A low voltage series arc fault identification and monitoring method for power collection terminal

Abstract

The present invention provides a low-voltage series arc fault identification and monitoring method for a power acquisition terminal, which relates to the technical field of current signal arc fault detection. The method includes: calculating the modulus of the spectrum map matrix according to the spectrum map matrix of the signal; converting the spectrum map matrix from linear units to logarithmic units to obtain the logarithmic form of the modulus of the spectrum map matrix; visualizing the logarithmic form data of the modulus of the spectrum map matrix as an image, and converting the image format into a grayscale image; and calculating the grayscale integral projection in the column direction according to the grayscale image to obtain a feature sequence composed of integral projection values. The present invention establishes a description of high-frequency fault characteristics through time-frequency domain analysis and grayscale integral projection processes, determines the series arc fault through threshold comparison and fault point number statistics, and solves the problem that the high-frequency fault component characteristics in the identification of series arc faults in low-voltage systems are difficult to characterize in the time domain.

Description

Technical Field

[0001] The present invention relates to the technical field of current signal arc fault detection, and in particular to a low-voltage series arc fault identification and monitoring method for a power collection terminal. Background Art

[0002] Low-voltage series arc is one of the common electrical faults in low-voltage power distribution and power consumption systems. However, this type of fault will not cause an increase in line current when it occurs. Moreover, the line current signal when the fault occurs is similar to the line current when some nonlinear loads operate normally, which greatly increases the difficulty of detecting low-voltage series arc faults.

[0003] When a series arc fault occurs in a low-voltage line, the line current signal contains a normal current component, a noise component, and a series arc fault component. The normal current component is a power frequency component, and the noise component and the series arc fault are high-frequency components. It is difficult to effectively distinguish the high-frequency noise component from the high-frequency arc fault component by directly analyzing the non-stationary time-domain current fault waveform. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a low-voltage series arc fault identification and monitoring method for a power collection terminal. Through time-frequency domain analysis and grayscale integral projection process, a description of high-frequency fault characteristics is established, and the series arc fault is determined by threshold comparison and fault point number statistics. This solves the problem that the high-frequency fault component characteristics in the identification of low-voltage system series arc faults are difficult to characterize in the time domain, and the high-frequency fault characteristics cannot be better used for fault identification.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0006] In a first aspect, a low-voltage series arc fault identification and monitoring method for a power collection terminal is provided, the method comprising:

[0007] Collecting current signals;

[0008] Perform short-time transformation on the current signal to obtain the spectrum matrix of the signal;

[0009] According to the spectrogram matrix of the signal, a modulus value of the spectrogram matrix is ​​calculated;

[0010] Convert the spectrogram matrix from linear units to logarithmic units to obtain a logarithmic form of the modulus of the spectrogram matrix;

[0011] Visualize the logarithmic data of the modulus of the spectrum matrix as an image and convert the image format to a grayscale image;

[0012] According to the grayscale image, the grayscale integral projection in the column direction is calculated to obtain a feature sequence composed of integral projection values;

[0013] A grayscale integral projection value threshold is set, all values ​​in the integral projection value sequence are compared with the grayscale integral projection value threshold, and based on the comparison result; the series arc fault current is obtained based on the comparison result.

[0014] Furthermore, the current signal is short-term transformed to obtain a spectrum matrix of the signal, including:

[0015] The current signal is divided into several segments with length W short periods of time and choose the overlap length , for each short period of time The coefficients for each short period are calculated, where It is k The sample value of the current signal in the time domain, is the length of each short period, is the overlap point between short periods, is the index of the time domain sampling point, Indicates taking multiple scales , Represents different frequency components, with values ​​of 1, 2, 4 or 8; Indicates the translation of a short period of time. ; k is the index of the current short period; is the wavelet basis function, and the calculation formula is:

[0016] , It is k A short period of time and pan The wavelet transform coefficients under The local characteristics of the signal under the representation position;

[0017] The wavelet transform coefficients of each short period are combined to form time-frequency coefficients, and the obtained time-frequency coefficients are displayed in matrix form to obtain the spectrum matrix of the signal, where the horizontal axis of the spectrum matrix represents time, the vertical axis represents frequency, and the numerical values ​​in the spectrum matrix represent the corresponding amplitudes.

