Low-voltage AC / DC hybrid power distribution network electric leakage fault identification method and related device

Abstract

The invention discloses a low-voltage AC / DC hybrid power distribution network electric leakage identification method and a related device, and belongs to the technical field of fault identification, and the method comprises the following steps: collecting leakage current data of different states of a low-voltage AC / DC hybrid power distribution network; the leakage current data is decomposed by adopting optimized variational mode decomposition to obtain mode components of different frequency bands, and the optimized variational mode decomposition is obtained by optimizing parameters of the variational mode decomposition through a mucus algorithm; and calculating a sample entropy of each modal component to construct an electric leakage fault feature set, inputting the electric leakage fault feature set into a convolutional neural network model to establish an electric leakage fault recognition model, and completing electric leakage fault recognition through the electric leakage fault recognition model. The residual current protection device can solve the problem that an existing residual current protection device is difficult to accurately identify the electric leakage type.

Description

Technical Field

[0001] The present invention belongs to the technical field of fault identification, and in particular relates to a method for identifying leakage in a low-voltage AC / DC hybrid distribution network and a related device. Background Art

[0002] As an important device to prevent electrical disasters (electric shock accidents and fire accidents), residual current protection devices have been widely used in low-voltage distribution networks. In recent years, with the continuous growth of photovoltaic installed capacity, large-scale distributed photovoltaics have been connected to the distribution network. In the low-voltage AC / DC hybrid distribution network connected to distributed photovoltaics, due to the complex actual operating environment, residual current detection is easily affected by factors such as photovoltaic system leakage current and normal line leakage current. When a leakage fault occurs in the system, the existing residual current protection technology cannot distinguish between the leakage current of electric shock fault, the normal line leakage current and the leakage current of the photovoltaic system, causing the residual current protection device to malfunction, making the system unable to operate safely and reliably, and even threatening personal safety and causing electrical fires. Therefore, it is necessary to establish an efficient and reliable leakage fault identification model, accurately identify the type of leakage fault, ensure the correct operation of the residual current protection device, provide technical support for the construction of photovoltaic grid-connected projects and the safe operation of the system, and realize the personal safety, equipment safety and power grid safety of low-voltage AC / DC hybrid distribution networks with large-scale distributed photovoltaic access. Summary of the invention

[0003] The purpose of the present invention is to provide a method and related device for identifying leakage faults in a low-voltage AC / DC hybrid distribution network, so as to solve the problem that it is difficult for existing residual current protection devices to accurately identify the leakage type.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] In a first aspect, a method for identifying leakage faults in a low voltage AC / DC hybrid distribution network comprises the following steps:

[0006] S1: Collect leakage current data of different states of low-voltage AC / DC hybrid distribution network;

[0007] S2: Decomposing the leakage current data by using optimized variational modal decomposition to obtain modal components of different frequency bands;

[0008] S3: Calculate the sample entropy of each of the modal components to construct a leakage fault feature set, input the leakage fault feature set into a convolutional neural network model to establish a leakage fault identification model, and complete leakage fault identification through the leakage current data through the leakage fault identification model.

[0009] In some embodiments, the parameters of the variational modal decomposition include a penalty factor and the number of modal components.

[0010] In some embodiments, the convolutional neural network model includes an input layer, a first convolutional layer, a second convolutional layer, a first pooling layer, a second pooling layer, a fully connected layer, and an output layer;

[0011] The input layer is connected to the first convolution layer, the first convolution layer is connected to the first pooling layer, the first pooling layer is connected to the second convolution layer, the second convolution layer is connected to the second pooling layer, the second pooling layer is connected to the fully connected layer, and the fully connected layer is connected to the output layer.

[0012] In some embodiments, the size of the convolution kernel of the convolutional neural network model is 2×1, and the size of the first pooling layer and the second pooling layer is 2×1.

[0013] In some embodiments, the optimized variational modal decomposition is obtained by optimizing the parameters of the variational modal decomposition using a slime mold algorithm, specifically comprising the following steps:

[0014] Set the slime mold population size and maximum number of iterations;

[0015] Initialize the individual positions of slime molds by the size of the slime mold population;

[0016] After calculating the fitness according to the current slime mold individual position and sorting the fitness, the optimal fitness and its corresponding optimal slime mold individual position are updated;

[0017] After updating the search range and search direction of the slime mold population food according to the current distribution of the slime mold individual positions and the fitness, updating the slime mold individual positions and determining whether the maximum number of iterations has been reached, outputting the optimal slime mold individual positions if the maximum number of iterations has been reached, and repeating the steps of updating the optimal fitness and its corresponding optimal slime mold individual positions after updating the slime mold individual positions, until the maximum number of iterations has been reached;

[0018] The individual positions of the slime mold are the parameters of the variational mode decomposition to be optimized.

