Fault data fingerprint construction method and system based on energy difference

Abstract

The invention relates to the technical field of power grid automation, and provides a fault data fingerprint construction method and system based on energy difference, and the method comprises the steps: collecting the electrical quantity data waveforms of a bus and a line side, selecting fault waveform data from the electrical quantity data waveforms, and generating a first waveform data matrix; performing time domain framing processing on the first waveform matrix, dividing the first waveform matrix into time domain frames with overlapped parts through a window function, performing Fourier transform on each time domain subframe, calculating the amplitude and frequency of data, and realizing conversion from a time domain to a frequency domain of the signal; frequency domain sub-band division is carried out according to the length of each time domain sub-frame signal, frequency domain signals are divided into non-overlapping frequency domain sub-bands, an energy matrix of each frequency domain sub-band is calculated, differential operation is carried out according to the energy matrix to construct a binary feature matrix, coding of fault features is achieved, and fault data fingerprints are formed. According to the invention, the accuracy of identifying the fault type by the device is improved, and the protection device is prevented from refusing to operate or malfunctioning.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid automation, and particularly to a method and system for constructing fault data fingerprints based on energy difference. Background Art

[0002] With the rapid development of the power system, especially the increasing proportion of distributed new energy power systems connected to the grid, the correct detection of transmission line faults is the fundamental guarantee for the stable operation of the power grid. When traditional mechanism-based fault detection methods face faults in a complex power grid structure, fault location often fails to achieve the expected effect due to the diverse, random, and non-linear coupling of fault data characteristics, as well as the existence of a large amount of interference noise.

[0003] For example, in a complex distribution network with a large number of distributed photovoltaic power stations connected, when a transmission line fails, the access of distributed power sources changes the power flow distribution of the traditional power grid. Traditional fault detection methods rely on fixed current, voltage amplitude, and phase relationships to determine the fault type. However, the output power of distributed photovoltaics is affected by random factors such as light intensity and temperature, making the current and voltage data characteristics during faults complex and diverse. At the same time, the widespread application of power electronic devices in distributed new energy systems brings a large amount of harmonic interference, and these harmonics are coupled with fault signals to form non-linear characteristics. Traditional methods are difficult to accurately extract fault characteristics from these complex, random, and non-linearly coupled fault data by artificially setting a single fault criterion, thus unable to accurately determine the fault type. In addition, components such as capacitors and inductors in long-distance transmission lines generate transient processes at the moment of fault, and these transient signals contain a large amount of noise, further interfering with the accurate capture of fault characteristics by traditional detection methods and resulting in misjudgment of the fault type. Summary of the Invention

[0004] The purpose of the present invention is to solve at least one technical problem in the background art and provide a method and system for constructing fault data fingerprints based on energy difference.

[0005] To achieve the above object, the present invention provides a method for constructing fault data fingerprints based on energy difference, including:

[0006] Collect the waveform of electrical quantity data on the busbar and the line side through a protection recording device, select the waveform data of N cycles before and after the fault occurrence moment from the obtained electrical quantity data waveform, and arrange the selected waveforms in the order of phase A, phase B, and phase C from top to bottom to generate a first waveform data matrix;

[0007] Perform data standardization preprocessing on the first waveform data matrix;

[0008] Perform time-domain framing processing on the first waveform matrix. Divide the first waveform matrix into time-domain frames with overlapping parts through a window function, and perform Fourier transform on each time-domain sub-frame to calculate the amplitude and frequency of the data, realizing the conversion of signal time-domain information to frequency-domain information;

[0009] Divide frequency-domain sub-bands according to the signal length of each time-domain sub-frame, divide the frequency-domain signal into non-overlapping frequency-domain sub-bands, calculate the energy matrix of each frequency-domain sub-band, perform differential operation according to the energy matrix to construct a binary feature matrix, realize the encoding of fault features, and form a fault data fingerprint.

[0010] According to one aspect of the present invention, the data standardization preprocessing of the first waveform data matrix is as follows: perform standardization processing on the collected electrical quantity time series s(t), where the electrical quantity time series s(t) is the one-dimensional vector value of each phase electrical waveform in the first waveform matrix, and the standardization formula is:

[0011]

[0012] In the formula, t is the sampling time of the electrical quantity time series s(t).