[0018] Furthermore, the modulus value of the spectrum matrix is ​​calculated to obtain the modulus value of the spectrum matrix, including:

[0019] According to the spectrum matrix, obtain the complex number of each signal element in the frequency domain. The calculation formula of the complex number is ,in, a is the real part, b is the imaginary part;

[0020] According to the complex number of each element in the matrix, by Calculate the modulus of the elements;

[0021] By taking the modulus value of each element, Calculate the modulus of the spectrogram matrix, where is the spectrogram matrix, is the modulus of the spectrogram matrix.

[0022] Furthermore, the spectrogram matrix is ​​converted from linear units to logarithmic units to obtain the logarithmic form of the modulus value of the spectrogram matrix, including:

[0023] The spectrum matrix is ​​passed through Convert to the logarithmic form of the modulus of the spectrum matrix, where is the modulus value corresponding to the element in the spectrum matrix; The data is in logarithmic form.

[0024] Furthermore, the logarithmic data of the modulus of the spectrum matrix is ​​visualized as an image, and the image format is converted into a grayscale image, including:

[0025] Logarithmic data Visualize as an image and through

[0026] Convert the image to grayscale ,in, is a grayscale image, , , Representing images The red, green, and blue channel color values ​​of , and are the standard deviations of the red, green, and blue channels, respectively.

[0027] Further, according to the grayscale image, the grayscale integral projection in the column direction is calculated to obtain a feature sequence composed of integral projection values, including:

[0028] Determine the grayscale image The resolution is r × c ,in, r Indicates line, c Represents a column;

[0029] pass Calculate the Grayscale integral projection of the column, where Indicates The integral projection value of the column, Indicates Line The grayscale value of the column position, The value range is 1 to r , The range of transformation is 1 to c , is the row mean, is the standard deviation of the row.

[0030] Further, a grayscale integral projection value threshold is set, all values ​​in the integral projection value sequence are compared with the grayscale integral projection value threshold, and the series arc fault current is obtained according to the comparison result, including:

[0031] Set the grayscale integral projection value threshold T , compare all the values ​​in the integral projection value sequence with the grayscale integral projection value threshold to obtain the comparison result, where Grayscale integral projection of the column When , it means that the corresponding sampling point is a series arc fault point; when Grayscale integral projection of the column When , it indicates that the corresponding sampling point is not a series arc fault point;

[0032] Calculate the Grayscale integral projection of the column The number of arc fault points at the time of the series arc fault current is obtained, where the number of arc fault points is ,when , then a series arc fault occurs in the current detection unit signal; otherwise, no series arc fault occurs in the current detection unit.

[0033] In a second aspect, a low-voltage series arc fault identification and monitoring system for a power collection terminal includes:

[0034] An acquisition module is used to collect current signals and perform short-time transformation on the current signals to obtain a spectrum matrix of the signals; and calculate a modulus value of the spectrum matrix according to the spectrum matrix of the signals;

[0035] The processing module is used to convert the spectrum map matrix from linear units to logarithmic units to obtain the logarithmic form of the spectrum map matrix modulus value; visualize the logarithmic form data of the spectrum map matrix modulus value as an image, and convert the image format into a grayscale image; calculate the grayscale integral projection in the column direction according to the grayscale image to obtain a feature sequence composed of integral projection values; set a grayscale integral projection value threshold, compare all values ​​in the integral projection value sequence with the grayscale integral projection value threshold, and according to the comparison result; according to the comparison result, obtain the series arc fault current.

[0036] According to a third aspect, a computing device includes:

[0037] one or more processors;

[0038] The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the method described.

[0039] In a fourth aspect, a computer-readable storage medium stores a program, and when the program is executed by a processor, the method described is implemented.

[0040] The above solution of the present invention includes at least the following beneficial effects:

[0041] It is used to solve the problem that the high-frequency fault component characteristics in the detection of series arc faults in low-voltage systems are difficult to characterize in the time domain, and the high-frequency fault characteristics cannot be better used for fault identification. Through the current sequence acquisition mode, short-time transformation time-frequency domain feature description, time-frequency domain feature visualization, grayscale integral projection and threshold comparison method provided by the present invention, it is possible to effectively describe the high-frequency characteristics of series arc faults in low-voltage line currents and achieve efficient and accurate identification of series arc faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flow chart of a method for identifying and monitoring low-voltage series arc faults for a power collection terminal provided by an embodiment of the present invention.

