Distribution network automation equipment abnormal state detection method and device, electronic equipment and storage medium

Abstract

The invention discloses an abnormal state detection method and device for distribution network automation equipment, electronic equipment and a storage medium. The method comprises the following steps: acquiring a first time sequence abnormal signal of the distribution network automation equipment; decomposing the first time sequence abnormal signal to generate a first feature map matrix for representing the working state of the equipment; and inputting the first time sequence abnormal signal and the first feature map matrix into a multi-tensor convolutional network model, so that the multi-tensor convolutional network model generates an abnormal state type of the first time sequence abnormal signal. By implementing the method, the anomaly detection capability of the distribution network automation equipment can be remarkably improved, and safe and efficient operation of the distribution network is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of power distribution network safety analysis, and in particular to a method, device, electronic device and storage medium for detecting abnormal state of distribution network automation equipment. Background Art

[0002] Distribution network automation equipment is an important part of the modern power distribution network. It significantly improves the reliability and operating efficiency of the distribution system through real-time monitoring and intelligent control of the power grid operation status. However, due to the complex operating environment, frequent load fluctuations, and long-term operation, distribution network automation equipment is prone to various abnormal conditions, such as equipment aging, excessive load, and communication failures. These abnormalities will not only affect the stable operation of the distribution network, but may also cause local power outages or equipment damage. Therefore, timely detection and identification of abnormal conditions of distribution network automation equipment is of great significance to ensure the safe operation of the power system.

[0003] Existing distribution network automation equipment anomaly detection technologies are usually based on preset rules or simple abnormal signal recognition methods. These methods mainly rely on thresholds summarized by experience to determine whether the equipment is abnormal. For example, when the current, voltage or communication signal strength exceeds the set range, the system will mark it as an abnormal state. However, this type of method can only preliminarily identify the existence of anomalies and cannot further distinguish the specific state type of the anomaly. Due to the complex operating environment of distribution network automation equipment, different abnormal states, such as switch jams, excessive equipment vibration, decreased communication signal strength, aging of electrical components, etc., may produce similar abnormal signal characteristics under specific conditions, making it difficult to clearly diagnose the specific fault category of the equipment only through simple rules or signal thresholds. This not only limits the accuracy and efficiency of distribution network automation equipment anomaly detection, but may also lead to reduced reliability of power grid operation, increased operation and maintenance costs, and increased safety hazards. Summary of the invention

[0004] The embodiment of the present invention provides a method, device, electronic device and storage medium for detecting abnormal state of distribution network automation equipment. By implementing the present invention, the accuracy and efficiency of abnormal state detection of distribution network automation equipment can be improved.

[0005] An embodiment of the present invention provides a method for detecting abnormal state of distribution network automation equipment, comprising:

[0006] Obtain the first timing abnormal signal of the distribution network automation equipment;

[0007] Decomposing the first timing abnormality signal to generate a first characteristic graph matrix for characterizing a working state of the device;

[0008] Inputting the first timing abnormality signal and the first feature map matrix into a multi-tensor convolutional network model, so that the multi-tensor convolutional network model generates an abnormal state type of the first timing abnormality signal;

[0009] The training of the multi-tensor convolutional network model includes:

[0010] Acquire a plurality of second timing abnormal signals for training and abnormal state labels corresponding to the second timing abnormal signals; wherein the abnormal state label is the abnormal state type corresponding to the second timing abnormal signal;

[0011] Decomposing the second timing abnormality signal to generate a second characteristic graph matrix for characterizing the working state of the device;

[0012] Constructing a training set according to the second time series abnormal signal, the second feature map matrix, and the abnormal state label corresponding to the second time series abnormal signal;

[0013] Randomly dividing the training set into several batches of training samples according to a preset batch size;

[0014] Input each batch of training samples into the multi-tensor convolutional network model in turn, and train the multi-tensor convolutional network model until a preset number of training times is reached; wherein, the multi-tensor convolutional network model outputs the abnormal state type corresponding to the training sample each time it receives a batch of training samples; calculates the loss function value through the loss function according to the abnormal state type and the corresponding abnormal state label; and updates the multi-tensor convolutional network model according to the loss function value.