[0019] In some embodiments, the different states of the leakage current data include: a normal line leakage state, a photovoltaic system leakage state, and an electric shock fault state.

[0020] In a second aspect, a low voltage AC / DC hybrid distribution network leakage fault identification system includes:

[0021] A data acquisition module is used to collect leakage current data of different states of a low-voltage AC / DC hybrid distribution network;

[0022] An optimized variational modal decomposition module is used to decompose the leakage current data using the optimized variational modal decomposition to obtain modal components of different frequency bands;

[0023] The leakage fault identification module is used to calculate the sample entropy of each modal component to construct a leakage fault feature set, input the leakage fault feature set into a convolutional neural network model to establish a leakage fault identification model, and the leakage current data is used to complete the leakage fault identification through the leakage fault identification model.

[0024] In a third aspect, an electronic device comprises a memory, a processor, and a computer program stored in the memory and executable in the processor, wherein the processor implements the steps of a method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network when executing the computer program.

[0025] In a fourth aspect, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network are implemented.

[0026] In a fifth aspect, a computer program product is provided, the computer product comprising a computer program, characterized in that when the computer program is executed by a processor, the steps of the method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network are implemented.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] The present invention provides a method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network. After optimizing the parameters of variational modal decomposition using a slime mold algorithm, the leakage current data is decomposed to construct a leakage fault feature set and establish a leakage fault identification model for leakage fault identification. The modal components of different frequency bands in the leakage current data can be more accurately decomposed, and more refined and distinguishable fault features can be extracted, thereby improving the accuracy of leakage fault identification. In addition, the present invention inputs the leakage fault feature set into a convolutional neural network model to establish a leakage fault identification model, and can automatically learn and extract leakage fault features through convolution and pooling operations, thereby improving the efficiency and accuracy of leakage fault identification. Therefore, the present invention can accurately identify different leakage fault types in low-voltage AC / DC hybrid distribution networks, provide a reference for the research and development of residual current protection devices suitable for low-voltage AC / DC hybrid distribution networks, and is of great significance to ensuring the safety and reliability of power supply to low-voltage AC / DC hybrid distribution networks.

[0029] Furthermore, the convolutional neural network model of the present invention sets a specific convolution kernel size and pooling layer, which helps to reduce the amount of calculation and improve the operating efficiency of the leakage fault identification model. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A flow chart of a method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network provided in Example 1;

[0031] Figure 2 Schematic diagram of the process of optimizing VMD parameters for SMA;

[0032] Figure 3 This is a schematic diagram of the structure of the convolutional neural network model used in Example 1;

[0033] Figure 4 A topological structure diagram of a low-voltage AC / DC hybrid distribution network simulation model provided in Example 1;

[0034] Figure 5(a) to Figure 5(c) The leakage current data collected in different states in Example 1, wherein FIG5

[0035] (a) is the leakage current signal of the normal leakage state of the line, Figure 5(b) is the leakage current signal of the photovoltaic system leakage state, and Figure 5(c) is the leakage current signal of the electric shock fault state;

[0036] Figure 6 The fitness change curve of the SMA optimization VMD parameter process provided in Example 1;

[0037] Figure 7(a) to Figure 7(c) The VMD decomposition result diagram after the SMA optimizes the VMD parameters provided in Example 1, FIG7 (a) is the VMD decomposition result of the normal leakage state of the line, FIG7 (b) is the VMD decomposition result of the leakage state of the photovoltaic system, and FIG7 (c) is the VMD decomposition result of the electric shock fault state;

[0038] Figure 8 This is a recognition result diagram of the CNN-based leakage fault recognition model provided in Example 1;

[0039] Fig. 9 A schematic diagram of the structure of a low-voltage AC / DC hybrid distribution network leakage fault identification system provided in Example 2;

[0040] Fig.10 This is a schematic diagram of the computer device structure used in Example 2. DETAILED DESCRIPTION

[0041] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solution of the present invention will be further described in detail below in conjunction with the accompanying drawings, and the described content is intended to explain the present invention rather than to limit it.