[0043] Figure 2 It is a schematic diagram of a low-voltage series arc fault identification and monitoring system for a power collection terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0044] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in a form and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0045] like Figure 1 As shown, an embodiment of the present invention provides a low-voltage series arc fault identification and monitoring method for a power collection terminal, the method comprising the following steps:

[0046] Step 11, collecting current signals;

[0047] Step 12, performing a short-time transformation on the current signal to obtain a spectrum matrix of the signal;

[0048] Step 13, calculating the modulus value of the spectrum matrix according to the spectrum matrix of the signal;

[0049] Step 14, converting the spectrogram matrix from linear units to logarithmic units to obtain a logarithmic form of the modulus value of the spectrogram matrix;

[0050] Step 15, visualizing the logarithmic data of the modulus value of the spectrum map matrix into an image, and converting the image format into a grayscale image;

[0051] Step 16, calculating the grayscale integral projection in the column direction according to the grayscale image to obtain a feature sequence composed of integral projection values;

[0052] Step 17, setting a grayscale integral projection value threshold, comparing all values ​​in the integral projection value sequence with the grayscale integral projection value threshold, and obtaining the series arc fault current based on the comparison result.

[0053] In an embodiment of the present invention, by processing the current signal, the spectral characteristics of the signal can be analyzed more accurately. By visualizing the logarithmic form of the modulus value of the spectrum diagram matrix, the abnormal pattern in the current signal can be more easily identified. By setting the grayscale integral projection value threshold and comparing the integral projection value sequence with this threshold, the normal current and the fault current can be effectively distinguished, thereby reducing the false alarm rate. Accurate arc fault detection can promptly discover and deal with potential safety hazards, thereby improving the reliability of the entire electrical system. If the arc fault is not discovered and handled in time, it may cause equipment damage. Arc faults may cause safety accidents such as fires. Through the precise detection of this method, the fault current can be cut off in time, thereby enhancing the safety of the operation process. Timely detection and handling of arc faults can prevent energy waste and unnecessary heat generation, thereby improving the energy efficiency of the entire system.

[0054] In a preferred embodiment of the present invention, the above step 11 may include:

[0055] Initialize the particle swarm according to the acquisition parameters, and each particle represents a set of acquisition parameters;

[0056] according to Evaluate the quality of each particle to obtain the evaluation result, where is the power of the signal, is the power of the noise, is the signal-to-noise ratio, , is the maximum signal-to-noise ratio, is the minimum signal-to-noise ratio, is the variance of the signal noise, is the minimum variance of the signal noise, yes Maximum value;

[0057] Update the particle speed and position according to the evaluation results, iterate multiple times until the maximum number of iterations is reached to obtain the optimal acquisition parameters;

[0058] The current signal is collected according to the optimal collection parameters to obtain the current signal.

[0059] In an embodiment of the present invention, by optimizing the acquisition parameters, the quality of the collected current signal can be significantly improved, noise interference can be reduced, and the signal-to-noise ratio can be improved. The optimized acquisition parameters enable the system to process and analyze the current signal more efficiently, thereby improving the performance and accuracy of the overall system. Reasonable sampling frequency and point settings can reduce energy consumption during data acquisition and processing while ensuring signal quality, thereby achieving more environmentally friendly operations. The optimization algorithm can adaptively adjust the search strategy to adapt to different current signal environments and acquisition requirements. Through the automated parameter optimization process, the cost of manual intervention and trial and error can be reduced, the operating process can be simplified, and work efficiency can be improved.

[0060] In a specific embodiment of the present invention, the number of particles is set, and the position and speed of a set of acquisition parameters are randomly initialized according to the acquisition parameters. For each particle in the particle group, the acquisition parameters represented by it are used to simulate or experiment the signal acquisition, calculate the power of the signal and the power of the noise, and then calculate the signal-to-noise ratio. The quality of the particles is evaluated according to the signal-to-noise ratio and the variance of the signal noise, and the individual optimal position of each particle is updated. The optimal positions of all particles are compared to update the global optimal position. Until the preset maximum number of iterations is reached, the global optimal position and its corresponding acquisition parameters are recorded and updated in each iteration. After the iteration is completed, the acquisition parameters corresponding to the global optimal position are the optimal acquisition parameters sought, and the obtained optimal acquisition parameters are used to perform actual current signal acquisition.