[0013] According to one aspect of the present invention, the differential operation based on the energy matrix to construct a binary feature matrix is as follows:

[0014]

[0015] In the formula, △P(m, n) refers to the fingerprint bit of the m-th frequency-domain sub-band in the n-th time-domain sub-frame; E(m, n) refers to the energy of the m-th frequency-domain sub-band in the n-th time-domain sub-frame.

[0016] According to one aspect of the present invention, the waveform data sampling frequency of N cycles before and after the fault occurrence time is 8 kHz, and N is taken as 2.

[0017] According to one aspect of the present invention, the window function is a Hamming window.

[0018] To achieve the above object, the present invention also provides a fault data fingerprint construction system based on energy difference, including:

[0019] A first waveform data matrix generation module, which collects the electrical quantity data waveforms of the bus and the line side through a protection recording device, selects the waveform data of N cycles before and after the fault occurrence time from the obtained electrical quantity data waveforms, arranges the selected waveforms from top to bottom according to phase A, phase B, and phase C, and generates a first waveform data matrix;

[0020] A data preprocessing module, which performs data standardization preprocessing on the first waveform data matrix;

[0021] The data information conversion module performs time-domain framing processing on the first waveform matrix, divides the first waveform matrix into time-domain frames with overlapping parts through a window function, and performs Fourier transform on each time-domain sub-frame to calculate the amplitude and frequency of the data, realizing the conversion of signal time-domain information to frequency-domain information;

[0022] The fault data fingerprint formation module divides frequency-domain sub-bands according to the signal length of each time-domain sub-frame, divides the frequency-domain signal into non-overlapping frequency-domain sub-bands, calculates the energy matrix of each frequency-domain sub-band, performs differential operation according to the energy matrix to construct a binary feature matrix, realizes the encoding of fault features, and forms a fault data fingerprint.

[0023] To achieve the above object, the present invention also provides an electronic device, including a processor, a memory, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, the method for constructing a fault data fingerprint based on energy difference as described above is realized.

[0024] To achieve the above object, the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the method for constructing a fault data fingerprint based on energy difference as described above is realized.

[0025] According to the solution of the present invention, the fault data fingerprint constructed by the present invention is two-dimensional energy difference data. Through two digital domain region divisions of one-dimensional signals, the details of fault signals are depicted. Through the binary processing of frequency band energy difference, the encoding of fault feature quantities is realized.

[0026] According to the solution of the present invention, the method for constructing a fault data fingerprint of the present invention encodes fault data by using a mature and reliable fingerprint generation algorithm, generates a two-dimensional characteristic energy difference matrix, performs feature encoding on the signal from the energy distribution characteristics of the signal, and the generated two-dimensional binary encoding diagram (i.e., the fault data fingerprint) is beneficial to the later use of a deep learning network to learn and identify fault types, enriching the fault encoding method of artificial intelligence in the identification of power grid fault types.

[0027] According to the solution of the present invention, by using the one-dimensional time series electrical quantity data of each channel recorded during a fault, a two-dimensional data matrix fingerprint diagram is constructed. Based on this, it is beneficial to use the constructed two-dimensional encoded data as the input data of an artificial intelligence network to realize the effective training of the model. Thereby improving the accuracy of the device in identifying fault types and preventing the protection device from refusing to operate or malfunctioning. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 Schematically shows a flowchart of a method for constructing a fault data fingerprint based on energy difference according to an embodiment of the present invention;

[0029] Figure 2 It is a waveform data diagram of three-phase currents collected when a single-phase grounding fault occurs in a 10 kV distribution line via an arc suppression coil system;

[0030] Figure 3 For the Figure 2 corresponding fault data fingerprint diagram constructed from the three-phase current waveform data by the fault data fingerprint construction method based on energy difference;

[0031] Figure 4 It is a three-phase current data diagram collected when a circuit breaker closes instantaneously and the line is closed to a single-phase grounding fault in a 220 kV high-voltage transmission line;

[0032] Figure 5 For the Figure 4 corresponding fault data fingerprint diagram constructed from the three-phase current waveform data by the fault data fingerprint construction method based on energy difference. Detailed implementation manners

[0033] Now, the content of the present invention will be described with reference to exemplary implementation manners. It should be understood that the described implementation manners are only for enabling those of ordinary skill in the art to better understand and thus implement the content of the present invention, rather than implying any limitation to the scope of the present invention.