[0015] Further, obtaining a first timing abnormal signal of the distribution network automation equipment includes:

[0016] Obtain the original timing abnormality signal of distribution network automation equipment;

[0017] Calculating the autocorrelation function of the original time series abnormal signal;

[0018] Analyzing the autocorrelation function to identify random noise of the original time series abnormal signal;

[0019] The random noise of the original timing abnormality signal is removed according to a preset global threshold value to generate a first timing abnormality signal of the distribution network automation equipment.

[0020] Furthermore, the decomposing the first timing abnormality signal to generate a first characteristic diagram matrix for characterizing the working state of the device includes:

[0021] Performing multi-frequency decomposition on the first time series abnormal signal using a collective empirical mode decomposition method to generate a plurality of first fixed mode components;

[0022] A first characteristic diagram matrix for characterizing the working state of the device is generated according to the first fixed modal components; wherein each element in the first characteristic diagram matrix is ​​a value of the first fixed modal component at a certain moment.

[0023] Furthermore, the loss function value is calculated by the following formula:

[0024]

[0025] in, w is the weight of the multi-tensor convolutional network model; b is the bias coefficient of the multi-tensor convolutional network model; x is the second feature map matrix; The abnormal status type actually output; n is the number of training samples; y Abnormal status label.

[0026] Based on the above method embodiment, the present invention provides a corresponding device embodiment.

[0027] An embodiment of the present invention provides a distribution network automation equipment abnormal state detection device, including: a timing abnormal signal acquisition module, a timing abnormal signal decomposition module, an abnormal state type detection module and a model training module.

[0028] The timing abnormality signal acquisition module is used to acquire the first timing abnormality signal of the distribution network automation equipment;

[0029] The timing abnormal signal decomposition module is used to decompose the first timing abnormal signal to generate a first characteristic diagram matrix for characterizing the working state of the device;

[0030] The abnormal state type detection module is used to input the first timing abnormal signal and the first feature map matrix into a multi-tensor convolutional network model, so that the multi-tensor convolutional network model generates an abnormal state type of the first timing abnormal signal;

[0031] The model training module is used to obtain several second timing abnormality signals for training and abnormal state labels corresponding to the second timing abnormality signals; wherein the abnormal state label is the abnormal state type corresponding to the second timing abnormality signal; decompose the second timing abnormality signal to generate a second feature map matrix for characterizing the working state of the equipment; construct a training set according to the second timing abnormality signal, the second feature map matrix and the abnormal state label corresponding to the second timing abnormality signal; randomly divide the training set into several batches of training samples according to a preset batch size; input the training samples of each batch into the multi-tensor convolutional network model in turn, and train the multi-tensor convolutional network model until a preset number of training times is reached; wherein the multi-tensor convolutional network model outputs the abnormal state type corresponding to the training sample when receiving a batch of training samples; calculate the loss function value through the loss function according to the abnormal state type and the corresponding abnormal state label; and update the multi-tensor convolutional network model according to the loss function value.

[0032] Further, the timing abnormal signal acquisition module includes: an original timing abnormal signal acquisition unit, an autocorrelation function calculation unit, a noise identification unit and a first timing abnormal signal generation unit;

[0033] The original timing abnormal signal acquisition unit is used to acquire the original timing abnormal signal of the distribution network automation equipment;

[0034] The autocorrelation function calculation unit is used to calculate the autocorrelation function of the original time series abnormal signal;

[0035] The noise identification unit is used to analyze the autocorrelation function and identify the random noise of the original time series abnormal signal;

[0036] The first timing abnormality signal generating unit is used to remove the random noise of the original timing abnormality signal according to a preset global threshold value to generate the first timing abnormality signal of the distribution network automation equipment.

