Method and device for detecting high-resistance grounding fault of power distribution network

Abstract

The invention provides a power distribution network high-resistance grounding fault detection method and device, and the method comprises the steps: carrying out the variational mode decomposition calculation of the transient zero-sequence current of a power distribution network, and obtaining a plurality of IMFs; selecting the IMF with the maximum kurtosis as a characteristic intrinsic mode function; obtaining a Teager energy operator of the characteristic intrinsic mode function, and converting the Teager energy operator into an integral histogram; based on the integral histogram, obtaining a kurtosis judgment coefficient, a Shannon entropy judgment coefficient and a deviation degree judgment coefficient, and further obtaining a comprehensive judgment coefficient; judging whether the comprehensive judgment coefficient is greater than a preset threshold value or not, and if yes, judging that the fault is a high-resistance grounding fault; and otherwise, judging that the switching is capacitor switching or load switching. According to the method, the decomposition accuracy is improved, the problems of mode aliasing and endpoint effect of a traditional method are solved, and the high-resistance grounding fault and capacitor switching or load switching can be accurately recognized through comparison of the comprehensive judgment coefficient and the threshold value.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network fault detection, and in particular to a distribution network high-resistance grounding fault detection method and device. Background Art

[0002] With the continuous development of distribution networks and the increasing complexity of power systems, traditional protection mechanisms face many challenges in dealing with high-resistance grounding faults. High-resistance grounding faults usually occur when the conductor contacts the ground or other media (such as grass, sand, concrete, etc.), resulting in a significant weakening of the conductor current. The current of this type of fault is often only a few hundredths to a dozenths of the normal operating current, which makes the traditional protection scheme based on overcurrent relays, fuses and reclosing switches commonly used in distribution networks appear to be powerless when facing high-resistance grounding faults. Although the high-resistance grounding fault current is relatively weak and does not immediately cause significant damage to the distribution system, its potential dangers cannot be ignored. First, the live wires on the ground may cause electric shock hazards to contact personnel, and in severe cases may endanger personal life; second, due to the arc effect at the contact point, high-resistance grounding faults may also cause safety hazards such as fire. In addition, with the expansion of the scale of the distribution network and the increase in environmental complexity, the frequency of high-resistance grounding faults in the distribution system has a tendency to increase year by year. Therefore, how to effectively detect and isolate high-resistance grounding faults has become an urgent problem to be solved in the design of distribution network protection systems.

[0003] In recent years, although some new protection schemes such as fault detection technology based on signal processing methods and intelligent algorithms based on pattern recognition have been widely studied, the existing methods still face many challenges in practical applications because the current waveform characteristics of high-resistance grounding faults are relatively weak and are greatly affected by the fault environment and medium.

[0004] The following are several existing technologies for fault detection:

[0005] 1. Finite impulse response filtering: By setting appropriate filter parameters in the time domain to obtain the filter results, interharmonics can be filtered out. For example, in related studies, by setting parameters reasonably, specific frequency components in the signal can be effectively removed, thereby highlighting the signal part related to the fault characteristics.

[0006] 2. Mathematical morphology filtering: Use mathematical morphology to process current and estimate features for generating decision tree models. Through morphological processing of current, key features are extracted, and a decision tree model is constructed to achieve classification and judgment of fault types.

[0007] 3. Wavelet transform: Use wavelet transform to decompose the signal into different frequency bands in time and frequency, and use these frequency bands to extract the characteristics of high-resistance grounding faults and detect their occurrence. By performing multi-scale decomposition on the signal, the information of different frequency bands can reflect the fault characteristics of different degrees, which helps to locate and identify high-resistance grounding faults more accurately.

[0008] However, the above prior art still has the following defects:

[0009] Feature extraction: Most existing methods such as fast Fourier transform, mathematical morphology, wavelet transform, and S transform use fixed basis functions, resulting in low feature representation capabilities. These fixed basis functions cannot adapt well to the complex changes of different fault signals and cannot accurately capture fault features. They lack adaptability during the extraction process and are prone to obtaining components without physical meaning. For example, for some complex fault signals, traditional methods may not be able to accurately decompose and extract the true features related to the fault, resulting in the extracted components not reflecting the actual physical process.

[0010] Although the empirical mode decomposition algorithm has adaptive characteristics, its decomposition is prone to modal aliasing and endpoint effects. Modal aliasing causes modes of different frequencies to interfere with each other, affecting the accurate judgment of fault characteristics; endpoint effects introduce errors at both ends of the signal, reducing the accuracy of feature extraction.

