Fault identification method and device, computer equipment, readable storage medium and program product

Abstract

The invention relates to a fault identification method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring a three-phase voltage signal; performing multi-scale analysis on the three-phase voltage signal to obtain a multi-fractal feature; performing feature extraction based on the multi-fractal features to obtain a fault feature vector; the fault feature vector comprises a multi-fractal spectrum width of each phase voltage and a fractal dimension difference; and identifying the fault feature vector to obtain a fault type. By adopting the method, the fault type can be accurately identified.

Description

Technical Field

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

[0002] The power system is an important infrastructure of modern society, and its stability and reliability directly affect economic development and social life. As a key component of the power system, high-voltage transmission lines are responsible for the important task of transmitting electricity from power stations to various places. However, since high-voltage transmission lines are exposed to the outside, they are susceptible to various natural factors (such as lightning, wind, and freezing) and human factors (such as misoperation and mechanical damage), and the risk of failure is relatively high.

[0003] Once a transmission line fault occurs, it may cause a large-scale power outage or even trigger a chain failure of the power system. Therefore, accurate and rapid identification of the fault type is crucial for locating the fault and restoring the line in a timely manner.

[0004] However, related technologies are often complex and computationally intensive in the feature extraction process and are sensitive to the algorithm parameter selection, which leads to inaccurate recognition results. Summary of the invention

[0005] Based on this, it is necessary to provide a fault identification method, apparatus, computer equipment, computer-readable storage medium and computer program product that can accurately identify the fault type in response to the above technical problems.

[0006] In a first aspect, the present application provides a fault identification method, the method comprising:

[0007] Obtain three-phase voltage signals;

[0008] Performing multi-scale analysis on the three-phase voltage signal to obtain multi-fractal features;

[0009] Perform feature extraction based on the multifractal feature to obtain a fault feature vector; the fault feature vector includes the multifractal spectrum width and fractal dimension difference of each phase voltage;

[0010] The fault feature vector is identified to obtain the fault type.

[0011] In one embodiment, the performing multi-scale analysis on the three-phase voltage signal to obtain multi-fractal features includes:

[0012] Calculating the cumulative deviation of the three-phase voltage signal, and obtaining a time series to be processed according to the cumulative deviation;

[0013] The time series to be processed is fitted to obtain a fitting polynomial, and local fluctuation characteristics of different scales are calculated according to the fitting polynomial to obtain the multifractal feature.

[0014] In one embodiment, fitting the time series to be processed to obtain a fitting polynomial, and calculating local fluctuation characteristics of different scales according to the fitting polynomial, includes:

[0015] Dividing the time series to be processed to obtain multiple subsequences;

[0016] Performing second-order polynomial fitting on each of the subsequences to obtain a plurality of fitting polynomials;

[0017] The wave functions corresponding to the local waves at different orders are calculated to obtain the multi-fractal characteristics.

[0018] In one embodiment, the multifractal feature includes a multifractal spectrum width and a fractal dimension difference; and feature extraction is performed based on the multifractal feature to obtain the multifractal feature, including:

[0019] Performing logarithmic fitting on the fluctuation function to obtain the Hurst index;

[0020] Obtaining a quality index value according to the Hurst index and the order;

[0021] The multifractal spectrum width and the fractal dimension difference are calculated according to the quality indicator value.

[0022] In one embodiment, the calculating the multifractal spectrum width and the fractal dimension difference according to the quality indicator value comprises:

[0023] Calculating a first singular index, a second singular index, a first fractal dimension, and a second fractal dimension respectively according to the quality index value;

[0024] Calculating the difference between the first singular index and the second singular index to obtain the multifractal spectrum width;

[0025] The fractal dimension difference is obtained by calculation according to the first fractal dimension and the second fractal dimension.

[0026] In one embodiment, the multi-fractal feature is identified to obtain the fault type through a pre-trained recognition model; the training process of the recognition model includes:

[0027] Obtaining a sample fault feature vector; the sample fault feature vector carries a type label;

[0028] Inputting the sample fault feature vector into the initial model for training to obtain a prediction type;

[0029] A target loss function is obtained based on the predicted type and the type label, and the parameters of the initial model are adjusted according to the target loss function until the training is completed to obtain the recognition model.