[0042] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units that are not explicitly listed or are inherent to these processes, methods, systems, products or apparatus.

[0043] Embodiment 1

[0044] In order to solve the problem that the existing residual current protection device is difficult to accurately identify the leakage type in the low-voltage AC / DC hybrid distribution network connected to distributed photovoltaics, which often causes the residual current protection device to malfunction, this embodiment provides a low-voltage AC / DC hybrid distribution network leakage identification method based on optimized variational mode decomposition and convolutional neural network to achieve accurate identification of different leakage types. The following first introduces the principles of variational mode decomposition, slime mold algorithm optimization of VMD parameters, and convolutional neural network:

[0045] 1. Variational mode decomposition (VMD) is a non-recursive signal processing method. It abandons the recursive method in the traditional decomposition method in the signal decomposition process. Compared with empirical mode decomposition, it has stronger robustness and better effect in decomposing complex signals. The VMD algorithm can be divided into constructing variational problems and solving variational problems. The specific steps are as follows:

[0046] (1) Constructing a variational problem

[0047] Decompose the original signal f(t) into K modal components u k (t), the envelope signal of each modal component is obtained by Hilbert transform; between f(t) and each modal component u k (t) is equal to the constraint condition, the envelope signal is compared with the estimated center frequency ω k Mix, as shown in formula (1):

[0048]

[0049] Among them, δ(t) is the impulse function; * is the convolution operation.

[0050] (2) Solving variational problems

[0051] To solve the variational problem, by introducing λ(t) and α, the above equation (1) can be transformed into equation (2):

[0052]

[0053] Among them, λ(t) is the Lagrangian operator; α is the quadratic penalty factor.

[0054] In order to solve the optimal parameter u in equation (1) k and ω k , the specific steps are as follows:

[0055] (1) Initialization λ 1 , VMD iteration number n = 0. Set the loop process, let n = n + 1, u k ,ω k and λ are updated in turn according to equations (3) to (6).

[0056] (2) Update u in the non-negative frequency range k :

[0057]

[0058] in, f(ω),λ n (ω) are f(t),λ n (t) The corresponding Fourier transform.

[0059] (3) Update ω k :

[0060]

[0061] (4) In the non-negative frequency τ interval, update λ:

[0062]

[0063] (5) For a given judgment accuracy ε>0, we have

[0064]

[0065] (6) If equation (6) is satisfied, stop the iteration; otherwise, return to step (2).

[0066] 2. Slime Mold Algorithm Optimization of VMD Parameters

[0067] The slime mold algorithm is introduced to optimize the penalty factor α and the number of modal components K of VMD to obtain the optimal parameter combination of VMD. The slime mold algorithm (Slime mould algorithm, SMA) is a swarm intelligence optimization algorithm that simulates the behavior and morphological changes of slime mold foraging process. The iterative process of SMA mainly includes three stages: initialization stage: random diffusion, looking for the food with the strongest smell; exploration stage: approaching and diffusing to the food with the strongest smell; development stage: completing the package of the target.

[0068] (1) Initialization phase

[0069] In SMA, the initialization position update formula of slime mold population is shown in formula (7):

[0070] X i =rand(Ub-Lb)+Lb(7)

[0071] Where, X i is the specific location of the i-th slime mold; Ub and Lb are the boundaries of the slime mold's activity range; rand is an arbitrary value ranging from 0 to 1.

[0072] (2) Exploration phase

[0073] Even if the slime mold finds a better food source, some of them will continue to look for better food, and the position of some slime molds will change. The position change of the slime mold in the exploration stage can be expressed as:

[0074] X(t+1)=rand(Ub-Lb)+Lb,rand <z(8)

[0075] Where: z is the threshold parameter, its value is relatively low, generally 0.03.

[0076] (3) Development stage

[0077] When the slime mold finds food, it immediately takes action to capture the food. The process can be expressed as:

[0078]

[0079] Where: vb is a random number in [-a, a], vc is a value that oscillates in [-1, 1] and finally approaches zero, X b (t) indicates that the slime mold has found the best location for food at this time, that is, the optimal location for the current search. X(t) indicates the specific location of the current slime mold individual. X A (t) and X B (t) represents the position of two random slime mold individuals, t is the current iteration number, W is the weight coefficient, and p is the control parameter.