[0011] In terms of high-resistance fault detection criteria: Most existing methods are based on obtaining characteristic components to construct network detection methods such as neural networks, support vector machines, and decision trees. Although they have high accuracy, they require a sampling library and must be trained in advance. In practical applications, this increases the complexity and cost of the system and takes a long time to detect. For example, in scenarios where a rapid response to a fault is required, long training and complex sampling library preparation may result in the failure to detect the fault in a timely manner, affecting the safe and stable operation of the power system.

[0012] Finally, existing methods often ignore the essential difference in waveform between high-resistance grounding fault detection and capacitor switching and load switching. High-resistance fault current has intermittent reignition and extinction near the zero crossing point, while capacitor switching and load switching are only instantaneous disturbances, and the current shows an oscillating attenuation trend. This neglect of essential characteristics makes it difficult for existing methods to distinguish high-resistance grounding fault detection from capacitor switching and load switching, and it is easy to make misjudgments, which reduces the accuracy and reliability of fault detection. Summary of the invention

[0013] In view of this, the present invention provides a method and device for detecting a high-resistance grounding fault in a distribution network to solve at least one of the above-mentioned problems.

[0014] In order to achieve the above object, the present invention adopts the following scheme:

[0015] According to a first aspect of the present invention, a flow chart of a method for detecting a high-resistance grounding fault in a distribution network is provided, the method comprising: performing variational modal decomposition calculations on a transient zero-sequence current of the distribution network using an improved variational modal decomposition algorithm with adaptive characteristics to obtain a plurality of intrinsic modal functions; obtaining the kurtosis of each of the intrinsic modal functions, and selecting the eigenmodal function with the largest kurtosis as a characteristic intrinsic modal function; obtaining the Teager energy operator of the characteristic intrinsic modal function, and converting the Teager energy operator into an integral histogram; based on the integral histogram, obtaining a kurtosis determination coefficient, a Shannon entropy determination coefficient, and a deviation determination coefficient; obtaining a comprehensive determination coefficient based on the kurtosis determination coefficient, the Shannon entropy determination coefficient, and the deviation determination coefficient; judging whether the comprehensive determination coefficient is greater than a preset threshold, and if so, determining that it is a high-resistance grounding fault; otherwise, determining that it is a capacitor switching or a load switching.

[0016] As an embodiment of the present invention, the above method uses an improved variational modal decomposition algorithm with adaptive characteristics to perform variational modal decomposition calculation on the transient zero-sequence current of the distribution network to obtain a number of intrinsic modal functions, including: step 1, converting the modal signal into an analytical signal through Hilbert transform, and obtaining a single-sided spectrum, and then converting the spectrum of the modal signal to the baseband region through exponential mixing tuned to the respective estimated center frequencies; step 2, estimating the bandwidth of the modal signal through the Gaussian smoothness of the demodulated signal; step 3, after obtaining the bandwidth of each modal signal, iteratively updating the modal signal and the center frequency through a variational optimization method To minimize the bandwidth, the variational problem in the variational optimization method is transformed into an unconstrained optimization problem by introducing a quadratic penalty term and a Lagrange multiplier, and is solved by an alternating direction multiplier method; step 4, setting an iteration termination condition, stopping the iteration when the condition is met and outputting the modal component; step 5, using the sparrow algorithm to search for the best solution of the variational modal decomposition parameters by taking the sample entropy as the fitness function; step 6, repeating the above steps 1 to 5 until the sample entropy is minimized, outputting the variational modal decomposition parameters at this moment, and using the output parameters to perform variable modal decomposition on the transient zero-sequence current, decomposing the transient zero-sequence current into a number of intrinsic modal functions.

[0017] As an embodiment of the present invention, in the above method, obtaining the kurtosis of each of the intrinsic mode functions, and selecting the eigenmode function with the largest kurtosis as the characteristic intrinsic mode function includes:

[0018] The kurtosis of each of the intrinsic mode functions is calculated using the following formula:

[0019]

[0020] In the above formula: γ, μ and σ are the kurtosis, mean value and standard deviation of the intrinsic mode function signal x(n) respectively, and E is the expected value.

[0021] The kurtosis of each of the intrinsic mode functions is compared, and the eigenmode function with the largest kurtosis is selected as the characteristic intrinsic mode function of the transient zero-sequence current.