[0030] In a second aspect, the present application provides a fault identification device, the device comprising:

[0031] An acquisition module, used for acquiring three-phase voltage signals;

[0032] An analysis module, used for performing multi-scale analysis on the three-phase voltage signal to obtain multi-fractal features;

[0033] A feature extraction module, used for performing feature extraction based on the multi-fractal feature to obtain the multi-fractal feature;

[0034] The identification module is used to identify the multi-fractal features to obtain the fault type.

[0035] In a third aspect, the present application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0036] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the above method when executed by a processor.

[0037] In a fifth aspect, the present application provides a computer program product, including a computer program, which implements the steps of the above method when executed by a processor.

[0038] The above-mentioned fault identification method, device, computer equipment, computer-readable storage medium and computer program product obtain multiple fractal features by performing multi-scale analysis on the three-phase voltage signal, then extract features based on the multiple fractal features to obtain a fault feature vector, and finally identify the fault feature vector to obtain the fault type, so as to accurately identify the fault type. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 A schematic diagram of a flow chart of a fault identification method in an embodiment;

[0041] Figure 2 is a schematic diagram of a three-phase voltage signal in an embodiment;

[0042] Figure 3 is a schematic diagram of a multifractal spectrum in an embodiment;

[0043] Figure 4 is a structural block diagram of a fault identification device in one embodiment;

[0044] Figure 5 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0046] In one embodiment, Figure 1 As shown, a fault identification method is provided. This embodiment takes the method applied to a terminal as an example. It can be understood that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. In this embodiment, the method includes the following steps:

[0047] Step 102, obtaining three-phase voltage signals.

[0048] The three-phase voltage signal refers to the voltage signal between each phase in the three-phase power system. The three-phase power system is one of the most commonly used systems in power transmission and distribution. It consists of three phase lines with a fixed phase difference between each phase line. The three-phase voltage signal can be composed of three phase voltages: phase A, phase B, and phase C.

[0049] Optionally, data acquisition equipment may be used to collect data and convert the analog voltage signal into a data signal for processing.

[0050] Optionally, the collected signal may be filtered to remove noise or unnecessary frequency components.

[0051] Optionally, sampling data of 240 points before and after the fault moment is intercepted as a single voltage signal to be processed.

[0052] Exemplary, combined Figure 2 As shown, Figure 2 is a schematic diagram of a three-phase voltage signal in an embodiment. It should be noted that: Figure 2It is the three-phase voltage signal collected by the relay protection device when a C-phase fault occurs in a certain line.

[0053] Step 104: Perform multi-scale analysis on the three-phase voltage signal to obtain multi-fractal features.

[0054] Optionally, multi-scale analysis is performed on each phase voltage in the three-phase voltage signal to obtain multi-fractal features.

[0055] Among them, the multifractal characteristics describe the complexity and multifractal characteristics of the three-phase voltage signal at different scales.

[0056] In signal processing, scale usually refers to the size of the time window or the width of the frequency range used to analyze the signal. Different scales can reveal the characteristics of the signal in different time or frequency ranges.

[0057] Optionally, the fluctuation characteristics of each phase voltage in the three-phase voltage signal at different scales are calculated respectively to obtain multi-fractal characteristics. The multi-fractal characteristics describe the multi-fractal characteristics of the signal at each scale and reflect the multi-scale behavior of the signal.

[0058] Optionally, the accumulated deviations can be extracted from the three-phase voltage signal, and the time series to be processed can be constructed based on the accumulated deviations, and the time series to be processed can be fitted to obtain a fitting polynomial. By fitting the polynomial, some trends or low-frequency components in the signal can be eliminated, so as to focus on analyzing the local fluctuation characteristics of the signal.

[0059] Step 106 , extracting features based on the multifractal features to obtain a fault feature vector; the fault feature vector includes the multifractal spectrum width and fractal dimension difference of each phase voltage.

[0060] Among them, the multifractal spectrum width reflects the complexity variation range of the three-phase voltage signal at different scales. The larger the spectrum width, the higher the complexity of the signal; the fractal dimension difference indicates the difference in geometric complexity of the signal in different regions. The larger the dimension difference, the more significant the local change of the signal.

[0061] Optionally, the singularity index and the fractal dimension may be calculated based on the multifractal feature, and then the multifractal spectrum width and the fractal dimension difference may be calculated based on the singularity index and the fractal dimension.

[0062] Step S108: Identify the fault feature vector to obtain the fault type.

[0063] Optionally, the fault feature vector may be input into a pre-trained recognition model to identify different types of faults.