[0037] Furthermore, the timing abnormal signal decomposition module includes: a multi-frequency decomposition unit and a first characteristic graph matrix generation unit;

[0038] The multi-frequency decomposition unit is used to perform multi-frequency decomposition on the first time series abnormal signal by using a collective empirical mode decomposition method to generate a plurality of first fixed mode components;

[0039] The first characteristic diagram matrix generating unit is used to generate a first characteristic diagram matrix for characterizing the working state of the device according to a plurality of the first fixed modal components; wherein each element in the first characteristic diagram matrix is ​​the value of the first fixed modal component at a certain moment.

[0040] Furthermore, the loss function value is calculated by the following formula:

[0041]

[0042] in, w is the weight of the multi-tensor convolutional network model; b is the bias coefficient of the multi-tensor convolutional network model; x is the second feature map matrix; The abnormal status type actually output; n is the number of training samples; y Abnormal status label.

[0043] Based on the above method embodiment, the present invention provides a corresponding electronic device embodiment.

[0044] An embodiment of the present invention provides 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, the method for detecting abnormal state of distribution network automation equipment described in any one of the above method embodiments can be implemented.

[0045] Based on the above method item embodiments, the present invention provides a corresponding storage medium item embodiment.

[0046] An embodiment of the present invention provides a storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for detecting abnormal state of distribution network automation equipment described in any one of the above method embodiments can be implemented.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The embodiment of the present invention provides a method, device, electronic device and storage medium for detecting abnormal state of distribution network automation equipment. The method obtains the timing abnormal signal of the distribution network automation equipment, decomposes the signal, generates a series of feature map matrices to characterize the working state of the equipment, and effectively extracts the multi-dimensional information of the working state of the equipment; these timing abnormal signals and feature map matrices are input into the multi-tensor convolutional network model, so that the multi-tensor convolutional network model generates the type of abnormal state. The timing abnormal signal of the equipment and its corresponding feature map matrix are jointly analyzed by the multi-tensor network model. The network can extract meaningful feature information from multiple dimensions and multiple signal sources, effectively reduce the information loss caused by a single signal or feature dimension, and can accurately generate the abnormal state type according to the input timing data and feature map matrix. In the model training process, by using the historical timing abnormal signal and its corresponding abnormal state label, the training sample is constructed, and after multiple trainings, the model can accurately identify the abnormal state type of the equipment. By implementing the present invention, the abnormal detection capability of the distribution network automation equipment can be significantly improved, and the safe and efficient operation of the distribution network is guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 The present invention is a flowchart of a method for detecting abnormal state of distribution network automation equipment provided by an embodiment of the present invention.

[0050] Figure 2 It is a schematic diagram of a process for training a multi-tensor convolutional network model provided by one embodiment of the present invention.

[0051] Figure 3 It is a structural schematic diagram of a distribution network automation equipment abnormal state detection device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] like Figure 1 As shown, an embodiment of the present invention provides a method for detecting abnormal state of distribution network automation equipment, which at least includes the following steps:

[0054] Step S1, obtaining a first timing abnormal signal of a distribution network automation device;

[0055] Specifically, in a preferred embodiment, obtaining a first timing abnormality signal of a distribution network automation device includes:

[0056] Obtain the original timing abnormality signal of distribution network automation equipment;

[0057] Calculating the autocorrelation function of the original time series abnormal signal;

[0058] Analyzing the autocorrelation function to identify random noise of the original time series abnormal signal;

[0059] The random noise of the original timing abnormality signal is removed according to a preset global threshold value to generate a first timing abnormality signal of the distribution network automation equipment.

[0060] It can be understood that through this implementation step, the quality and analysis accuracy of the abnormal signal of the distribution network automation equipment can be effectively improved. By obtaining the original timing abnormal signal of the distribution network automation equipment and calculating its autocorrelation function, the time dependence of the signal can be fully utilized to reveal its periodic characteristics and random noise. By analyzing the autocorrelation function, it is possible to accurately distinguish between periodic signals and random noise, thereby effectively removing the noise in the signal. This denoising process is based on a preset global threshold and can flexibly respond to different types of noise, thereby retaining effective information related to the equipment status. Finally, the generated first timing abnormal signal is purer and can provide high-quality data input for subsequent feature extraction and anomaly detection.