[0080] 3. Convolutional Neural Network

[0081] Convolutional Neural Networks (CNN) is one of the representative algorithms of deep learning. It is mainly composed of input layer, convolution layer, pooling layer, fully connected layer and output layer connected in sequence. Among them, the convolution layer is the core component of CNN. The convolution layer obtains the main features of the data through convolution operation and outputs it to the pooling layer. The calculation formula of the convolution operation is shown in formula (10):

[0082]

[0083] Where Conv is the output of the convolutional layer, W C is the data weight, b c is the bias, ReLU is the activation function, is the corresponding convolution operation.

[0084] The pooling layer is a processing method after the convolution operation. The pooling layer mainly compresses the output results of the convolution layer and reduces the complexity of the network. The calculation formula is shown in formula (11):

[0085] P = max(Conv) + b p (11)

[0086] In the formula, P is the output of the pooling layer, max represents the maximum pooling, and b p is the residual.

[0087] The fully connected layer, convolutional layer, and pooling layer form an integral structure. The fully connected layer is used to convert the output into a one-dimensional vector of the classification result. The calculation formula of the fully connected layer is shown in formula (12):

[0088] F=S(PW z +b z )(12)

[0089] Where, F is the output result of the fully connected layer, which is the transformer operating status output code in this embodiment, P is the output of the pooling layer, and W z and b z are the corresponding weights and biases, and S is the activation function.

[0090] Next, the method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network provided by this embodiment is described in detail. Figure 1 As shown, the following steps are included:

[0091] Step 1: Collect leakage current data of different states of low-voltage AC / DC hybrid distribution network;

[0092] Through the low-voltage AC / DC hybrid distribution network simulation model, 300 sets of leakage current data of normal line leakage state, photovoltaic system leakage state and electric shock fault state are obtained. Figure 4 FIG. 5 shows a topological structure of a low-voltage AC / DC hybrid power distribution simulation model, and FIG. 5 shows a leakage current signal obtained by simulation in each state.

[0093] Step 2: After optimizing the parameters of variational mode decomposition using the slime mold algorithm, the leakage current data is decomposed to obtain the modal components of different frequency bands [IMF 1 , IMF 2 , …, IMF K]; Based on the principle of optimizing VMD parameters by the slime mold algorithm, the process of optimizing VMD parameters by the slime mold algorithm is as follows: Figure 2 As shown;

[0094] The slime mold algorithm is used to optimize the parameters in VMD. The relevant parameters are set as follows: the search range of the optimal number of modal components K is [2, 10], the search range of the penalty factor is [1, 1000], the population size N = 5, the maximum number of iterations T = 10, the search space dimension d = 2, and the appendix Figure 6 The figure shows the fitness change curve of SMA optimizing VMD parameters. Figure 6 It can be seen that when SMA is used to optimize VMD parameters, the best fitness curve gradually converges with the increase of optimization times, and converges to the minimum value after 9 iterations. At this time, the best fitness value is 0.005352, and the corresponding optimal parameter combination is [9, 984].

[0095] VMD decomposition is performed on 300 leakage current signals in three states under the optimal parameter combination. FIG7 shows the VMD decomposition results of the leakage current signals in three states in FIG5.

[0096] Step 3: Calculate the sample entropy values ​​of K IMF components respectively to establish a K-dimensional feature vector, that is, to construct a leakage fault feature set;

[0097] The sample entropy is calculated for the 9 IMF components decomposed from each sample, as shown in Appendix 1.

[0098] Table 1 Sample entropy of modal components in different states

[0099]

[0100] Step 4: Randomly select 70% of the samples as the training set and the remaining 30% of the samples as the test set;

[0101] 210 samples (70 samples in each state) are randomly selected as the training set, and the remaining 90 samples (30 samples in each state) are used as the test set.

[0102] Step 5: Use the training sample set to train the convolutional neural network model and establish a CNN-based leakage fault recognition model; the CNN model structure of this embodiment is as follows: Figure 3 As shown in the figure, it consists of an input layer, two convolutional layers, two pooling layers, a fully connected layer and an output layer, where the size of the convolution kernel is 2×1, the size of the pooling layer is 2×1, and the fully connected layer is composed of the output category state encoding.

[0103] The CNN model is trained using the training set. The learning rate of the CNN model is set to 0.012, the maximum number of iterations is 600, the batch size is 30, and the other CNN structure parameters are shown in Appendix 2.