[0064] Optionally, the fault type may be a ground fault, a short circuit fault, an unbalanced load, etc. Different fault types are manifested in different fractal characteristics.

[0065] Exemplarily, the multifractal spectrum width and fractal dimension difference corresponding to each phase voltage of phase A, phase B and phase C, a total of 6 features, are input into a pre-trained recognition model to obtain the fault type.

[0066] In the above fault identification method, multi-fractal features are obtained by performing multi-scale analysis on the three-phase voltage signal, and then feature extraction is performed based on the multi-fractal features to obtain a fault feature vector. Finally, the fault feature vector is identified to obtain the fault type, so that the fault type can be accurately identified.

[0067] In one of the embodiments, the above-mentioned multi-scale analysis of the three-phase voltage signal to obtain multiple fractal features includes: calculating the cumulative deviation of the three-phase voltage signal, and obtaining a time series to be processed based on the cumulative deviation; fitting the time series to be processed to obtain a fitting polynomial, and calculating local fluctuation characteristics of different scales based on the fitting polynomial; and constructing multiple fractal features based on the local fluctuation characteristics of different scales.

[0068] Optionally, the cumulative deviation of each phase voltage is calculated, and the to-be-processed time series corresponding to each phase voltage is obtained according to the cumulative deviation.

[0069] Furthermore, the time series {X1, ...X n}Intercept the sampling data of 240 points before and after the fault, is the mean of the time series. Then the cumulative deviation is calculated based on the time series and the mean of the time series.

[0070] Exemplarily, the time series data of phase A, phase B and phase C are obtained respectively, the mean of the time series of each phase is calculated, and then the cumulative deviation corresponding to each phase is calculated based on the time series mean expected to correspond to the time series data of each phase.

[0071] Exemplarily, in combination with formula (1), the cumulative deviation of a time series from a time series mean can be calculated according to the formula.

[0072]

[0073] Calculate the cumulative deviation of each data in the time series from the average value and obtain {Y1..., Y i}, which is the time series to be processed.

[0074] Furthermore, in one of the embodiments, the above-mentioned fitting of the time series to be processed to obtain a fitting polynomial, and calculating the local fluctuation characteristics of different scales based on the fitting polynomial, includes: dividing the time series to be processed to obtain multiple subsequences; fitting each subsequence with a second-order polynomial to obtain multiple fitting polynomials; calculating the deviation between each fitting polynomial and each subsequence to obtain local fluctuation characteristics of different scales; calculating the fluctuation function corresponding to each of the local fluctuations at different orders to obtain the multiple fractal characteristics.

[0075] The cumulative deviation Y(i) of the constructed time series from the mean is divided into N k In order to ensure that the data length can be fully utilized when k cannot be divided by N, the data is divided from the positive direction and the reverse direction. First, the accumulated deviation is divided into N non-overlapping k segment, and then divide it from the negative direction into N non-overlapping k A total of 2N segments are divided k Segment data.

[0076] Optionally, the time series to be processed is divided into multiple non-overlapping subsequences, and a polynomial is fitted for each subsequence.

[0077] For example, the subsequence can be calculated according to formula (2):

[0078] N k =int(240 / k) Formula (2)

[0079] Among them, the value of k is {4,8,16,32,64,128}

[0080] Optionally, a second-order polynomial is fitted using the least squares method for each subsequence. v (j) is the least squares 2nd order fitting polynomial of the vth subsequence.

[0081] y v (j) = a0 + a1j + a2j 2 Formula (3)

[0082] The time series to be processed corresponding to each phase is fitted by the least square method and a second-order polynomial to fit the polynomial of each subsequence.

[0083] Finally, the deviation between the polynomial and each sub-series is calculated by local root mean square to obtain the local fluctuation characteristics, which reflect the fluctuation amplitude and change of the signal.

[0084] By analyzing the fluctuations of signals at different scales, we can understand the multi-scale structure and potential fractal properties of the signal.

[0085] Exemplarily, the local fluctuation characteristics are calculated according to formula (4).

[0086]

[0087] Finally, multifractal features are constructed based on local fluctuation characteristics.

[0088] Optionally, the local fluctuation characteristics of all subsequences are averaged to calculate the q-order fluctuation function F q (k).

[0089] Exemplarily, the multifractal feature can be calculated according to formula (5).

[0090]

[0091] Among them, q is a parameter with different orders.

[0092] In the above embodiment, through this multi-scale analysis method, the multi-fractal features in the three-phase voltage signal can be extracted more accurately and comprehensively, providing strong support for subsequent fault identification.