[0034] As used herein, the term "comprising" and its variants are to be construed as open-ended terms meaning "including but not limited to". The term "based on" is to be construed as "at least partially based on". The terms "one implementation manner" and "an implementation manner" are to be construed as "at least one implementation manner".

[0035] Figure 1 Schematically shows a flowchart of a fault data fingerprint construction method based on energy difference according to one implementation manner of the present invention. As Figure 1 shown, in this implementation manner, the fault data fingerprint construction method based on energy difference includes:

[0036] Collect electrical quantity data waveforms of the busbar and the line side through a protection recording device. From the obtained electrical quantity data waveforms, starting from the selected fault recording start moment (i.e., the fault moment), select waveform data of the previous and subsequent N cycles each, and arrange the selected waveforms in the order of phase A, phase B, and phase C from top to bottom to generate a first waveform data matrix X, where X = [Ia, Ib, Ic];

[0037] Perform data standardization preprocessing on the first waveform data matrix;

[0038] Perform time-domain framing processing on the first waveform matrix. Divide the first waveform matrix into time-domain frames with overlapping parts through a window function, and perform Fourier transform on each time-domain sub-frame to calculate the amplitude and frequency of the data, realizing the conversion of signal time-domain information to frequency-domain information;

[0039] Divide the frequency-domain sub-bands according to the signal length of each time-domain sub-frame, divide the frequency-domain signal into non-overlapping frequency-domain sub-bands, calculate the energy matrix of each frequency-domain sub-band, and perform differential operation according to the energy matrix to construct a binary feature matrix, realizing the encoding of fault features and forming a two-dimensional binary encoding map (fault data fingerprint).

[0040] In this embodiment, the electrical quantity dimension of the first waveform matrix X can be increased according to the specific actual application scenario, and the types of original data can be appropriately added to enrich the fault feature dimension.

[0041] Further, according to an embodiment of the present invention, the data standardization preprocessing of the first waveform data matrix is as follows: perform standardization processing on the collected electrical quantity time series s(t), where the electrical quantity time series s(t) is the one-dimensional vector value of each-phase electrical waveform in the first waveform matrix, and the standardization formula is:

[0042]

[0043] In the formula, t is the sampling time of the electrical quantity time series s(t).

[0044] Further, according to an embodiment of the present invention, the differential operation is performed according to the energy matrix to construct a binary feature matrix as:

[0045]

[0046] In the formula, △P(m, n) refers to the fingerprint bit of the m-th frequency-domain sub-band in the n-th time-domain sub-frame; E(m, n) refers to the energy of the m-th frequency-domain sub-band in the n-th time-domain sub-frame.

[0047] Further, according to an embodiment of the present invention, the waveform data sampling frequency of N power frequency cycles before and after the fault occurrence time is 8 kHz, and N is taken as 2. The size of N selected needs to be determined according to the sampling frequency of the actual device. Generally, the lower the sampling frequency, the greater the probability of fault feature distortion in the first waveform matrix, and the corresponding number of sampling power frequency cycles N before and after needs to be increased to ensure that the fault features can be included in the first waveform matrix as much as possible.

[0048] Further, according to an embodiment of the present invention, the above time-domain framing number and frequency-domain sub-bands can be flexibly set according to the specific power grid fault type and the device sampling frequency limit, and the selected window function type is Hamming window.

[0049] According to the above solution of the present invention, the fault data fingerprint constructed by the present invention is two-dimensional energy difference data. Through two digital domain region divisions of the one-dimensional signal, the details of the fault signal are depicted, and the fault feature quantity coding is realized through the binarization processing of the frequency band energy difference.

[0050] According to the above solution of the present invention, the fault data fingerprint construction method of the present invention encodes the fault data by using a mature and reliable fingerprint generation algorithm, generates a two-dimensional characteristic energy difference matrix, and performs feature coding on the signal from the energy distribution characteristics of the signal. The generated two-dimensional binary coding map (i.e., the fault data fingerprint) is beneficial for the subsequent use of a deep learning network to learn and identify the fault type, enriching the fault coding method of artificial intelligence in the identification of power grid fault types.