[0022] As an embodiment of the present invention, the method of obtaining the Teager energy operator of the characteristic intrinsic mode function and converting the Teager energy operator into an integral histogram includes:

[0023] The Teager energy operator waveform of the characteristic intrinsic mode function is calculated using the following formula:

[0024] ψ[x(n)]=x 2 (n)-x(n+1)x(n-1);

[0025] In the above formula: ψ[x(n)]= is the Teager energy operator, x(n-1), x(n) and x(n+1) are three adjacent sampling points;

[0026] Normalizing the Teager energy operator waveform;

[0027] The normalized Teager energy operator waveform is divided into intervals, the area enclosed by each interval and the X-axis is calculated, and an integral histogram is drawn with the area value of each interval as the height of the histogram.

[0028] As an embodiment of the present invention, in the above method, based on the integral histogram, obtaining the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient includes:

[0029] Calculate the kurtosis of the integral histogram of the Teager energy operator waveform, denoted as the kurtosis determination coefficient k1.

[0030] The Shannon entropy value SE of each interval of the integral histogram is calculated using the following formula: i :

[0031] SE i =-P i (Z l )ln[P i (Z l )];

[0032] In the above formula, P m (Z l ) indicates that the Teager energy operator waveform falls within the interval Z l The probability of falling on Z l The ratio of the area of ​​the histogram to the total area of ​​the histogram;

[0033] The Shannon entropy determination coefficient k2 is obtained by summing the Shannon entropy of each interval of the integral histogram:

[0034] k2=∑SE i ,n=1,2,...,N;

[0035] In the above formula, N is the number of intervals of the integral histogram;

[0036] Calculate the deviation between the maximum value and the mean value of the Teager energy operator waveform integral histogram, and record it as the deviation determination coefficient k3.

[0037] As an embodiment of the present invention, in the above method, the comprehensive determination coefficient is obtained according to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient, including:

[0038] According to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient, the comprehensive determination coefficient is obtained using the following formula:

[0039] k=k2(1 / k1+1 / k3).

[0040] According to a second aspect of the present invention, there is provided a distribution network high-resistance grounding fault detection device, the device comprising: a variational mode decomposition unit, used to perform variational mode decomposition calculation on the transient zero-sequence current of the distribution network by using an improved variational mode decomposition algorithm with adaptive characteristics, and obtain a plurality of intrinsic mode functions; a characteristic mode selection unit, used to obtain the kurtosis of each of the intrinsic mode functions, and select the eigenmode function with the largest kurtosis as the characteristic intrinsic mode function; a histogram acquisition unit, used to obtain the Teager energy operator of the characteristic intrinsic mode function, and convert the Teager energy operator into an integral histogram; a coefficient acquisition unit, used to obtain the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient based on the integral histogram; a comprehensive coefficient acquisition unit, used to obtain the comprehensive determination coefficient according to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient; a fault detection unit, used to judge whether the comprehensive determination coefficient is greater than a preset threshold value, and if so, it is judged as a high-resistance grounding fault; otherwise, it is judged as capacitor switching or load switching.

[0041] As an embodiment of the present invention, the variational modal decomposition unit includes: a conversion module, which is used to convert the modal signal into an analytical signal through Hilbert transform and obtain a single-sided spectrum, and then convert the spectrum of the modal signal to a baseband region through exponential mixing tuned to the respective estimated center frequencies; a bandwidth estimation module, which is used to estimate the bandwidth of the modal signal through the Gaussian smoothness of the demodulated signal; a variational optimization module, which is used to iteratively update the modal signal and the center frequency through a variational optimization method to minimize the bandwidth after obtaining the bandwidth of each modal signal, and the variational problem in the variational optimization method is solved by The quadratic penalty term and Lagrange multiplier are introduced to transform the problem into an unconstrained optimization problem, and the problem is solved by the alternating direction multiplier method. An iteration termination module is used to set the iteration termination condition. When the condition is met, the iteration is stopped and the modal component is output. A parameter optimization module is used to use the sparrow algorithm for optimization, and the sample entropy is used as the fitness function to find the optimal solution of the variational modal decomposition parameters. A parameter output module is used to output the variational modal decomposition parameters at the moment when the sample entropy is the smallest. A variable modal decomposition module is used to use the output parameters to perform variable modal decomposition on the transient zero-sequence current, and decompose the transient zero-sequence current into several intrinsic modal functions.