[0093] In one of the embodiments, the feature extraction based on multi-fractal features to obtain the fault feature vector includes: performing logarithmic fitting on the wave function to obtain the Hurst index; obtaining the quality index value according to the Hurst index and the order; and calculating the multi-fractal spectrum width and the fractal dimension difference according to the quality index value.

[0094] Optionally, a logarithmic fit is performed on the relationship between the fluctuation function and the scale K to obtain the Hurst index. The Hurst index describes the scale dependence of the signal at different q values, and the specific calculation formula is shown in formula (6).

[0095]

[0096] Optionally, based on the Hurst index, the quality index value can be further calculated. The specific calculation formula can be shown in formula (7).

[0097] τ q =qH q -1 Formula (7)

[0098] Furthermore, in one of the embodiments, the above-mentioned calculation of the multifractal spectrum width and the fractal dimension difference according to the quality index value includes: calculating the first singular index, the second singular index, the first fractal dimension and the second fractal dimension according to the quality index value; calculating the difference between the first singular index and the second singular index to obtain the multifractal spectrum width; and calculating the fractal dimension difference according to the first fractal dimension and the second fractal dimension.

[0099] Optionally, the first singular index may be a maximum singular index, and the second singular numerical index may be a minimum singular index.

[0100] Optionally, the singular index is obtained by differentiating the quality index. Exemplarily, the singular index can be calculated by combining formula (8).

[0101]

[0102] Optionally, the first fractal dimension may be a maximum fractal dimension, and the second fractal dimension may be a minimum fractal dimension.

[0103] Optionally, the fractal dimension is calculated based on the singular index and the quality index value. Exemplarily, the fractal dimension can be calculated in combination with formula (9).

[0104] D=qα-τ q Formula (9)

[0105] Furthermore, the first fractal dimension is the maximum value of the fractal dimension, and the second fractal dimension is the minimum value of the fractal dimension.

[0106] It should be noted that the fractal dimension and the singular index value constitute a multifractal spectrum. Figure 3 As shown, Figure 3 Schematic diagram of a multifractal spectrum in one embodiment.

[0107] Optionally, the difference between the first singular index and the second singular index is calculated to obtain the multifractal spectrum width. Exemplarily, the multifractal spectrum width is calculated in combination with formula 10.

[0108] Δ α =α max -α min Formula (10)

[0109] Optionally, a fractal dimension difference is calculated based on the first fractal dimension and the second fractal dimension. Exemplarily, the fractal dimension difference is calculated in combination with formula 11.

[0110] Δ D =D max -D min Formula (11)

[0111] Exemplarily, the multi-fractal spectrum width and the maximum / minimum fractal dimension difference of the three-phase voltage in the aforementioned fault big data are shown in the following table.

[0112] Multifractal spectral width Maximum / minimum fractal dimension difference Phase A 0.64 0.47 Phase B 1.02 0.82 Phase C 4.00 1.76

[0113] In one of the embodiments, the fault feature vector is identified and the fault type is obtained through a pre-trained recognition model; the training process of the recognition model includes: obtaining a sample fault feature vector; the sample fault feature vector carries a type label; the sample fault feature vector is input into the initial model for training to obtain a predicted type; based on the predicted type and the type label, a target loss function is obtained, and the parameters of the initial model are adjusted according to the target loss function until the training is completed to obtain a recognition model.

[0114] Optionally, the multifractal spectrum width of phase A voltage, the maximum / minimum fractal dimension difference of phase A voltage, the multifractal spectrum width of phase B voltage, the maximum / minimum fractal dimension difference of phase B voltage, the multifractal spectrum width of phase C voltage, and the maximum / minimum fractal dimension difference of phase C voltage are used to form the feature vector of the fault and input it into the trained recognition model.