[0051] According to the above solution of the present invention, by using the one-dimensional time series electrical quantity data collected by each channel during the fault, a two-dimensional data matrix fingerprint map is constructed. Based on this, it is beneficial to use the constructed two-dimensional coded data as the input data of the artificial intelligence network to realize the effective training of the model. Thereby, the accuracy of the device in identifying the fault type is improved, and the protection device is prevented from refusing to operate or malfunctioning.

[0052] Furthermore, to achieve the above object, the present invention also provides a fault data fingerprint construction system based on energy difference, including:

[0053] The first waveform data matrix generation module collects the waveform data of the electrical quantities on the busbar and the line side through a protection recording device, selects the waveform data of N cycles before and after the fault occurrence moment from the obtained electrical quantity waveform data, arranges the selected waveforms in the order of phase A, phase B, and phase C from top to bottom, and generates the first waveform data matrix;

[0054] The data preprocessing module performs data standardization preprocessing on the first waveform data matrix;

[0055] The data information conversion module performs time domain framing processing on the first waveform matrix, divides the first waveform matrix into time domain frames with overlapping parts through a window function, and performs Fourier transform on each time domain sub-frame to calculate the amplitude and frequency of the data, realizing the conversion of the signal time domain information to the frequency domain information;

[0056] The fault data fingerprint formation module divides the frequency domain sub-bands according to the signal length of each time domain sub-frame, divides the frequency domain signal into non-overlapping frequency domain sub-bands, calculates the energy matrix of each frequency domain sub-band, performs differential operation according to the energy matrix to construct a binary characteristic matrix, realizes the coding of the fault features, and forms the fault data fingerprint.

[0057] In this implementation, the electrical quantity dimension of the first waveform matrix X can be increased according to the specific practical application scenario, and the types of original data can be appropriately increased to enrich the fault feature dimension.

[0058] Further, according to an embodiment of the present invention, the data standardization preprocessing of the first waveform data matrix is: standardizing the collected electrical quantity time series s(t), the electrical quantity time series s(t) is a one-dimensional vector value of the electrical waveform of each phase in the first waveform matrix, and the standardization formula is:

[0059]

[0060] Where t is the sampling time of the electrical quantity time series s(t).

[0061] Further, according to an embodiment of the present invention, a binary feature matrix is ​​constructed by performing a differential operation on the energy matrix:

[0062]

[0063] Wherein, △P(m,n) refers to the fingerprint bit of the mth frequency domain subband in the nth time domain subframe; E(m,n) refers to the energy of the mth frequency domain subband in the nth time domain subframe.

[0064] Further, according to an embodiment of the present invention, the sampling frequency of the waveform data of N cycles before and after the fault occurs is 8 kHz, and N is 2.

[0065] Further, according to an embodiment of the present invention, the above-mentioned time domain frame number and frequency domain sub-band can be flexibly set according to the specific power grid fault type and the device sampling frequency limit, and the selected window function type is the Hamming window.

[0066] According to the above scheme of the present invention, the fault data fingerprint constructed by the present invention is two-dimensional energy differential data. By dividing the one-dimensional signal into two digital domain areas, the detailed characterization of the fault signal is achieved, and the encoding of the fault feature quantity is achieved through the binarization processing of the frequency band energy difference.

[0067] According to the above scheme of the present invention, the fault data fingerprint construction method of the present invention uses a mature and reliable fingerprint generation algorithm to encode fault data, generate a two-dimensional feature energy difference matrix, and feature encode the signal from the energy distribution characteristics of the signal. The generated two-dimensional binary coding image (i.e., the fault data fingerprint) is conducive to the subsequent use of a deep learning network to learn and identify the fault type, enriching the fault coding method of artificial intelligence in identifying power grid fault types.

[0068] According to the above solution of the present invention, using the one-dimensional time series electrical quantity data of each channel recorded during a fault, a two-dimensional data matrix fingerprint map is constructed. Based on this, it is beneficial to use the constructed two-dimensional encoded data as the input data of the artificial intelligence network to effectively train the model. Thereby improving the accuracy of the device in identifying fault types and preventing the protection device from refusing to operate or malfunctioning.