[0061] Step S2, decomposing the first timing abnormality signal to generate a first characteristic diagram matrix for characterizing the working state of the device;

[0062] Specifically, in a preferred embodiment, the decomposing the first timing abnormality signal to generate a first characteristic diagram matrix for characterizing the working state of the device includes:

[0063] Performing multi-frequency decomposition on the first time series abnormal signal using a collective empirical mode decomposition method to generate a plurality of first fixed mode components;

[0064] A first characteristic diagram matrix for characterizing the working state of the device is generated according to the first fixed modal components; wherein each element in the first characteristic diagram matrix is ​​a value of the first fixed modal component at a certain moment.

[0065] In an optional embodiment, when the first time series abnormal signal is decomposed by the ensemble empirical mode decomposition (EEMD) method, multiple groups of different Gaussian white noise sequences are first superimposed on the first time series abnormal signal to reduce the mode aliasing problem in the first time series abnormal signal and improve the stability of the decomposition, wherein the amplitude of the Gaussian white noise is an important parameter of the decomposition and needs to be reasonably set to balance the noise introduction and signal integrity. Next, each group of the first time series abnormal signal superimposed with noise is subjected to empirical mode decomposition (EMD), and the specific process of EMD includes: determining the local extreme value points of the first time series abnormal signal and generating the upper envelope and the lower envelope by interpolation respectively, then calculating the average value of the two as the local trend term and subtracting it from the first time series abnormal signal to obtain a candidate first intrinsic mode component; the candidate first intrinsic mode component is screened multiple times until the definition condition of the first intrinsic mode component is met, that is, the upper envelope and the lower envelope are symmetrical about the zero point within the entire first time series abnormal signal range, and the number of zero crossings is equal to or close to the number of extreme value points, and finally a first intrinsic mode component is extracted. Repeat the above steps for the remaining first time series abnormal signals until the remaining terms become monotonic functions, and complete the EMD decomposition of a group of first time series abnormal signals. EEMD averages the EMD results of multiple groups of first time series abnormal signals, thereby eliminating the randomness introduced by Gaussian noise and obtaining a set of first intrinsic modal components with stability and physical significance. These first intrinsic modal components can fully characterize the characteristics of the first time series abnormal signals at different frequency components, providing key data support for subsequent feature extraction and equipment status diagnosis.

[0066] It can be understood that the use of the ensemble empirical mode decomposition (EEMD) method to decompose the first timing abnormal signal can analyze the high and low frequency characteristics of the first timing abnormal signal. For example, abnormal fluctuations in temperature and voltage are usually more obvious in the low-frequency components, while the high-frequency components may contain abnormal vibration information of the equipment, which can more effectively distinguish the long-term trend, periodic changes and instantaneous anomalies of the equipment.

[0067] In a preferred embodiment, based on these first fixed modal components, a first characteristic diagram matrix for characterizing the working state of the device is constructed. The first characteristic diagram matrix is ​​generated by splicing all the first intrinsic modal components in a time series manner, wherein each column corresponds to a first intrinsic modal component, representing the evolution process of the first timing abnormal signal under a certain frequency component; and each row represents the amplitude of all frequency components of the first timing abnormal signal at a certain moment. Each element of the first characteristic diagram matrix represents the value of a first fixed modal component at a specific moment.

[0068] For example, taking the first abnormal time series signal as an example, the N intrinsic mode components {IMF1(t), IMF1(t), ..., IMFN (t)}, each IMF represents the part of the first time series abnormal signal at different frequency components, which are concatenated into a two-dimensional matrix M, where each column represents an IMF component and each row represents the value of the signal at a certain moment.