[0115] Optionally, first collect the raw data of different types of faults. Each set of fault data needs to undergo signal processing and feature extraction to generate corresponding feature vectors. These feature vectors are usually multifractal features, such as multifractal spectrum width, fractal dimension difference, etc. For each sample fault, extract its feature vector. For example, for a three-phase voltage signal, a six-dimensional feature vector containing multifractal spectrum width and fractal dimension difference can be generated. Next, annotate the data, and each sample feature vector carries a fault type label, which represents the true fault type of the sample, such as A-phase fault, B-phase fault, C-phase fault, short circuit fault, etc. Finally, select or build an initial model and initialize its parameters, and send the sample fault feature vector as input data to the initial model. The model calculates according to the input feature vector and outputs the predicted fault type. The initial model can be any machine learning model or deep learning model. During the training process, the predicted output of the model is compared with the true type label to calculate the target loss function. Use an optimization algorithm, such as a gradient descent algorithm, to minimize the target loss function. The optimization algorithm calculates the gradient of the loss function relative to the model parameters through back propagation and updates the parameters of the model. In this way, the process of feature vector input, loss function calculation and parameter update is repeated until the model converges or reaches the preset number of training rounds to obtain a trained recognition model.

[0116] In the above embodiment, a recognition model that can accurately identify the fault type is obtained through model training, which can achieve more efficient, more accurate and more intelligent fault detection.

[0117] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0118] Based on the same inventive concept, the embodiment of the present application also provides a fault identification device for implementing the fault identification method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more fault identification device embodiments provided below can refer to the limitations on the fault identification method above, and will not be repeated here.

[0119] In an exemplary embodiment, Figure 4 As shown, a fault identification device is provided, comprising: an acquisition module, an analysis module, a feature extraction module and an identification module, wherein:

[0120] The acquisition module 410 is used to acquire three-phase voltage signals.

[0121] The analysis module 420 is used to perform multi-scale analysis on the three-phase voltage signal to obtain multi-fractal features.

[0122] The feature extraction module 430 is used to perform feature extraction based on the multifractal feature to obtain a fault feature vector; the fault feature vector includes the multifractal spectrum width and fractal dimension difference of each phase voltage.

[0123] The identification module 440 is used to identify the fault feature vector and obtain the fault type.

[0124] In one embodiment, the analysis module includes:

[0125] The time series calculation unit is used to calculate the cumulative deviation of the three-phase voltage signal and obtain the time series to be processed according to the cumulative deviation.

[0126] The fitting unit is used to fit the time series to be processed to obtain a fitting polynomial, and calculate the local fluctuation characteristics of different scales according to the fitting polynomial to obtain the multi-fractal characteristics.

[0127] In one embodiment, the fitting unit comprises:

[0128] The division sub-unit is used to divide the time series to be processed into multiple sub-sequences.

[0129] The polynomial fitting subunit is used to perform second-order polynomial fitting on each subsequence to obtain multiple fitting polynomials.

[0130] The feature calculation subunit is used to calculate the deviation between each fitting polynomial and each subsequence to obtain local fluctuation characteristics of different scales.

[0131] The feature calculation subunit is used to calculate the wave functions corresponding to each local wave at different orders to obtain the multi-fractal features.

[0132] In one embodiment, the feature extraction module includes:

[0133] The first exponent calculation unit is used to perform logarithmic fitting on the fluctuation function to obtain the Hurst exponent.

[0134] The second index calculation unit is used to obtain a quality index value according to the Hurst index and the order.

[0135] The multi-fractal feature extraction unit is used to calculate the multi-fractal spectrum width and the fractal dimension difference according to the quality index value.

[0136] In one embodiment, the polymorphic feature extraction unit comprises:

[0137] The fractal dimension calculation unit is used to calculate the first singular index, the second singular index, the first fractal dimension and the second fractal dimension respectively according to the quality index value.

[0138] The difference calculation unit is used to calculate the fractal dimension difference according to the first fractal dimension and the second fractal dimension.

[0139] In one embodiment, the above device further includes a training module, and the training module includes:

[0140] The sample acquisition unit is used to acquire a sample fault feature vector; the sample fault feature vector carries a type label.

[0141] The training unit is used to input the sample fault feature vector into the initial model for training to obtain the prediction type.

[0142] The adjustment unit is used to obtain a target loss function based on the predicted type and the type label, and adjust the parameters of the initial model according to the target loss function until the training is completed to obtain a recognition model.

[0143] Each module in the above fault identification device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0144] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 5 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store three-phase voltage signals. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a fault identification method is implemented.

[0145] Those skilled in the art will understand that Figure 5 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0146] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the following steps when executing the computer program: acquiring a three-phase voltage signal; performing a multi-scale analysis on the three-phase voltage signal to obtain a multi-fractal feature; performing feature extraction based on the multi-fractal feature to obtain a fault feature vector; the fault feature vector includes the multi-fractal spectrum width and fractal dimension difference of each phase voltage; and identifying the fault feature vector to obtain a fault type.

