-fault line is generally obvious. The construction of the first neural network model can be efficiently completed by applying BP neural network instead of other deep neural networks. Its output data is the line selection result of the BP neural network model based on this feature. Therefore, the output result of the first neural network model is the fault line selection result based on five different BP neural network models.

[0049] In MATLAB / Simulink environment, build Figure 6 , Figure 7 The neutral point is grounded through the arc suppression coil system model and the neutral point is not grounded system model shown. The lengths of the three feeders L1, L2, and L3 are set to 10km, 8km, and 6km respectively. The arc suppression coil inductance is set to 0.165H. The positive sequence parameters of the line are: R1 = 0.1273Ω / km, X1 = 0.9377mh / km, C1 = 0.01274μF / km; the zero sequence parameters are: R0 = 0.38643Ω / km, X0 = 4.1264mh / km, C0 = 0.00775μF / km.

[0050] Different fault conditions are set: the transition resistance is set to 0 ohm, 10 ohm, and 100 ohm; the fault location is 10% of the line, 50% of the line, and 90% of the line; the fault time is 0.04s, 0.045s, 0.05s, and 0.055s; the fault lines are L1, L2, and L3. The fault data under different fault conditions are obtained by fast Fourier transform. The fault data includes zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic amplitude, fifth harmonic phase angle, and transient current amplitude under different fault conditions. Some data are shown in Table 1.

[0051] Table 1

[0052]

[0053]

[0054] Each BP neural network model based on a single fault feature has its input as the fault feature value of the fault feature in different lines. This solution sets 216 different fault conditions through the simulink simulation model, measures the zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, fifth harmonic current phase angle and transient zero-sequence current amplitude of line L1, line L2 and line L3 under each fault condition and uses them as the data set of the BP neural network model in the first neural network model. Taking the BP neural network model trained based on the fifth harmonic current phase angle as the feature in the first neural network model as an example, the fault condition is set as the grounding resistance of 0.01 ohms, the fault time of 0.045s, and the fault position of 10% of line L2 through the simulink simulation model. The fifth harmonic current phase angles of line L1, line L2, and line L3 are measured by Fourier transform to be 90.57 degrees, -172.5 degrees, and 89.72 degrees, respectively. Therefore, a set of training data for the BP neural network model trained based on the fifth harmonic current phase angle under this fault condition can be obtained. The input data of this set of training data is 90.57, -172.5, and 89.72, and the output data is fault line 2. This solution sets 216 fault conditions through the simulink simulation model, and obtains a total of 216 sets of fifth harmonic current phase angle characteristic data consistent with the above format and the corresponding fault lines to form a fifth harmonic current phase angle fault characteristic data set. The data set is divided into a training set, a test set, and a verification set according to the ratio of 70%, 15%, and 15%, completing the training and construction of the BP neural network model based on the fifth harmonic current phase angle in the first neural network model. Similarly, the training and construction of the other four BP neural network models in the first neural network model can be completed by obtaining the zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, and transient zero-sequence current amplitude fault characteristic data sets through the simulink simulation model.

[0055] S104, input all initial line selection results into the second neural network model to obtain the final single-phase grounding fault line selection result.

[0056] The second neural network model is a GA-BP neural network model. The GA-BP neural network is an optimization model that combines the genetic algorithm (GA) with the BP neural network. It optimizes the weights and biases of the BP neural network through the genetic algorithm, thereby improving the performance of the network. Its input data is the fault line selection results of the neural network model based on five different fault characteristics output by the first neural network model. Compared with the input data with obvious differences in the first neural network model, the input data of the second neural network model is more complex, and it is difficult to fit it well through the BP neural network. The GA-BP neural network combines the advantages of the genetic algorithm and the BP neural network. Through global search, it can effectively avoid overfitting and has stronger adaptability. Its output data is the fault line selection result given by the single-phase grounding fault line selection model of the small current grounding system based on multiple groups of neural network models.

[0057] Five different fault data under different fault conditions are used to train five BP neural network line selection models in the first neural network model. The input layer data of each neural network model is a single fault feature of three feeders under one fault condition, and the output layer data is the line selection result based on this feature. The output results of the first neural network model are recorded and used to train the second GA-BP neural network model. The input layer data of the second neural network model is the line selection results given by the five neural network models under one fault condition, and the output layer data is the real fault line.