[0069]

[0070] Among them, each matrix element M i，j Indicates that at time point t i At time , the value of the first time series abnormal signal on the jth IMF component.

[0071] It can be understood that this construction method of the feature map matrix can represent the complex first time series abnormal signal characteristics in a two-dimensional structured form, providing a clear feature description for subsequent analysis and processing. At the same time, since each column in the matrix independently reflects the changes in a specific frequency component, it can better distinguish the operating status and potential abnormalities of the device on multi-frequency characteristics.

[0072] In a preferred embodiment, before using the ensemble empirical mode decomposition method to perform multi-frequency decomposition on the first abnormal time series signal to generate a plurality of first fixed mode components, the method further includes:

[0073] The hyperparameters in the ensemble empirical mode decomposition (EEMD) method, such as the number of integrations, the amplitude of Gaussian white noise and the number of decomposed modes, are taken as population individuals, and the vertical and horizontal crossover algorithm is used to optimize the hyperparameters.

[0074] Step S3, inputting the first timing abnormality signal and the first feature map matrix into a multi-tensor convolutional network model, so that the multi-tensor convolutional network model generates an abnormal state type of the first timing abnormality signal;

[0075] like Figure 2 As shown, in a preferred embodiment of the present invention, the training of the multi-tensor convolutional network model includes:

[0076] Step S3.1, obtaining a plurality of second timing abnormal signals for training and abnormal state labels corresponding to the second timing abnormal signals; wherein the abnormal state label is the abnormal state type corresponding to the second timing abnormal signal;

[0077] Step S3.2, decomposing the second timing abnormality signal to generate a second characteristic diagram matrix for characterizing the working state of the device;

[0078] Step S3.3, constructing a training set according to the second time series abnormal signal, the second feature map matrix, and the abnormal state label corresponding to the second time series abnormal signal;

[0079] Step S3.4, randomly dividing the training set into several batches of training samples according to a preset batch size;

[0080] Step S3.5, input each batch of training samples into the multi-tensor convolutional network model in turn, and train the multi-tensor convolutional network model until a preset number of training times is reached; wherein, the multi-tensor convolutional network model outputs the abnormal state type corresponding to the training sample each time it receives a batch of training samples; calculates the loss function value through the loss function according to the abnormal state type and the corresponding abnormal state label; and updates the multi-tensor convolutional network model according to the loss function value.

[0081] Optionally, the multi-tensor convolutional network model consists of a one-dimensional convolutional neural network and a two-dimensional convolutional neural network. In this study, the convolutional network mainly includes four layers: convolution layer, pooling layer, fusion layer and fully connected layer.

[0082] The convolution layer performs one-dimensional convolution and two-dimensional convolution operations respectively through two convolution kernels with different weights. Each convolution kernel extracts features from the input first feature map matrix in a moving window manner. The process of extracting features by the convolution kernel is described as follows:

[0083] y 2d =a 2d (W 2d *H+b 2d )

[0084] y 1d =a 1d (W 1d *V+b 1d )

[0085] Where * is the convolution operator; y 2d With y 1d Output two-dimensional features and one-dimensional features after convolution operation; W 2d is the two-dimensional convolution kernel; W 1d is a one-dimensional convolution kernel; b 2d is the bias coefficient of the two-dimensional convolution operation; b 1d is the bias coefficient of the one-dimensional convolution operation; H is the first feature map matrix; V is the first timing abnormal signal.

[0086] The pooling layer is used to further remove redundant information and improve the robustness and generalization ability of the overall network. The pooling layer in the convolutional neural network of this study adopts the maximum pooling method and applies the maximum pooling once. Its mathematical description is as follows:

[0087] z i =max{z i-1 [mL,(m+1)L]}

[0088]

[0089] z i is the feature matrix after maximum pooling; z i-1 is the feature matrix output by the previous convolutional layer; L is the window size of the pooling operation; S is the stride of the pooling operation; m is the number or index of the pooling window, which is used to gradually slide the pooling window, select a local area and calculate the maximum value.