[0104] Table 2 Structural parameters of CNN model

[0105]

[0106] Step 6: Use the test set to test the recognition effect of the constructed CNN-based leakage fault recognition model, output the leakage state type, and use the recognition accuracy as the performance evaluation indicator.

[0107] The test set is input into the CNN model to identify the leakage fault type of the test sample, and the leakage type of the test sample is obtained. The confusion matrix of a certain recognition result is shown in the attached figure. Figure 8 As shown, the recognition accuracy is 93.33%.

[0108] The present embodiment provides a method for identifying leakage faults in low-voltage AC / DC hybrid distribution networks. First, the penalty factor and the number of IMF components of VMD are optimized by SMA, and the optimized VMD method is used to decompose leakage signals of different states of low-voltage AC / DC hybrid distribution networks; then, the sample entropy of each modal component after decomposition is calculated to construct a leakage fault feature set; finally, the entropy features of the training samples are input into the CNN model for training to establish a leakage fault identification model based on CNN, and the entropy features of the samples to be tested are input into the trained leakage fault identification model to obtain the leakage fault type. It can more accurately identify different leakage fault types of low-voltage AC / DC hybrid distribution networks, provide a reference for the research and development of residual current protection devices suitable for low-voltage AC / DC hybrid distribution networks, and is of great significance to ensuring the safety and reliability of power supply of low-voltage AC / DC hybrid distribution networks.

[0109] Embodiment 2

[0110] This embodiment provides a low voltage AC / DC hybrid distribution network leakage fault identification system. Fig. 9 As shown, it includes: a data acquisition module, an optimized variational mode decomposition module and a leakage fault identification module;

[0111] A data acquisition module is used to collect leakage current data of different states of a low-voltage AC / DC hybrid distribution network;

[0112] Specifically, different states of leakage current data include: normal line leakage state, photovoltaic system leakage state and electric shock fault state;

[0113] An optimized variational modal decomposition module is used to decompose the leakage current data using the optimized variational modal decomposition to obtain modal components of different frequency bands, wherein the optimized variational modal decomposition is obtained by optimizing the parameters of the variational modal decomposition using a slime mold algorithm;

[0114] Specifically, the parameters of variational mode decomposition include penalty factor and the number of modal components;

[0115] Specifically, the steps of optimizing the parameters of variational mode decomposition by using the slime mold algorithm include:

[0116] S1, set the slime mold population size and maximum number of iterations;

[0117] S2, initializing the individual positions of slime molds according to the size of the slime mold population;

[0118] S3, after calculating the fitness according to the current slime mold individual position and sorting the fitness, updating the optimal fitness and its corresponding optimal slime mold individual position;

[0119] S4, after updating the search range and search direction of the slime mold population food according to the current distribution of the slime mold individual positions and the fitness, the slime mold individual positions are updated and it is determined whether the maximum number of iterations is reached. If the maximum number of iterations is reached, the optimal slime mold individual positions are output. If the maximum number of iterations is not reached, after the slime mold individual positions are updated, the steps of updating the optimal fitness and its corresponding optimal slime mold individual positions are repeated until the maximum number of iterations is reached; the slime mold individual positions are the parameters of the variational mode decomposition to be optimized.

[0120] The leakage fault identification module is used to calculate the sample entropy of each modal component to construct a leakage fault feature set, input the leakage fault feature set into a convolutional neural network model to establish a leakage fault identification model, and complete the leakage fault identification through the leakage fault identification model.

[0121] Specifically, the convolutional neural network model includes an input layer, a first convolutional layer, a second convolutional layer, a first pooling layer, a second pooling layer, a fully connected layer and an output layer, wherein the input layer is connected to the first convolutional layer, the first convolutional layer is connected to the first pooling layer, the first pooling layer is connected to the second convolutional layer, the second convolutional layer is connected to the second pooling layer, the second pooling layer is connected to the fully connected layer, and the fully connected layer is connected to the output layer; the size of the convolution kernel is 2×1, and the size of the first pooling layer and the second pooling layer is 2×1;

[0122] The division of modules in the embodiments of the present invention is schematic and is only a logical function division. There may be other division methods in actual implementation. In addition, each functional module in each embodiment of the present invention may be integrated into one processor, or may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated modules may be implemented in the form of hardware or in the form of software functional modules.