[0069] Furthermore, to achieve the above object, the present invention also provides an electronic device, including a processor, a memory, and a computer program stored on the memory and executable on the processor. When the computer program is executed by the processor, it implements the method for constructing a fault data fingerprint based on energy difference as described above.

[0070] Furthermore, to achieve the above object, the present invention also provides a computer-readable storage medium with a computer program stored thereon. When the computer program is executed by the processor, it implements the method for constructing a fault data fingerprint based on energy difference as described above.

[0071] To make the object, technical solution, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only the best embodiments of the present invention, only used to explain the present invention, and do not limit the protection scope of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0072] Embodiment 1

[0073] The method for constructing a fault data fingerprint based on energy difference includes:

[0074] Step 1. Collect the waveform of electrical quantity data on the busbar and line side through a protection oscillograph device, select the waveform data of N cycles before and after the fault occurrence moment from the obtained electrical quantity data waveform, and arrange the selected waveforms in the order of phase A, phase B, and phase C from top to bottom to generate the first waveform data matrix;

[0075] Step 2. Perform data standardization preprocessing on the first waveform data matrix;

[0076] Step 3. Perform time-domain framing processing on the first waveform matrix, divide the first waveform matrix into time-domain frames with overlapping parts through a window function, and perform Fourier transform on each time-domain sub-frame to calculate the amplitude and frequency of the data, realizing the conversion of signal time-domain information to frequency-domain information;

[0077] Step 4. Divide the frequency-domain subbands according to the signal length of each time-domain subframe, divide the frequency-domain signal into non-overlapping frequency-domain subbands, calculate the energy matrix of each frequency-domain subband, perform differential operations based on the energy matrix to construct a binary feature matrix, and realize the encoding of fault features to form a fault data fingerprint.

[0078] In this embodiment, the system sampling frequency is 8 kHz, the sampling duration is 0.2 s, which includes the waveform data length of two cycles before and after the fault occurrence moment. According to Step 1, the collected data is arranged from top to bottom in the order of three phases A, B, and C, and an initial current waveform matrix with a dimension of 3×1600 is output;

[0079] Further, according to the data after the normalization preprocessing in Step 2, perform time-domain framing once according to Step 3 to generate a time-domain waveform array with n = 125.

[0080] Optionally, the time-domain framing window function selected in Step 3 is a Hamming window.

[0081] Further, according to Step 4, construct an energy matrix E, and construct a binary feature matrix △P through a differential algorithm to realize the encoding of fault features and form a fault data fingerprint.

[0082] Figure 2 It is a three-phase current waveform data diagram collected when a single-phase grounding fault occurs in a 10 kV distribution line via an arc suppression coil system; Figure 3 It is for Figure 2 the corresponding fault data fingerprint diagram constructed for the three-phase current waveform data;

[0083] Figure 4 It is a three-phase current data diagram collected when a circuit breaker closes instantaneously on a single-phase grounding fault in a 220 kV high-voltage transmission line; Figure 5 It is for Figure 4 the corresponding fault data fingerprint diagram constructed for the three-phase current waveform data;

[0084] It should be noted that: when a single-phase grounding fault occurs in a 10 kV distribution system, due to the resonance compensation effect, there is only a very small capacitive current at the grounding point. In a 220 kV high-voltage transmission line, when the circuit breaker closes instantaneously, high-frequency transient signals will be generated in the line. When closing on a faulty line, the closing current will initially change sinusoidally at nearly 50 Hz, while when closing on a sound line, it will oscillate rapidly at a natural frequency much higher than 50 Hz. Therefore, the selected embodiment encodes the fault waveform using a mature and reliable fingerprint generation algorithm to generate two-dimensional characteristic energy difference data, which is beneficial for later use of a deep learning network to learn and identify fault types, and enriches the fault encoding method of artificial intelligence in power grid fault type identification.

[0085] Those of ordinary skill in the art can realize that the modules and algorithm steps described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. A professional technician can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present invention.

[0086] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described devices and equipment can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0087] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or modules can be electrical, mechanical or other forms.