[0090] The fusion layer is used to fuse the one-dimensional features and the two-dimensional features after the convolution kernel pooling operation. The operation of the fusion layer is as follows:

[0091] F fused =Concat(z 1d , z 2d )

[0092] Among them, z 1d With z 2d It is the output result of the main channel and auxiliary channel pooling layer, and Concat is the concatenation operation on the feature dimension.

[0093] The fully connected layer, as the last layer in the hidden layer of the convolutional neural network, is flattened into a one-dimensional feature vector as the input of the fully connected layer. The neurons in the layer are fully connected with all neurons in the previous layer, and the fused features are used as the input of the classifier Softmax to output the probability of the abnormal state in the form of probability coding. According to the probability of the abnormal state, the abnormal state type corresponding to the first time series signal is determined.

[0094] Specifically, the loss function value is calculated by the following formula:

[0095]

[0096] Among them, w is the weight of the multi-tensor convolutional network model; b is the bias coefficient of the multi-tensor convolutional network model; x is the second feature map matrix; is the abnormal state type actually output; n is the number of training samples; y is the abnormal state label.

[0097] Based on the above method embodiment, the present invention provides a corresponding device embodiment.

[0098] like Figure 3 As shown, an embodiment of the present invention provides a distribution network automation equipment abnormal state detection device, including: a timing abnormal signal acquisition module, a timing abnormal signal decomposition module, an abnormal state type detection module and a model training module;

[0099] The timing abnormality signal acquisition module is used to acquire the first timing abnormality signal of the distribution network automation equipment;

[0100] The timing abnormal signal decomposition module is used to decompose the first timing abnormal signal to generate a first characteristic diagram matrix for characterizing the working state of the device;

[0101] The abnormal state type detection module is used to input the first timing abnormal signal and the first feature map matrix into a multi-tensor convolutional network model, so that the multi-tensor convolutional network model generates an abnormal state type of the first timing abnormal signal;

[0102] The model training module is used to obtain several second timing abnormality signals for training and abnormal state labels corresponding to the second timing abnormality signals; wherein the abnormal state label is the abnormal state type corresponding to the second timing abnormality signal; decompose the second timing abnormality signal to generate a second feature map matrix for characterizing the working state of the equipment; construct a training set according to the second timing abnormality signal, the second feature map matrix and the abnormal state label corresponding to the second timing abnormality signal; randomly divide the training set into several batches of training samples according to a preset batch size; input the training samples of each batch into the multi-tensor convolutional network model in turn, and train the multi-tensor convolutional network model until a preset number of training times is reached; wherein the multi-tensor convolutional network model outputs the abnormal state type corresponding to the training sample when receiving a batch of training samples; calculate the loss function value through the loss function according to the abnormal state type and the corresponding abnormal state label; and update the multi-tensor convolutional network model according to the loss function value.

[0103] Specifically, the timing abnormal signal acquisition module includes: an original timing abnormal signal acquisition unit, an autocorrelation function calculation unit, a noise identification unit and a first timing abnormal signal generation unit;

[0104] The original timing abnormal signal acquisition unit is used to acquire the original timing abnormal signal of the distribution network automation equipment;

[0105] The autocorrelation function calculation unit is used to calculate the autocorrelation function of the original time series abnormal signal;

[0106] The noise identification unit is used to analyze the autocorrelation function and identify the random noise of the original time series abnormal signal;

[0107] The first timing abnormality signal generating unit is used to remove the random noise of the original timing abnormality signal according to a preset global threshold value to generate the first timing abnormality signal of the distribution network automation equipment.

[0108] Specifically, the timing abnormal signal decomposition module includes: a multi-frequency decomposition unit and a first characteristic graph matrix generation unit;

[0109] The multi-frequency decomposition unit is used to perform multi-frequency decomposition on the first time series abnormal signal by using a collective empirical mode decomposition method to generate a plurality of first fixed mode components;

[0110] The first characteristic diagram matrix generating unit is used to generate a first characteristic diagram matrix for characterizing the working state of the device according to a plurality of the first fixed modal components; wherein each element in the first characteristic diagram matrix is ​​the value of the first fixed modal component at a certain moment.