[0123] This embodiment also provides a computer device, such as Fig.10 As shown, the computer device includes a processor and a memory, the memory is used to store a computer program (in this embodiment, the computer program includes a computing component and an iterative component, which can perform model calculation and model update), the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of a method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network.

[0124] This embodiment also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It is understandable that the computer-readable storage medium here can include both built-in storage media in a computer device and, of course, an extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by a processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of a method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network in the above embodiment.

[0125] This embodiment further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the corresponding steps of the method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network in the above embodiment are implemented.

[0126] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0127] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0128] These computer program instructions 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 an article of manufacture comprising 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.

[0129] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0130] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for identifying leakage faults in a low voltage AC / DC hybrid distribution network, characterized in that: The following steps are involved: Collect leakage current data of different states of low-voltage AC / DC hybrid distribution network; After decomposing the leakage current data using optimized variational modal decomposition, modal components of different frequency bands are obtained; The sample entropy of each modal component is calculated to construct a leakage fault feature set, the leakage fault feature set is input into a convolutional neural network model to establish a leakage fault identification model, and the leakage current data is used to complete leakage fault identification through the leakage fault identification model.

2. A method for identifying leakage faults in a low voltage AC / DC hybrid distribution network according to claim 1, characterized in that: The parameters of the variational modal decomposition include a penalty factor and the number of modal components.

3. A method for identifying leakage faults in a low voltage AC / DC hybrid distribution network according to claim 1, characterized in that: The convolutional neural network model includes an input layer, a first convolutional layer, a second convolutional layer, a first pooling layer, a second pooling layer, a fully connected layer and an output layer; The input layer is connected to the first convolution layer, the first convolution layer is connected to the first pooling layer, the first pooling layer is connected to the second convolution layer, the second convolution layer is connected to the second pooling layer, the second pooling layer is connected to the fully connected layer, and the fully connected layer is connected to the output layer.

4. A method for identifying leakage faults in a low voltage AC / DC hybrid distribution network according to claim 3, characterized in that: The size of the convolution kernel of the convolutional neural network model is 2×1, and the size of the first pooling layer and the second pooling layer is 2×1.

5. A method for identifying leakage faults in a low voltage AC / DC hybrid distribution network according to claim 1, characterized in that: The optimized variational mode decomposition is obtained by optimizing the parameters of the variational mode decomposition by using a slime mold algorithm, which specifically includes the following steps: Set the slime mold population size and maximum number of iterations; Initialize the individual positions of slime molds by the size of the slime mold population; After calculating the fitness according to the current slime mold individual position and sorting the fitness, the optimal fitness and its corresponding optimal slime mold individual position are updated; After updating the search range and search direction of the slime mold population food according to the current distribution of the slime mold individual positions and the fitness, updating the slime mold individual positions and determining whether the maximum number of iterations has been reached, outputting the optimal slime mold individual positions if the maximum number of iterations has been reached, and repeating the steps of updating the optimal fitness and its corresponding optimal slime mold individual positions after updating the slime mold individual positions, until the maximum number of iterations has been reached; The individual positions of the slime mold are the parameters of the variational mode decomposition to be optimized.

6. A method for identifying leakage faults in a low voltage AC / DC hybrid distribution network according to claim 1, characterized in that: The different states of leakage current data include: normal line leakage state, photovoltaic system leakage state and electric shock fault state.

7. A low voltage AC / DC hybrid distribution network leakage fault identification system, characterized in that: include: A data acquisition module is used to collect leakage current data of different states of a low-voltage AC / DC hybrid distribution network; An optimized variational modal decomposition module is used to decompose the leakage current data using the optimized variational modal decomposition to obtain modal components of different frequency bands; The leakage fault identification module is used to calculate the sample entropy of each modal component to construct a leakage fault feature set, input the leakage fault feature set into a convolutional neural network model to establish a leakage fault identification model, and the leakage current data is used to complete the leakage fault identification through the leakage fault identification model.

8. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable in the processor, wherein the processor implements the steps of a method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network as claimed in any one of claims 1 to 6 when executing the computer program.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network according to any one of claims 1 to 6 are implemented.

10. A computer program product, the computer product comprising a computer program, characterized in that: When the computer program is executed by a processor, the steps of a method for identifying leakage faults in a low-voltage AC / DC hybrid distribution network as claimed in any one of claims 1 to 6 are implemented.

Images