[0042] As an embodiment of the present invention, the above-mentioned characteristic mode selection unit is specifically used for:

[0043] The kurtosis of each of the intrinsic mode functions is calculated using the following formula:

[0044]

[0045] In the above formula: γ, μ and σ are the kurtosis, mean value and standard deviation of the intrinsic mode function signal x(n) respectively, and E is the expected value.

[0046] The kurtosis of each of the intrinsic mode functions is compared, and the eigenmode function with the largest kurtosis is selected as the characteristic intrinsic mode function of the transient zero-sequence current.

[0047] As an embodiment of the present invention, the histogram acquisition unit includes:

[0048] The waveform calculation module is used to calculate the Teager energy operator waveform of the characteristic intrinsic mode function using the following formula:

[0049] ψ[x(n)]=x 2 (n)-x(n+1)x(n-1);

[0050] In the above formula: ψ[x(n)]= is the Teager energy operator, x(n-1), x(n) and x(n+1) are three adjacent sampling points;

[0051] A normalization processing module, used for normalizing the Teager energy operator waveform diagram;

[0052] The histogram drawing module is used to divide the normalized Teager energy operator waveform into intervals, calculate the area enclosed by each interval and the X-axis, and draw an integral histogram using the area value of each interval as the height of the histogram.

[0053] As an embodiment of the present invention, the coefficient acquisition unit includes:

[0054] The kurtosis coefficient acquisition module is used to calculate the kurtosis of the Teager energy operator waveform integral histogram, which is recorded as the kurtosis determination coefficient k1.

[0055] The Shannon coefficient acquisition module is used to calculate the Shannon entropy value SE of each interval of the integral histogram using the following formula: i :

[0056] SE i =-P i (Z l )ln[P i (Z l )];

[0057] In the above formula, P m (Z l ) indicates that the Teager energy operator waveform falls within the interval Z l The probability of falling on Z l The ratio of the area of ​​the histogram to the total area of ​​the histogram;

[0058] The Shannon entropy determination coefficient k2 is obtained by summing the Shannon entropy of each interval of the integral histogram:

[0059] k2=∑SE i ,n=1,2,...,N

[0060] In the above formula, N is the number of intervals of the integral histogram;

[0061] The deviation coefficient acquisition module is used to calculate the deviation between the maximum value and the mean value of the Teager energy operator waveform integral histogram, which is recorded as the deviation determination coefficient k3.

[0062] As an embodiment of the present invention, the comprehensive coefficient acquisition unit is specifically used for:

[0063] According to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient, the comprehensive determination coefficient is obtained using the following formula:

[0064] k=k2(1 / k1+1 / k3).

[0065] According to a third aspect of the present invention, there is provided an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0066] According to a fourth aspect of the present invention, there is provided 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.

[0067] The method and device for detecting high-resistance grounding faults in distribution networks proposed by the present invention can more accurately obtain the intrinsic mode function of transient zero-sequence current, and select the characteristic intrinsic mode function by kurtosis, thereby improving the credibility and representativeness of the features, overcoming the shortcomings of the existing fixed-base feature extraction method, and eliminating the problems of mode aliasing and endpoint effect of the traditional adaptive decomposition method. In addition, the detection criteria constructed by the present application based on the Teager energy operator, integral histogram and comprehensive criterion have a small amount of calculation, do not require a complex training process, and can quickly and accurately determine the type of fault. It can still work effectively under 1dB strong noise interference, and has a high accuracy rate for judging high-resistance grounding faults, capacitor switching and load switching in distribution networks, effectively solving the problems of long detection time and high misjudgment rate of existing detection methods. Finally, the present application makes full use of the essential difference between the intermittent reignition and extinguishing characteristics of high-resistance grounding faults in distribution networks and capacitor switching and load switching in current waveforms. The constructed detection criteria are highly targeted and can accurately distinguish different fault types, providing a more effective solution for distribution network fault detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0069] Figure 1 It is a flow chart of a method for detecting a high-resistance grounding fault in a distribution network provided in an embodiment of the present application;

[0070] Figure 2 It is a schematic diagram of a flow chart of performing variational modal decomposition calculation on transient zero-sequence current of a distribution network provided in an embodiment of the present application;