[0111] Exemplarily, the loss function value is calculated by the following formula:

[0112]

[0113] Among them, w is the weight of the multi-tensor convolutional network model; b is the bias coefficient of the multi-tensor convolutional network model; x is the second feature map matrix; is the abnormal state type actually output; n is the number of training samples; y is the abnormal state label.

[0114] It should be noted that the embodiment of the device described above corresponds to the above-mentioned embodiment of the present invention, and it can implement any of the above-mentioned distribution network automation equipment abnormal state detection methods of the present invention. In addition, the embodiment of the above-mentioned device is merely schematic, wherein the modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the accompanying drawings of the embodiment of the device provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which can be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement it without paying any creative labor.

[0115] Based on the above method embodiment of the present invention, a corresponding electronic device embodiment is provided.

[0116] An embodiment of the present invention provides an electronic device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the abnormal state detection method of distribution network automation equipment described in any one of the present invention is implemented, or when the processor executes the computer program, the functions of each module in the above-mentioned device embodiments are implemented.

[0117] Exemplarily, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program in the terminal device.

[0118] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0119] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) 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, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0120] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the terminal device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0121] Based on the above method embodiment, the present invention provides a storage medium embodiment;

[0122] Another embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute any one of the above-mentioned distribution network automation equipment abnormal state detection methods of the present invention.

[0123] The storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0124] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0125] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for detecting abnormal state of distribution network automation equipment, characterized in that: include: Obtain the first timing abnormal signal of the distribution network automation equipment; Decomposing the first timing abnormality signal to generate a first characteristic graph matrix for characterizing a working state of the device; Inputting the first timing abnormality signal and the first feature map matrix into a multi-tensor convolutional network model, so that the multi-tensor convolutional network model generates an abnormal state type of the first timing abnormality signal; The training of the multi-tensor convolutional network model includes: Acquire a plurality of second timing abnormal signals for training and abnormal state labels corresponding to the second timing abnormal signals; wherein the abnormal state label is the abnormal state type corresponding to the second timing abnormal signal; Decomposing the second timing abnormality signal to generate a second characteristic graph matrix for characterizing the working state of the device; Constructing a training set according to the second time series abnormal signal, the second feature map matrix, and the abnormal state label corresponding to the second time series abnormal signal; Randomly dividing the training set into several batches of training samples according to a preset batch size; Input each batch of training samples into the multi-tensor convolutional network model in turn, and train the multi-tensor convolutional network model until a preset number of training times is reached; wherein, the multi-tensor convolutional network model outputs the abnormal state type corresponding to the training sample each time it receives a batch of training samples; calculates the loss function value through the loss function according to the abnormal state type and the corresponding abnormal state label; and updates the multi-tensor convolutional network model according to the loss function value.

2. A method for detecting abnormal state of distribution network automation equipment according to claim 1, characterized in that: Obtain the first timing abnormal signal of the distribution network automation equipment, including: Obtain the original timing abnormality signal of distribution network automation equipment; Calculating the autocorrelation function of the original time series abnormal signal; Analyzing the autocorrelation function to identify random noise of the original time series abnormal signal; The random noise of the original timing abnormality signal is removed according to a preset global threshold value to generate a first timing abnormality signal of the distribution network automation equipment.

3. A method for detecting abnormal state of distribution network automation equipment according to claim 2, characterized in that: The step of decomposing the first timing abnormality signal to generate a first characteristic graph matrix for characterizing the working state of the device includes: Performing multi-frequency decomposition on the first time series abnormal signal using a collective empirical mode decomposition method to generate a plurality of first fixed mode components; A first characteristic diagram matrix for characterizing the working state of the device is generated according to the first fixed modal components; wherein each element in the first characteristic diagram matrix is ​​a value of the first fixed modal component at a certain moment.