[0061] In a preferred embodiment of the present invention, the above step 12 may include:

[0062] The current signal is divided into several segments with length W short periods of time and choose the overlap length , for each short period of time The coefficients for each short period are calculated, where It is k The sample value of the current signal in the time domain, is the length of each short period, is the overlap point between short periods, is the index of the time domain sampling point, Indicates taking multiple scales = , Represents different frequency components, with values ​​of 1, 2, 4 or 8; Indicates the translation of a short period of time. ; k is the index of the current short period; is the wavelet basis function, and the calculation formula is:

[0063] , It is k A short period of time and pan The wavelet transform coefficients under The local characteristics of the signal under the representation position;

[0064] The wavelet transform coefficients of each short period are combined to form time-frequency coefficients, and the obtained time-frequency coefficients are displayed in matrix form to obtain the spectrum matrix of the signal, where the horizontal axis of the spectrum matrix represents time, the vertical axis represents frequency, and the numerical values ​​in the spectrum matrix represent the corresponding amplitudes.

[0065] In the embodiment of the present invention, the wavelet transform provides the ability to analyze signals in both the time domain and the frequency domain. By decomposing the signal at different scales, the wavelet transform can show the characteristics of the signal at different times and frequencies. The wavelet transform can effectively extract the local features of the signal. In the short-time wavelet transform, by dividing the signal into short time periods and calculating the wavelet coefficients of each time period, the frequency components of the signal in different time periods can be accurately identified, and by selecting different wavelet scales, the behavior of the signal in different frequency ranges can be analyzed. The wavelet transform has good denoising performance and provides greater flexibility.

[0066] In a specific embodiment of the present invention, the length of the short period and the overlap length are defined, and according to these parameters, the current signal is divided into several short periods, each of which contains W sample points, and there will be overlaps between adjacent short periods. Select the wavelet basis function and define the set of scales to be analyzed, which correspond to different frequency components. For example, the scale can be 1, 2, 4, 8, set the index of the time domain sampling point, initialize an empty wavelet coefficient array for each short period and each scale, translate the short period, one sample point at a time, and at each translation position, calculate the wavelet transform coefficient using the selected wavelet basis function and the current scale. This usually involves convolving the wavelet basis function with the signal segment and storing the calculated wavelet coefficients in an array. Combine the wavelet transform coefficients calculated for all short periods and all scales to form a time-frequency coefficient matrix, where each row of the matrix corresponds to a specific time window and each column corresponds to a specific scale (or frequency). Use the processed time-frequency coefficient matrix as a spectrum matrix, with the horizontal axis representing time and the vertical axis representing frequency. The values ​​in the matrix represent the signal amplitude at the corresponding time and frequency.

[0067] In a preferred embodiment of the present invention, the above step 13 may include:

[0068] According to the spectrum matrix, obtain the complex number of each signal element in the frequency domain. The calculation formula of the complex number is ,in, a is the real part, b is the imaginary part;

[0069] According to the complex number of each element in the matrix, by Calculate the modulus of the elements;

[0070] By taking the modulus value of each element, Calculate the modulus of the spectrogram matrix, where is the spectrogram matrix, is the modulus of the spectrogram matrix.

[0071] In an embodiment of the present invention, by calculating the modulus of the spectrum matrix, a complex complex signal can be converted into intuitive amplitude information, thereby simplifying the signal analysis process; the modulus matrix directly displays the amplitude of each frequency component in the signal, which helps to accurately identify the characteristic frequency and periodic components in the signal. By calculating the modulus matrix, the dimension and computational complexity of the data can be reduced while retaining key information, thereby improving processing efficiency. The modulus matrix can be easily converted into an image or chart for display, making the spectral characteristics of the signal more intuitive.

[0072] In the embodiment of the present invention, the signal is converted from the time domain to the frequency domain by a certain signal processing method. Each element in the matrix is ​​a complex number, representing the signal amplitude and phase at a specific frequency; in the spectrum matrix, each element can be expressed as

[0073] of the form, in which is the real part, b is the imaginary part. For each complex element , the modulus is calculated by the formula, and the modulus represents the amplitude of the signal at that frequency; after calculating the modulus of each element, a new matrix is ​​constructed, namely the modulus matrix of the spectrum matrix. Each element in this new matrix is ​​the modulus of the corresponding element in the original spectrum matrix.