[0071] Figure 3 It is a schematic diagram of a process of converting a Teager energy operator into an integral histogram provided by an embodiment of the present invention;

[0072] Figure 4It is a structural schematic diagram of a distribution network high-resistance grounding fault detection device provided in an embodiment of the present application;

[0073] Figure 5 is a schematic diagram of the structure of a variational mode decomposition unit provided in an embodiment of the present application;

[0074] Figure 6 is a schematic diagram of the structure of a histogram acquisition unit provided in an embodiment of the present application;

[0075] Figure 7 is a structural diagram of a coefficient acquisition unit provided in an embodiment of the present application;

[0076] Figure 8 It is a structural schematic diagram of a distribution network simulation model provided in an embodiment of the present application;

[0077] Fig. 9 It is a structural schematic diagram of a high-resistance grounding fault model of a distribution network provided in an embodiment of the present application;

[0078] Fig.10 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0079] To make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings. Here, the exemplary embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0080] like Figure 1 The figure is a flow chart of a method for detecting a high-resistance grounding fault in a distribution network provided in an embodiment of the present application, the method comprising the following steps:

[0081] Step S101: using an improved variational mode decomposition algorithm with adaptive characteristics to perform variational mode decomposition calculation on transient zero-sequence current of the distribution network to obtain a number of intrinsic mode functions (IMFs).

[0082] In this step, the transient zero-sequence current signal of the distribution network is first collected. Then the collected signal can be preprocessed, such as removing the DC component, filtering, etc., to eliminate noise and interference and improve the accuracy of subsequent decomposition. The preprocessing method can be selected according to the actual situation, such as wavelet filtering, mean filtering, etc., which is not limited in this application. The preprocessed signal is then input into the improved variational mode decomposition algorithm for decomposition.

[0083] Preferably, Figure 2 As shown, this step S101 may further include the following sub-steps:

[0084] Step 1: Use Hilbert transform to transform the modal signal uk The modal signal u is converted into an analytical signal and a single-sided spectrum is obtained, and then the modal signal u is mixed with exponentials tuned to the respective estimated center frequencies. k The spectrum is converted to the baseband region.

[0085] The Hilbert transform is shown below:

[0086]

[0087] In the above formula (1): δ(t) is the Dirac function; V k ,ω k are the Kth IMF components and their center frequencies respectively; * is the convolution calculation symbol; j is the imaginary unit; t is the time.

[0088] Step 2: Estimate the modal signal u by using the Gaussian smoothness of the demodulated signal k bandwidth.

[0089] Step 3: After obtaining the bandwidth of each modal signal, the modal signal u is iteratively updated by a variational optimization method. k and the center frequency ω k To minimize the bandwidth, the variational problem in the variational optimization method is transformed into an unconstrained optimization problem by introducing a quadratic penalty term and Lagrange multipliers, and is solved by the alternating direction multiplier method.

[0090] The constrained variational problem is expressed as follows (2) (the squared L2 norm of the gradient):

[0091]

[0092] Where: δ is the Dirac distribution; * represents convolution; {u k} represents the modal function {u1,u2,...,u k},{ω k} represents the center frequency {ω1,ω2,...,ω k}; j is the imaginary unit; t is the time.

[0093] By introducing the quadratic penalty term and Lagrange multiplier, the above constrained variational problem is transformed into an unconstrained optimization problem, which can be expressed by the following formula (3):

[0094]

[0095] In the above formula (3), a is the quadratic penalty factor and λ is the Lagrange multiplier.

[0096] The solution of the above formula (3) is obtained by the alternating direction multiplication method, u k and ω k The solution is expressed as:

[0097]

[0098] Where: and They represent u(t) and u i (t), λ(t), and The Fourier transform of , n is the number of iterations.

[0099] Step 4: Set the iteration termination condition. If the condition is met, the iteration is stopped and the modal components are output.

[0100] The iteration termination condition is determined. If the following equation (6) holds, the iteration is stopped and the modal components are output; if not, the process returns to step 3 to obtain the solution of equation (3) by the alternating direction multiplier method.

[0101]

[0102] In the above formula (6): ε is the judgment accuracy, represents the modal function after the nth iteration, represents the mode function after the n+1th iteration, and K represents the number of mode functions.

[0103] Step 5: Use the sparrow algorithm to find the optimal solution of the variational mode decomposition parameters [K, α] by taking the sample entropy as the fitness function.