[0058] The second neural network model realizes the fusion of different fault features by integrating the output results of the first neural network model. Taking the case where the fault condition is an ungrounded system, the grounding resistance is 100 ohms, the fault time is 0.05s, the fault position is 90%, and the fault line is L3 as an example, the fault line selection results given by the BP neural network model based on zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, fifth harmonic current phase angle and transient zero-sequence current amplitude in the first neural network model are 2.9, 2, 2, 1.8, and 1.6 respectively. The actual fault line is line 2, so a set of training data is obtained, and its input data is 2.9, 2, 2, 1.8, and 1.6, and the output data is line 2. Since a total of 216 fault conditions are set, the first neural network model has a total of 216 groups of fault line selection results and 216 corresponding actual fault lines. The 216 groups of input and output data constitute the GA-BP neural network model data set. The data set is divided into a training set and a test set according to the ratio of 90% and 10% to complete the training and construction of the GA-BP neural network model.

[0059] It utilizes the strong parallelism of the GA-BP neural network and uses the line selection results of the BP neural network model based on different fault characteristics under the same fault as the input of the second neural network model. It optimizes the initial weights and bias of the neural network model through the genetic algorithm to achieve the fusion of different fault characteristics.

[0060] Taking the case where the fault condition is metallic grounding of an ungrounded system, the fault time is 0.04s, the fault location is 10%, and the fault line is L1 as an example, the fault line selection results given by the BP neural network model based on zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, fifth harmonic current phase angle and transient zero-sequence current amplitude in the first neural network model are 1.5, 2.9, 1.9, 1.6, and 3.1 respectively. The actual fault line is line 3. It can be seen that the BP neural network model based on a single fault feature in the first neural network model cannot achieve accurate line selection in this case. The fault line selection results (1.5, 2.9, 1.9, 1.6, and 3.1) of the first neural network model are input into the trained second group of GA-BP neural network models. The fault line selection result given by the GA-BP neural network model is line 3, which is consistent with the actual fault line. It can be seen that the second neural network model uses the output data of the first neural network model as input data to achieve the fusion of different fault feature data and finally obtain the correct line selection result.

[0061] like Figure 5 As shown in the figure, the optimization of BP neural network by genetic algorithm is realized by writing code. The implementation steps of genetic algorithm are as follows:

[0062] (1) Encoding: Consider a neural network with input nodes i, hidden nodes j, and output nodes k, where the weight matrix from the input layer to the hidden layer is W, the threshold matrix of the hidden layer is r, the weight matrix from the hidden layer to the output layer is V, and the threshold matrix of the output layer is h. The number of input layer nodes of the second neural network model is i=5, the number of hidden layer nodes is j=5, and the number of output layer nodes is k=1. According to the formula

[0063] L=i×j+j×k+j+k (1)

[0064] The calculated individual coding length L of the genetic algorithm is 36. Where i is the number of input layer nodes, j is the number of hidden layer nodes, and k is the number of output layer nodes.

[0065] (2) Fitness function: The fitness function is defined as

[0066]

[0067] In formula (2) represents the i-th predicted fault line selection result obtained by the second neural network model for the k-th training sample, is the corresponding actual fault line, NO is the number of training set samples obtained by the first neural network model, Ne is the number of output nodes, and t is the number of iterations. Iteration adjustment factor (1+e -t ) is larger in the early stage of iteration, which can weaken the fitness value of the known optimal solution and increase the exploration of other potential optimal solutions. -t ) is close to 1 in the later stage of iteration, has no effect on the optimal solution, and can better determine the optimal solution.

[0068] (3) Population initialization: The simplest population initialization method is to randomly generate individual chromosome information, where each individual (chromosome) represents a possible weight and threshold configuration of the second neural network model. Each chromosome can be randomly generated to ensure that the individuals in the population have a certain degree of randomness in the solution space. The following process of selection, crossover, mutation, etc. is used to achieve the optimization of the weight and threshold configuration of the second neural network model.

[0069] (4) Selection: The selection operation refers to selecting individuals suitable as representatives of parents to evolve the next generation based on the fitness value of the individuals. The selection operation is an important link that affects the performance of genetic algorithms. Roulette selection is one of the most common selection methods. It selects based on the ratio of individual fitness values. The higher the fitness of an individual, the greater the probability of being selected. The selection can be made by calculating the probability of each individual being selected and then using random numbers.

[0070] (5) Crossover: Assume that there are two chromosomes a and b of length l. Use the real number crossover method to cross these two chromosomes.

[0071] a' k =(1-r)a k +rb k (3)

[0072] b' k =rb k +(1-r)b k (4)

[0073] In formula (3) and (4), r is a random number between [0,1], and k = 1,2…l. This method helps to retain the excellent genes of excellent individuals and generate more genetic diversity for future generations. The r value is usually used to control the hybridization probability, that is, to determine which individuals will undergo hybridization. An appropriate r value can increase the diversity of individuals, enhance the algorithm's ability to search the solution space, and help the algorithm find the best solution during the iteration process.