[0147] In one embodiment, when the processor executes the computer program, the following steps are also implemented: calculating the cumulative deviation of the three-phase voltage signal, and obtaining the time series to be processed based on the cumulative deviation; fitting the time series to be processed to obtain a fitting polynomial, and calculating the local fluctuation characteristics of different scales based on the fitting polynomial to obtain multiple fractal features.

[0148] In one embodiment, when the processor executes the computer program, the following steps are also implemented: dividing the time series to be processed to obtain multiple subsequences; fitting each subsequence with a second-order polynomial to obtain multiple fitting polynomials; calculating the deviation between each fitting polynomial and each subsequence to obtain local fluctuation characteristics of different scales; calculating the fluctuation function corresponding to each local fluctuation at different orders to obtain multiple fractal characteristics.

[0149] In one embodiment, when the processor executes the computer program, the following steps are also implemented: logarithmically fitting the wave function to obtain the Hurst index; obtaining the quality index value according to the Hurst index and the order; and calculating the multifractal spectrum width and the fractal dimension difference according to the quality index value.

[0150] In one embodiment, when the processor executes the computer program, the following steps are also implemented: according to the quality index value, the first singular index, the second singular index, the first fractal dimension and the second fractal dimension are respectively calculated; the difference between the first singular index and the second singular index is calculated to obtain the multi-fractal spectrum width; according to the first fractal dimension and the second fractal dimension, the fractal dimension difference is calculated.

[0151] In one embodiment, when the processor executes the computer program, the following steps are also implemented: obtaining a sample fault feature vector; the sample fault feature vector carries a type label; inputting the sample fault feature vector into the initial model for training to obtain a predicted type; obtaining a target loss function based on the predicted type and the type label, and adjusting the parameters of the initial model according to the target loss function until the training is completed to obtain a recognition model.

[0152] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: obtaining a three-phase voltage signal; performing a multi-scale analysis on the three-phase voltage signal to obtain a multi-fractal feature; performing feature extraction based on the multi-fractal feature to obtain a fault feature vector; the fault feature vector includes the multi-fractal spectrum width and the fractal dimension difference of each phase voltage; and identifying the fault feature vector to obtain a fault type.

[0153] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: calculating the cumulative deviation of the three-phase voltage signal, and obtaining the time series to be processed based on the cumulative deviation; fitting the time series to be processed to obtain a fitting polynomial, and calculating the local fluctuation characteristics of different scales based on the fitting polynomial to obtain multiple fractal features.

[0154] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: dividing the time series to be processed to obtain multiple subsequences; fitting each subsequence with a second-order polynomial to obtain multiple fitting polynomials; calculating the deviation between each fitting polynomial and each subsequence to obtain local fluctuation characteristics of different scales; calculating the fluctuation function corresponding to each local fluctuation at different orders to obtain multiple fractal characteristics.

[0155] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: logarithmically fitting the wave function to obtain the Hurst index; obtaining the quality index value according to the Hurst index and the order; and calculating the multifractal spectrum width and the fractal dimension difference according to the quality index value.

[0156] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented: a first singular index, a second singular index, a first fractal dimension and a second fractal dimension are respectively calculated according to the quality index value; a difference between the first singular index and the second singular index is calculated to obtain a multi-fractal spectrum width; and a fractal dimension difference is calculated according to the first fractal dimension and the second fractal dimension.

[0157] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: obtaining a sample fault feature vector; the sample fault feature vector carries a type label; inputting the sample fault feature vector into the initial model for training to obtain a predicted type; obtaining a target loss function based on the predicted type and the type label, and adjusting the parameters of the initial model according to the target loss function until the training is completed to obtain a recognition model.

[0158] In one embodiment, a computer program product is provided, including a computer program, which, when executed by a processor, implements the following steps: acquiring a three-phase voltage signal; performing a multi-scale analysis on the three-phase voltage signal to obtain a multi-fractal feature; performing feature extraction based on the multi-fractal feature to obtain a fault feature vector; the fault feature vector includes a multi-fractal spectrum width and a fractal dimension difference of each phase voltage; and identifying the fault feature vector to obtain a fault type.

[0159] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: calculating the cumulative deviation of the three-phase voltage signal, and obtaining the time series to be processed based on the cumulative deviation; fitting the time series to be processed to obtain a fitting polynomial, and calculating the local fluctuation characteristics of different scales based on the fitting polynomial to obtain multiple fractal features.