[0088] The modules described as separate components may or may not be physically separated. The components shown as modules may or may not be physical modules, that is, they can be located in one place, or can be distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the embodiments of the present invention.

[0089] In addition, the various functional modules in the embodiments of the present invention can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0090] When the above functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method for sending / receiving energy-saving signals in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical discs that can store program codes.

[0091] The above description is only a preferred embodiment of the present application and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the present application is not limited to the technical solution formed by the specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the technical solution formed by mutually replacing the above features with the (but not limited to) technical features with similar functions disclosed in the present application.

[0092] It should be understood that the magnitudes of the sequence numbers of the steps in the content and embodiments of the present invention do not absolutely mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

Claims

1. A method for constructing fault data fingerprint based on energy difference, characterized in that: include: The electrical quantity data waveforms on the bus and line sides are collected by the protection recording device, and the waveform data of N cycles before and after the fault occurs are selected from the acquired electrical quantity data waveforms, and the selected waveforms are arranged from top to bottom according to phase A, phase B and phase C to generate a first waveform data matrix; Performing data standardization preprocessing on the first waveform data matrix; Performing time domain frame processing on the first waveform matrix, dividing the first waveform matrix into time domain frames with overlapping parts through a window function, and performing Fourier transform on each time domain subframe to calculate the amplitude and frequency of the data, thereby realizing the conversion of signal time domain information into frequency domain information; The frequency domain subbands are divided according to the signal length of each time domain subframe, and the frequency domain signal is divided into non-overlapping frequency domain subbands. The energy matrix of each frequency domain subband is calculated, and a binary feature matrix is ​​constructed by performing differential operation based on the energy matrix to achieve the encoding of fault features and form a fault data fingerprint.

2. The method for constructing fault data fingerprint based on energy difference according to claim 1, characterized in that: The data standardization preprocessing of the first waveform data matrix is: standardizing the collected electrical quantity time series s(t), where the electrical quantity time series s(t) is a one-dimensional vector value of the electrical waveform of each phase in the first waveform matrix, and the standardization formula is: Where t is the sampling time of the electrical quantity time series s(t).

3. The method for constructing fault data fingerprint based on energy difference according to claim 1, characterized in that: The binary feature matrix constructed by performing differential operation according to the energy matrix is: Wherein, △P(m,n) refers to the fingerprint bit of the mth frequency domain subband in the nth time domain subframe; E(m,n) refers to the energy of the mth frequency domain subband in the nth time domain subframe.

4. The method for constructing fault data fingerprint based on energy difference according to claim 1, characterized in that: The sampling frequency of the waveform data of N cycles before and after the fault occurs is 8 kHz, and N is 2.

5. The method for constructing fault data fingerprint based on energy difference according to any one of claims 1 to 4, characterized in that: The window function is a Hamming window.

6. A fault data fingerprint construction system based on energy difference, characterized in that: include: The first waveform data matrix generation module collects the electrical quantity data waveforms of the bus and line sides through the protection recording device, selects the waveform data of N cycles before and after the fault occurs from the acquired electrical quantity data waveforms, arranges the selected waveforms from top to bottom according to phase A, phase B and phase C, and generates the first waveform data matrix; A data preprocessing module performs data standardization preprocessing on the first waveform data matrix; A data information conversion module, which performs time domain frame processing on the first waveform matrix, divides the first waveform matrix into time domain frames with overlapping parts through a window function, and performs Fourier transform on each time domain subframe, calculates the amplitude and frequency of the data, and realizes the conversion of signal time domain information to frequency domain information; The fault data fingerprint formation module divides the frequency domain sub-bands according to the signal length of each time domain sub-frame, divides the frequency domain signal into non-overlapping frequency domain sub-bands, calculates the energy matrix of each frequency domain sub-band, and constructs a binary feature matrix based on the difference operation of the energy matrix to realize the encoding of the fault features and form the fault data fingerprint.

7. An electronic device, characterized in that The invention comprises a processor, a memory and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the method for constructing a fault data fingerprint based on energy difference as described in any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for constructing a fault data fingerprint based on energy difference according to any one of claims 1 to 5 is implemented.

Images