[0074] (6) Mutation: When the mutation operator mutates the kth gene of an individual a with a certain probability, the mutation operation is

[0075]

[0076] In formula (5), a max and a min are the upper and lower bounds of the gene value, G is the current iteration number, and G max is the maximum number of iterations, and r1 and r2 are random numbers between [0,1]. The advantage of this method is that it can limit the variation of gene values ​​by controlling the probability and range of mutation, ensuring that the variation of gene values ​​is evenly distributed. The purpose of the mutation operation is to introduce new gene combinations to help the population escape from the local optimal solution and avoid falling into local search.

[0077] (7) Population update: The population update process is an important evolutionary process in genetic algorithms. By continuously iterating and updating the population, the individuals in the population gradually tend to a more optimized state. In the population update process, it is necessary to pay attention to the reasonable setting of the selection, crossover and mutation parameters, as well as the setting of the termination conditions, to ensure that the algorithm can effectively search for the global optimal solution or a better solution.

[0078] (8) Termination condition: Set the maximum number of iterations to 1000 as the termination condition. When the algorithm reaches the specified number of iterations, it stops searching and returns the result.

[0079] In order to verify the effectiveness and applicability of the method proposed in this embodiment for the fault line selection problem of small current grounding system, a MATLAB / Simulink software is used to build the following Figure 6 , Figure 7The fault simulation model of the small current grounding system (neutral point grounding system through arc suppression coil, neutral point ungrounded system) shown in the figure, in this model, there are 3 10.5kV voltage level distribution network feeder lines. In order to match the actual situation, the lengths of the three distribution network feeder lines should not be too long. The lengths of feeder lines L1, L2, and L3 are set to 10km, 8km, and 6km respectively. The positive sequence parameters of the line are: R1 = 0.1273Ω / km, X1 = 0.9377mh / km, C1 = 0.01274μF / km; the zero sequence parameters are: R0 = 0.38643Ω / km, X0 = 4.1264mh / km, C0 = 0.00775μF / km. In order to simulate the complex fault conditions in the actual distribution network, this embodiment sets different grounding resistances (metallic grounding, grounding resistance of 10 ohms, grounding resistance of 100 ohms), different fault occurrence times (0.04s, 0.045s, 0.05s, 0.055s), different fault lines (L1, L2, L3) and different fault occurrence locations on the line (line 10%, line 90%) for the fault simulation model. Figure 8 The FFT module shown measures the zero-sequence current amplitude and phase angle, fifth harmonic current amplitude and phase angle of each line under different fault conditions and reads transient current amplitude and other information through an oscilloscope.

[0080] Several groups of fault test data from some cases are selected, and this group of data is input into the first neural network model. The five fault line selection results are recorded, and the line selection results of the first neural network model are input into the second neural network model to obtain the final fault line selection results based on multiple groups of neural network models. The relevant data are shown in Table 2. The test results are consistent with the actual fault line.

[0081] Table 2

[0082]

[0083] Figure 3 Schematic diagram of the structure of a single-phase ground fault line selection system for a small current grounding system according to an embodiment of the present invention. Figure 1 The line selection method for single-phase ground fault in a small current grounding system corresponds to the following method: Figure 3 As shown, the single-phase grounding fault line selection system of a small current grounding system in this embodiment may include:

[0084] A fault current acquisition module 301 is used to acquire fault current information of a small current grounding system;

[0085] A fault feature extraction module 302 is used to extract fault features from fault current information; the fault features at least include zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, fifth harmonic current phase angle and transient zero-sequence current amplitude;

[0086] The line selection initial result prediction module 303 is used to input the extracted fault features into the first neural network model in parallel to obtain the initial line selection results of the single-phase grounding fault corresponding to the fault features; the first neural network model includes a plurality of parallel fault classifiers, and the number of the fault classifiers is the same as the number of the fault features;

[0087] The line selection result finalization module 304 is used to input all the initial line selection results into the second neural network model to obtain the final single-phase grounding fault line selection result.

[0088] Wherein, the fault classifiers are all BP neural network models; or / and the second neural network model is a GA-BP neural network model.

[0089] It should be noted here that Figure 3 The modules in the single-phase ground fault line selection system of the small current grounding system are Figure 1 The various steps in the single-phase grounding fault line selection method for a small current grounding system correspond one to one, and the specific implementation process is the same, so it will not be repeated here.