[0104] Step 6, repeat the above steps 1 to 5 until the sample entropy is minimum, output the variational modal decomposition parameters [K, α] at this moment, and use the output parameters [K, α] to perform variable modal decomposition on the transient zero-sequence current, decomposing the transient zero-sequence current into several intrinsic mode functions IMF1, IMF2, ..., IMF n .

[0105] Step S102: obtaining the kurtosis of each of the intrinsic mode functions, and selecting the eigenmode function with the largest kurtosis as the characteristic eigenmode function.

[0106] Kurtosis is a statistic that describes the sharpness of data distribution. The larger the kurtosis value, the sharper the data distribution and the stronger the pulse characteristics. The transient zero-sequence current signal of a high-resistance grounding fault usually has a sharp pulse characteristic, so the kurtosis can be used as a basis for feature extraction. This application compares the kurtosis values ​​of all IMFs and selects the IMF with the largest kurtosis as the characteristic intrinsic mode function, that is, this IMF is considered to be the most representative feature of a high-resistance grounding fault.

[0107] Preferably, in step S102, the kurtosis of each of the intrinsic mode functions can be calculated using the following formula:

[0108]

[0109] In the above formula: γ, μ and σ are the kurtosis, mean value and standard deviation of the intrinsic mode function signal x(n) respectively, and E is the expected value.

[0110] Then, the kurtosis of each of the intrinsic mode functions is compared, and the eigenmode function with the largest kurtosis is selected as the characteristic eigenmode function of the transient zero-sequence current.

[0111] Step S103: obtaining the Teager energy operator of the characteristic intrinsic mode function, and converting the Teager energy operator into an integral histogram.

[0112] This step first calculates the Teager energy operator for the characteristic intrinsic mode function selected in step S102. The Teager energy operator is a nonlinear energy operator that can effectively extract the instantaneous energy information of the signal. The calculated Teager energy operator reflects the energy change of the characteristic IMF. Then, the Teager energy operator is converted into an integral histogram, because the integral histogram can more clearly show the distribution characteristics of the Teager energy operator.

[0113] Preferably, Figure 3 As shown, this step may include the following sub-steps:

[0114] Step S1031: Calculate the Teager energy operator waveform of the characteristic intrinsic mode function using the following formula:

[0115] ψ[x(n)]=x 2 (n)-x(n+1)x(n-1);

[0116] In the above formula: ψ[x(n)]= is the Teager energy operator, x(n-1), x(n) and x(n+1) are three adjacent sampling points.

[0117] Step S1032: normalizing the Teager energy operator waveform.

[0118] Step S1033: divide the normalized Teager energy operator waveform into intervals, calculate the area enclosed by each interval and the X-axis, and draw an integral histogram using the area value of each interval as the height of the histogram.

[0119] Step S104: based on the integral histogram, obtaining the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient.

[0120] This step is based on the integral histogram obtained in step S103, and extracts three determination coefficients: the kurtosis determination coefficient, the Shannon entropy determination coefficient, and the deviation determination coefficient. These three coefficients reflect the characteristics of the integral histogram from different angles. For example, the kurtosis determination coefficient reflects the sharpness of the distribution, the Shannon entropy determination coefficient reflects the randomness of the distribution, and the deviation determination coefficient reflects the difference between the distribution and a certain ideal distribution.

[0121] Preferably, this step may specifically include:

[0122] Calculate the kurtosis of the integral histogram of the Teager energy operator waveform, denoted as the kurtosis determination coefficient k1.

[0123] The Shannon entropy value SE of each interval of the integral histogram is calculated using the following formula: i :

[0124] SE i =-P i (Z l )ln[P i (Z l )];

[0125] In the above formula, P m (Z l ) indicates that the Teager energy operator waveform falls within the interval Z l The probability of falling on Z l The ratio of the area of ​​the histogram to the total area of ​​the histogram;

[0126] The Shannon entropy determination coefficient k2 is obtained by summing the Shannon entropy of each interval of the integral histogram:

[0127] k2=∑SE i ,n=1,2,...,N

[0128] In the above formula, N is the number of intervals of the integral histogram;

[0129] Calculate the deviation between the maximum value and the mean value of the Teager energy operator waveform integral histogram, and record it as the deviation determination coefficient k3.

[0130] Through the above steps, the kurtosis determination coefficient k1, the Shannon entropy determination coefficient k2 and the deviation determination coefficient k3 can be obtained.