[0074] In a preferred embodiment of the present invention, the above step 14 may include:

[0075] The spectrum matrix is ​​passed through Convert to the logarithmic form of the modulus of the spectrum matrix, where is the modulus value corresponding to the element in the spectrum matrix; The data is in logarithmic form.

[0076] In an embodiment of the present invention, the modulus values ​​in the spectrum matrix span a large range, and direct processing may lead to difficulties in data representation. Through logarithmic conversion, the dynamic range of the data can be effectively compressed, so that frequency components of different amplitudes are more comparable on a logarithmic scale. On a logarithmic scale, smaller modulus values ​​(i.e., weak signals) will be represented relatively larger, thereby improving the visibility of weak signals. Converting the spectrum matrix to a logarithmic form is more consistent with auditory perception characteristics, and data in logarithmic form may be easier to perform mathematical operations or statistical analysis. When visualizing a spectrum, the use of a logarithmic scale can make the image clearer and easier to interpret when representing a large range of modulus values.

[0077] In a preferred embodiment of the present invention, the above step 15 may include:

[0078] Logarithmic data Visualize as an image and through

[0079] Convert the image to grayscale ,in, is a grayscale image, , , Representing images The red, green, and blue channel color values , ,and are the standard deviations of the red, green, and blue channels, respectively.

[0080] In the embodiment of the present invention, the grayscale image reduces the interference of color, and converting the color image into a grayscale image reduces the dimension of the data, which helps to improve the computing efficiency, especially when processing a large amount of data. The grayscale image is more compatible in a variety of devices and software. Through logarithmic conversion and appropriate grayscale mapping, the visibility of weaker signals in the spectrum diagram can be enhanced while maintaining the details of stronger signals, which helps to comprehensively and accurately analyze the spectral characteristics of the signal.

[0081] In the embodiment of the present invention, the modulus data of the spectrum matrix is ​​logarithmically transformed. get , where 1 is to avoid the zero value problem in logarithmic calculation, Is the data in logarithmic form; The value of is normalized to the range of [0, 255] to adapt to the image representation; Perform a linear mapping, where is the normalized data, The data is in logarithmic form. is the minimum value of the logarithmic data. It is the maximum value of the data in logarithmic form; it is implemented through the image processing library to map the data value to the pixel value of the image , use the normalized data to create an image, and use the weighted formula to convert the color image Convert to grayscale image .

[0082] In a preferred embodiment of the present invention, the above step 16 may include:

[0083] Determine the grayscale image The resolution is r × c ,in, r Indicates line, c Represents a column;

[0084] pass Calculate the Grayscale integral projection of the column, where Indicates The integral projection value of the column, Indicates Line The grayscale value of the column position, The value range is 1 to r , The range of transformation is 1 to c , is the row mean, is the standard deviation of the row;

[0085] According to the grayscale integral projection in the column direction, a feature sequence consisting of integral projection values ​​is obtained.

[0086] In an embodiment of the present invention, by calculating the grayscale integral projection in the column direction, the two-dimensional image data is converted into a one-dimensional feature sequence, which simplifies the data representation and reduces the complexity of processing and analysis; the grayscale integral projection can capture the structural information and grayscale distribution characteristics of the image in a specific direction. Due to the reduction of data dimension, subsequent image processing and analysis algorithms (such as feature extraction, classifier training) can run more efficiently, reducing the consumption of computing resources and time.

[0087] In the embodiment of the present invention, the grayscale image is determined with a resolution of r × c ,in, r Indicates the number of rows, c Indicates the number of columns, creating a c A one-dimensional array, for each column j (in, j The range of transformation is 1 to c , initialize the column integral projection value If it is 0, iterate through all rows in the column i (in i The range of transformation is 1 to r ), and the gray value of each row corresponding to the position Add to The cumulative result Stored in the one-dimensional array created previously.