[0090] Reference Figure 4 , a schematic diagram of an electronic device is given. It should be noted that, Figure 4 The electronic device 400 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0091] like Figure 4 As shown, electronic device 400 includes a central processing unit (CPU) 401, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 402 or a program loaded from a storage portion 408 into a random access memory (RAM) 403. In RAM 403, various programs and data required for system operation are also stored. Central processing unit 401, ROM 402 and RAM 403 are connected to each other via bus 404. Input / output (I / O) interface 405 is also connected to bus 404.

[0092] The following components are connected to the I / O interface 405: an input section 406 including a keyboard, a mouse, etc.; an output section 407 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a local area network (LAN) card, a modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the I / O interface 405 as needed. A removable medium 411, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 410 as needed so that a computer program read therefrom is installed into the storage section 408 as needed.

[0093] When the central processing unit 401 in the electronic device of this embodiment executes the program, the following is achieved: Figure 1 The steps in the single-phase grounding fault line selection method for a small current grounding system are shown.

[0094] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a computer readable medium, the computer program including a computer program for executing Figure 1 In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 409, and / or installed from the removable medium 411. When the computer program is executed by the central processing unit 401, various functions defined in the apparatus of the present application are executed.

[0095] in, Figure 1 The computer program instructions corresponding to the method shown may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0096] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0097] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for selecting a single-phase grounding fault line in a small current grounding system, characterized in that: include: Obtain fault current information of small current grounding system; Extract fault characteristics from fault current information; The fault characteristics at least include zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, fifth harmonic current phase angle and transient zero-sequence current amplitude; The extracted fault features are inputted into the first neural network model in parallel to obtain the initial results of single-phase grounding fault line selection corresponding to the fault features; The first neural network model includes a plurality of parallel fault classifiers, and the number of the fault classifiers is the same as the number of the fault features; All initial line selection results are input into the second neural network model to obtain the final single-phase grounding fault line selection result.

2. The method for selecting a single-phase ground fault line in a small current grounding system according to claim 1, characterized in that: The fault classifiers are all BP neural network models.

3. The method for selecting a single-phase ground fault line in a small current grounding system according to claim 1, characterized in that: The second neural network model is a GA-BP neural network model.

4. The method for selecting a single-phase ground fault line in a small current grounding system according to claim 3, characterized in that: The GA-BP neural network model is used to optimize the initial weights and biases of the BP neural network model for each input single-phase grounding fault line selection initial result through a genetic algorithm, so as to determine each weight value of the single-phase grounding fault line selection initial result and determine the final single-phase grounding fault line selection result.

5. The method for selecting a single-phase ground fault line in a small current grounding system according to claim 3, characterized in that: In the GA-BP neural network model, the individual coding length L of the GA algorithm is: L = i × j + j × k + j + k; where i is the number of input layer nodes, j is the number of hidden layer nodes, and k is the number of output layer nodes; In the GA-BP neural network model, the fitness function of the GA algorithm is defined as: In the formula represents the i-th predicted fault line selection result obtained by the k-th training sample through the GA-BP neural network model, is the corresponding actual fault line, NO is the number of training set samples obtained by the first neural network model, Ne is the number of output nodes, and t is the number of iterations; (1+e -t ) is the iteration adjustment factor.

6. A single-phase grounding fault line selection system for a small current grounding system, characterized in that: include: A fault current acquisition module, which is used to obtain fault current information of a small current grounding system; A fault feature extraction module is used to extract fault features from fault current information; The fault characteristics at least include zero-sequence current amplitude, zero-sequence current phase angle, fifth harmonic current amplitude, fifth harmonic current phase angle and transient zero-sequence current amplitude; The line selection initial result prediction module is used to extract each fault feature and input it into the first neural network model in parallel to obtain the initial line selection result of the single-phase grounding fault corresponding to each fault feature; the first neural network model includes a plurality of parallel fault classifiers, and the number of fault classifiers is the same as the number of fault features; The line selection result finalization module is used to input all the initial line selection results into the second neural network model to obtain the final single-phase grounding fault line selection result.

7. The single-phase grounding fault line selection system for a small current grounding system according to claim 6, characterized in that: The fault classifiers are all BP neural network models; or / and the second neural network model is a GA-BP neural network model.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the single-phase grounding fault line selection method for a small current grounding system as described in any one of claims 1 to 5 are implemented.

9. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the steps in the method for selecting a single-phase grounding fault line in a small current grounding system as described in any one of claims 1 to 5 are implemented.

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 program, the steps in the method for selecting a single-phase grounding fault line in a small current grounding system according to any one of claims 1 to 5 are implemented.

Images