[0160] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: dividing the time series to be processed to obtain multiple subsequences; fitting each subsequence with a second-order polynomial to obtain multiple fitting polynomials; calculating the deviation between each fitting polynomial and each subsequence to obtain local fluctuation characteristics of different scales; calculating the fluctuation function corresponding to each local fluctuation at different orders to obtain multiple fractal characteristics.

[0161] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: logarithmically fitting the wave function to obtain the Hurst index; obtaining the quality index value according to the Hurst index and the order; and calculating the multifractal spectrum width and the fractal dimension difference according to the quality index value.

[0162] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented: a first singular index, a second singular index, a first fractal dimension and a second fractal dimension are respectively calculated according to the quality index value; a difference between the first singular index and the second singular index is calculated to obtain a multi-fractal spectrum width; and a fractal dimension difference is calculated according to the first fractal dimension and the second fractal dimension.

[0163] In one embodiment, when the computer program is executed by the processor, the following steps are also implemented: obtaining a sample fault feature vector; the sample fault feature vector carries a type label; inputting the sample fault feature vector into the initial model for training to obtain a predicted type; obtaining a target loss function based on the predicted type and the type label, and adjusting the parameters of the initial model according to the target loss function until the training is completed to obtain a recognition model.

[0164] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0165] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0166] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A fault identification method, characterized in that: The method comprises: Obtain three-phase voltage signals; Performing multi-scale analysis on the three-phase voltage signal to obtain multi-fractal features; Perform feature extraction based on the multifractal feature to obtain a fault feature vector; the fault feature vector includes the multifractal spectrum width and fractal dimension difference of each phase voltage; The fault feature vector is identified to obtain the fault type.

2. The fault identification method according to claim 1, characterized in that: The multi-scale analysis of the three-phase voltage signal to obtain multiple fractal features includes: Calculating the cumulative deviation of the three-phase voltage signal, and obtaining a time series to be processed according to the cumulative deviation; The time series to be processed is fitted to obtain a fitting polynomial, and local fluctuation characteristics of different scales are calculated according to the fitting polynomial to obtain the multifractal feature.

3. The fault identification method according to claim 2, characterized in that: The fitting of the time series to be processed to obtain a fitting polynomial, and calculating the local fluctuation characteristics of different scales according to the fitting polynomial to obtain the multifractal feature, include: Dividing the time series to be processed to obtain multiple subsequences; Performing second-order polynomial fitting on each of the subsequences to obtain a plurality of fitting polynomials; Calculating the deviation between each of the fitting polynomials and each of the subsequences to obtain the local fluctuation characteristics of different scales; The wave functions corresponding to the local waves at different orders are calculated to obtain the multi-fractal characteristics.

4. The fault identification method according to claim 3, characterized in that: The extracting features based on the multi-fractal features to obtain the fault feature vector includes: Performing logarithmic fitting on the fluctuation function to obtain the Hurst index; Obtaining a quality index value according to the Hurst index and the order; The multifractal spectrum width and the fractal dimension difference are calculated according to the quality indicator value.

5. The fault identification method according to claim 4, characterized in that: The calculating the multifractal spectrum width and the fractal dimension difference according to the quality index value comprises: Calculating a first singular index, a second singular index, a first fractal dimension, and a second fractal dimension respectively according to the quality index value; Calculating the difference between the first singular index and the second singular index to obtain the multifractal spectrum width; The fractal dimension difference is obtained by calculation according to the first fractal dimension and the second fractal dimension.

6. The fault identification method according to claim 1, characterized in that: The multi-fractal feature is identified to obtain the fault type through a pre-trained recognition model; the training process of the recognition model includes: Obtaining a sample fault feature vector; the sample fault feature vector carries a type label; Inputting the sample fault feature vector into the initial model for training to obtain a prediction type; A target loss function is obtained based on the predicted type and the type label, and the parameters of the initial model are adjusted according to the target loss function until the training is completed to obtain the recognition model.

7. A fault identification device, characterized in that: The device comprises: An acquisition module, used for acquiring three-phase voltage signals; An analysis module, used for performing multi-scale analysis on the three-phase voltage signal to obtain multi-fractal features; A feature extraction module, used for performing feature extraction based on the multifractal feature to obtain a fault feature vector; the fault feature vector includes the multifractal spectrum width and fractal dimension difference of each phase voltage; The identification module is used to identify the fault feature vector to obtain the fault type.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images