[0131] Step S105: Obtain a comprehensive determination coefficient according to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient.

[0132] Preferably, in this step, the comprehensive determination coefficient k can be obtained by the following formula:

[0133] k=k2(1 / k1+1 / k3).

[0134] Step S106: Determine whether the comprehensive determination coefficient is greater than a preset threshold value. If so, determine it as a high-resistance grounding fault; otherwise, determine it as capacitor switching or load switching.

[0135] This step compares the comprehensive determination coefficient obtained in step S105 with the preset threshold. The preset threshold needs to be determined based on a large amount of experimental data to ensure a high detection accuracy and a low false positive rate. If the comprehensive determination coefficient is greater than the preset threshold, it is determined to be a high-resistance grounding fault; otherwise, it is determined to be a capacitor switching or load switching.

[0136] As can be seen from the above, the distribution network high-resistance grounding fault detection method proposed by the present invention can more accurately obtain the intrinsic mode function of the transient zero-sequence current, and select the characteristic intrinsic mode function by kurtosis, which improves the credibility and representativeness of the feature, overcomes the shortcomings of the existing fixed-base feature extraction method, and eliminates the modal aliasing and endpoint effect problems of the traditional adaptive decomposition method. In addition, the detection criteria constructed by the present application based on the Teager energy operator, integral histogram and comprehensive criterion have a small amount of calculation, do not require a complex training process, and can quickly and accurately determine the fault type. It can still work effectively under 1dB strong noise interference, and has a high accuracy rate for judging the high-resistance grounding fault, capacitor switching and load switching of the distribution network, effectively solving the problems of long detection time and high misjudgment rate of the existing detection method. Finally, the present application makes full use of the essential difference between the intermittent reignition and extinguishing characteristics of the high-resistance grounding fault of the distribution network and the capacitor switching and load switching on the current waveform. The constructed detection criteria are highly targeted and can accurately distinguish different fault types, providing a more effective solution for distribution network fault detection.

[0137] like Figure 4 The figure shows a schematic diagram of the structure of a distribution network high-resistance grounding fault detection device provided by an embodiment of the present application, the device comprises: a variational mode decomposition unit 410, a characteristic mode selection unit 420, a histogram acquisition unit 430, a coefficient acquisition unit 440, a comprehensive coefficient acquisition unit 450 and a fault detection unit 460, which are sequentially connected. Among them:

[0138] The variational modal decomposition unit 410 is used to perform variational modal decomposition calculation on the transient zero-sequence current of the distribution network by using an improved variational modal decomposition algorithm with adaptive characteristics to obtain a plurality of intrinsic mode functions.

[0139] The characteristic mode selection unit 420 is used to obtain the kurtosis of each of the intrinsic mode functions and select the eigenmode function with the largest kurtosis as the characteristic eigenmode function.

[0140] The histogram acquisition unit 430 is used to acquire the Teager energy operator of the characteristic intrinsic mode function and convert the Teager energy operator into an integral histogram.

[0141] The coefficient acquisition unit 440 is used to acquire the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient based on the integral histogram.

[0142] The comprehensive coefficient acquisition unit 450 is used to obtain a comprehensive determination coefficient according to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient.

[0143] The fault detection unit 460 is used to determine whether the comprehensive determination coefficient is greater than a preset threshold value. If so, it is determined to be a high-resistance grounding fault; otherwise, it is determined to be a capacitor switching or a load switching.

[0144] Preferably, Figure 5 As shown, the variational mode decomposition unit 410 includes:

[0145] The conversion module 411 is used to convert the modal signal into an analytical signal through Hilbert transform and obtain a single-sided spectrum, and then convert the spectrum of the modal signal into a baseband region through exponential mixing tuned to the respective estimated center frequencies.

[0146] The bandwidth estimation module 412 is used to estimate the bandwidth of the modal signal by using the Gaussian smoothness of the demodulated signal.

[0147] The variational optimization module 413 is used to obtain the bandwidth of each modal signal and then iteratively update the modal signal and the center frequency through a variational optimization method to minimize the bandwidth. The variational problem in the variational optimization method is converted into an unconstrained optimization problem by introducing a quadratic penalty term and a Lagrange multiplier, and is solved by an alternating direction multiplier method.

[0148] The iteration termination module 414 is used to set the iteration termination condition. If the condition is met, the iteration is stopped and the modal component is output.