4. A method for detecting abnormal state of distribution network automation equipment according to claim 3, characterized in that: The loss function value is calculated by the following formula: in, w is the weight of the multi-tensor convolutional network model; b is the bias coefficient of the multi-tensor convolutional network model; x is the second feature map matrix; The abnormal status type actually output; n is the number of training samples; y is the abnormal state label.

5. A distribution network automation equipment abnormal state detection device, characterized in that: include: Timing abnormal signal acquisition module, timing abnormal signal decomposition module, abnormal state type detection module and model training module; The timing abnormality signal acquisition module is used to acquire the first timing abnormality signal of the distribution network automation equipment; The timing abnormal signal decomposition module is used to decompose the first timing abnormal signal to generate a first characteristic diagram matrix for characterizing the working state of the device; The abnormal state type detection module is used to input the first timing abnormal signal and the first feature map matrix into a multi-tensor convolutional network model, so that the multi-tensor convolutional network model generates an abnormal state type of the first timing abnormal signal; The model training module is used to obtain several second timing abnormality signals for training and abnormal state labels corresponding to the second timing abnormality signals; wherein the abnormal state label is the abnormal state type corresponding to the second timing abnormality signal; decompose the second timing abnormality signal to generate a second feature map matrix for characterizing the working state of the equipment; construct a training set according to the second timing abnormality signal, the second feature map matrix and the abnormal state label corresponding to the second timing abnormality signal; randomly divide the training set into several batches of training samples according to a preset batch size; input the training samples of each batch into the multi-tensor convolutional network model in turn, and train the multi-tensor convolutional network model until a preset number of training times is reached; wherein the multi-tensor convolutional network model outputs the abnormal state type corresponding to the training sample when receiving a batch of training samples; calculate the loss function value through the loss function according to the abnormal state type and the corresponding abnormal state label; and update the multi-tensor convolutional network model according to the loss function value.

6. A distribution network automation equipment abnormal state detection device as claimed in claim 5, characterized in that: The timing abnormal signal acquisition module includes: an original timing abnormal signal acquisition unit, an autocorrelation function calculation unit, a noise identification unit and a first timing abnormal signal generation unit; The original timing abnormal signal acquisition unit is used to acquire the original timing abnormal signal of the distribution network automation equipment; The autocorrelation function calculation unit is used to calculate the autocorrelation function of the original time series abnormal signal; The noise identification unit is used to analyze the autocorrelation function and identify the random noise of the original time series abnormal signal; The first timing abnormality signal generating unit is used to remove the random noise of the original timing abnormality signal according to a preset global threshold value to generate the first timing abnormality signal of the distribution network automation equipment.

7. A distribution network automation equipment abnormal state detection device as claimed in claim 6, characterized in that: The timing abnormal signal decomposition module includes: a multi-frequency decomposition unit and a first characteristic graph matrix generation unit; The multi-frequency decomposition unit is used to perform multi-frequency decomposition on the first time series abnormal signal by using a collective empirical mode decomposition method to generate a plurality of first fixed mode components; The first characteristic diagram matrix generating unit is used to generate a first characteristic diagram matrix for characterizing the working state of the device according to a plurality of the first fixed modal components; wherein each element in the first characteristic diagram matrix is ​​the value of the first fixed modal component at a certain moment.

8. A distribution network automation equipment abnormal state detection device as claimed in claim 7, characterized in that: The loss function value is calculated by the following formula: in, w is the weight of the multi-tensor convolutional network model; b is the bias coefficient of the multi-tensor convolutional network model; x is the second feature map matrix; The abnormal status type actually output; n is the number of training samples; y is the abnormal state label.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for detecting abnormal state of distribution network automation equipment as described in any one of claims 1 to 4 can be implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it can implement the abnormal state detection method of distribution network automation equipment as described in any one of claims 1 to 4.

Images