[0088] In a preferred embodiment of the present invention, the above step 17 may include:

[0089] Set the grayscale integral projection value threshold T , compare all the values ​​in the integral projection value sequence with the grayscale integral projection value threshold to obtain the comparison result, where Grayscale integral projection of the column When , it means that the corresponding sampling point is a series arc fault point; when Grayscale integral projection of the column When , it indicates that the corresponding sampling point is not a series arc fault point;

[0090] Calculate the Grayscale integral projection of the column The number of arc fault points at the time of the series arc fault current is obtained, where the number of arc fault points is ,when , then a series arc fault occurs in the current detection unit signal; otherwise, no series arc fault occurs in the current detection unit.

[0091] In the embodiment of the present invention, by setting a reasonable threshold, the arc fault point can be identified more accurately, the false alarm and missed alarm can be reduced, and the fault detection can be automatically performed without manual intervention, thereby improving the efficiency and real-time performance of the detection. Timely detection of series arc faults helps prevent potential dangers such as fires. Through automated detection, the frequency and cost of manual inspections can be reduced, while reducing equipment damage and maintenance costs caused by faults.

[0092] In the embodiment of the present invention, based on experimental data, a grayscale integral projection value threshold T is set, and all values ​​in the integral projection value sequence are compared with the threshold, which is expressed as follows: When , it indicates that the sampling point is a series arc fault point; when When , it means that the sampling point is not a series arc fault point. Statistics When is the number of arc fault points .when , it indicates that a series arc fault has occurred in the current detection unit signal. Otherwise, it indicates that no series arc fault has occurred in the current detection unit.

[0093] like Figure 2 As shown, an embodiment of the present invention further provides a low-voltage series arc fault identification and monitoring system 20 for a power collection terminal, comprising:

[0094] The acquisition module 21 is used to collect the current signal; perform short-time transformation on the current signal to obtain a spectrum matrix of the signal; and calculate the modulus value of the spectrum matrix according to the spectrum matrix of the signal;

[0095] The processing module 22 is used to convert the spectrum map matrix from linear units to logarithmic units to obtain the logarithmic form of the spectrum map matrix modulus value; visualize the logarithmic form data of the spectrum map matrix modulus value as an image, and convert the image format into a grayscale image; based on the grayscale image, calculate the grayscale integral projection in the column direction to obtain a feature sequence composed of integral projection values; set a grayscale integral projection value threshold, compare all values ​​in the integral projection value sequence with the grayscale integral projection value threshold, and based on the comparison result; based on the comparison result, obtain the series arc fault current.

[0096] The embodiment of the present invention further provides a computing device, comprising: a processor, a memory storing a computer program, wherein when the computer program is executed by the processor, the method described above is executed. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

[0097] The embodiment of the present invention also provides a computer-readable storage medium storing instructions, which, when executed on a computer, enable the computer to execute the method described above. All implementations in the above method embodiment are applicable to this embodiment and can achieve the same technical effect.

Claims

1. A low-voltage series arc fault identification and monitoring method for a power collection terminal, characterized in that: The method comprises: Collecting current signals; Perform short-time transformation on the current signal to obtain the spectrum matrix of the signal; According to the spectrogram matrix of the signal, a modulus value of the spectrogram matrix is ​​calculated; Convert the spectrogram matrix from linear units to logarithmic units to obtain a logarithmic form of the modulus of the spectrogram matrix; Visualize the logarithmic data of the modulus of the spectrum matrix as an image and convert the image format to a grayscale image; According to the grayscale image, the grayscale integral projection in the column direction is calculated to obtain a feature sequence composed of integral projection values; A grayscale integral projection value threshold is set, all values ​​in the integral projection value sequence are compared with the grayscale integral projection value threshold, and based on the comparison result; the series arc fault current is obtained based on the comparison result.

2. The low-voltage series arc fault identification and monitoring method for a power collection terminal according to claim 1 is characterized in that: Perform a short-time transformation on the current signal to obtain the signal's spectrum matrix, including: The current signal is divided into several segments with length W short periods of time and choose the overlap length , for each short period of time The coefficients for each short period are calculated, where It is k The sample value of the current signal in the time domain, is the length of each short period, is the overlap point between short periods, is the index of the time domain sampling point, Indicates taking multiple scales = , Represents different frequency components, with values ​​of 1, 2, 4 or 8; Indicates the translation of a short period of time. ; k is the index of the current short period; is the wavelet basis function, and the calculation formula is: , It is k A short period of time and pan The wavelet transform coefficients under The local characteristics of the signal under the position; The wavelet transform coefficients of each short period are combined to form time-frequency coefficients, and the obtained time-frequency coefficients are displayed in matrix form to obtain the spectrum matrix of the signal, where the horizontal axis of the spectrum matrix represents time, the vertical axis represents frequency, and the numerical values ​​in the spectrum matrix represent the corresponding amplitudes.

3. According to the low-voltage series arc fault identification and monitoring method for a power collection terminal of claim 2, the modulus value of the spectrum matrix is ​​calculated to obtain the modulus value of the spectrum matrix, comprising: According to the spectrum matrix, obtain the complex number of each signal element in the frequency domain. The calculation formula of the complex number is ,in, a is the real part, b is the imaginary part; According to the complex number of each element in the matrix, by Calculate the modulus of the elements; By taking the modulus value of each element, Calculate the modulus of the spectrogram matrix, where is the spectrogram matrix, is the modulus of the spectrogram matrix.

4. The low-voltage series arc fault identification and monitoring method for a power collection terminal according to claim 3 is characterized in that: Convert the spectrogram matrix from linear units to logarithmic units to obtain the logarithmic form of the modulus of the spectrogram matrix, including: The spectrum matrix is ​​passed through Convert to the logarithmic form of the modulus of the spectrum matrix, where is the modulus value corresponding to the element in the spectrum matrix; The data is in logarithmic form.

5. The low-voltage series arc fault identification and monitoring method for a power collection terminal according to claim 4 is characterized in that: Visualize the logarithmic data of the spectrogram matrix modulus as an image and convert the image format to a grayscale image, including: Logarithmic data Visualize as an image and through Convert the image to grayscale ,in, is a grayscale image, , , Representing images The red, green, and blue channel color values ​​of , and are the standard deviations of the red, green, and blue channels, respectively.

6. The low-voltage series arc fault identification and monitoring method for a power collection terminal according to claim 5 is characterized in that: According to the grayscale image, the grayscale integral projection in the column direction is calculated to obtain a feature sequence composed of integral projection values, including: Determine the grayscale image The resolution is r × c ,in, r Indicates line, c Represents a column; pass Calculate the Grayscale integral projection of the column, where Indicates The integral projection value of the column, Indicates Line The grayscale value of the column position, The value range is 1 to r , The range of transformation is 1 to c , is the row mean, is the standard deviation of the row; According to the grayscale integral projection in the column direction, a feature sequence consisting of integral projection values ​​is obtained.

7. The low-voltage series arc fault identification and monitoring method for a power collection terminal according to claim 6 is characterized in that: A grayscale integral projection value threshold is set, all values ​​in the integral projection value sequence are compared with the grayscale integral projection value threshold, and the series arc fault current is obtained according to the comparison result, including: Set the grayscale integral projection value threshold T , compare all the values ​​in the integral projection value sequence with the grayscale integral projection value threshold to obtain the comparison result, where Grayscale integral projection of the column When , it means that the corresponding sampling point is a series arc fault point; when Grayscale integral projection of the column When , it indicates that the corresponding sampling point is not a series arc fault point; Calculate the Grayscale integral projection of the column The number of arc fault points at the time of the series arc fault current is obtained, where the number of arc fault points is ,when , then a series arc fault occurs in the current detection unit signal; otherwise, no series arc fault occurs in the current detection unit.

8. A low-voltage series arc fault identification and monitoring system for a power collection terminal, characterized in that: Applied to the method according to any one of claims 1 to 7, comprising: An acquisition module is used to collect current signals and perform short-time transformation on the current signals to obtain a spectrum matrix of the signals; and calculate a modulus value of the spectrum matrix according to the spectrum matrix of the signals; The processing module is used to convert the spectrum matrix from linear units to logarithmic units to obtain the logarithmic form of the spectrum matrix modulus value; visualize the logarithmic form data of the spectrum matrix modulus value as an image, and convert the image format into a grayscale image; calculate the grayscale integral projection in the column direction according to the grayscale image to obtain a feature sequence composed of integral projection values; set a grayscale integral projection value threshold, compare all values ​​in the integral projection value sequence with the grayscale integral projection value threshold, and according to the comparison result; obtain the series arc fault current according to the comparison result.

9. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, which, when executed by a processor, implements the method according to any one of claims 1 to 7.

Images