[0149] The parameter optimization module 415 is used to use the sparrow algorithm to perform optimization and use the sample entropy as the fitness function to find the optimal solution of the variational mode decomposition parameters.

[0150] The parameter output module 416 is used to output the variational mode decomposition parameters at the moment when the sample entropy is minimum.

[0151] The variable mode decomposition module 417 is used to perform variable mode decomposition on the transient zero-sequence current using the output parameters, and decompose the transient zero-sequence current into a plurality of intrinsic mode functions.

[0152] Preferably, the characteristic mode selection unit 420 is specifically used for:

[0153] The kurtosis of each of the intrinsic mode functions is calculated using the following formula:

[0154]

[0155] In the above formula: γ, μ and σ are the kurtosis, mean value and standard deviation of the intrinsic mode function signal x(n) respectively, and E is the expected value.

[0156] The kurtosis of each of the intrinsic mode functions is compared, and the eigenmode function with the largest kurtosis is selected as the characteristic intrinsic mode function of the transient zero-sequence current.

[0157] Preferably, Figure 6 As shown, the histogram acquisition unit 430 includes:

[0158] The waveform calculation module 431 is used to calculate the Teager energy operator waveform of the characteristic intrinsic mode function using the following formula:

[0159] ψ[x(n)]=x 2 (n)-x(n+1)x(n-1);

[0160] In the above formula: ψ[x(n)]= is the Teager energy operator, x(n-1), x(n) and x(n+1) are three adjacent sampling points.

[0161] The normalization processing module 432 is used to perform normalization processing on the Teager energy operator waveform diagram.

[0162] The histogram drawing module 433 is used to divide the normalized Teager energy operator waveform into intervals, calculate the area enclosed by each interval and the X-axis, and draw an integral histogram using the area value of each interval as the height of the histogram.

[0163] Preferably, Figure 7 As shown, the coefficient acquisition unit 440 includes:

[0164] The kurtosis coefficient acquisition module 441 is used to calculate the kurtosis of the Teager energy operator waveform integral histogram, which is recorded as the kurtosis determination coefficient k1.

[0165] The Shannon coefficient acquisition module 442 is used to calculate the Shannon entropy value SE of each interval of the integral histogram using the following formula: i :

[0166] SE i =-P i (Z l )ln[P i (Z l )];

[0167] In the above formula, P m (Z l ) indicates that the Teager energy operator waveform falls within the interval Z l The probability of falling on Z l The ratio of the area of ​​the histogram to the total area of ​​the histogram;

[0168] The Shannon entropy determination coefficient k2 is obtained by summing the Shannon entropy of each interval of the integral histogram:

[0169] k2=∑SE i ,n=1,2,...,N

[0170] In the above formula, N is the number of intervals of the integral histogram.

[0171] The deviation coefficient acquisition module 443 is used to calculate the deviation between the maximum value and the mean value of the Teager energy operator waveform integral histogram, which is recorded as the deviation determination coefficient k3.

[0172] Preferably, the comprehensive coefficient acquisition unit 450 is specifically used for:

[0173] According to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient, the comprehensive determination coefficient is obtained using the following formula:

[0174] k=k2(1 / k1+1 / k3).

[0175] For a detailed description of the above-mentioned units and modules, please refer to the corresponding description in the aforementioned method embodiment, which will not be further elaborated here.

[0176] As can be seen from the above, the distribution network high-resistance grounding fault detection device proposed by the present invention can more accurately obtain the intrinsic mode function of the transient zero-sequence current, and select the characteristic intrinsic mode function through the kurtosis, which improves the credibility and representativeness of the feature, overcomes the shortcomings of the existing fixed-base feature extraction method, and eliminates the modal aliasing and endpoint effect problems of the traditional adaptive decomposition method. In addition, the detection criteria constructed by the present application based on the Teager energy operator, integral histogram and comprehensive criterion have a small amount of calculation, do not require a complex training process, and can quickly and accurately determine the type of fault. It can still work effectively under 1dB strong noise interference, and has a high accuracy rate for judging the high-resistance grounding fault, capacitor switching and load switching of the distribution network, effectively solving the problems of long detection time and high misjudgment rate of the existing detection method. Finally, the present application makes full use of the essential difference between the intermittent reignition and extinguishing characteristics of the high-resistance grounding fault of the distribution network and the capacitor switching and load switching on the current waveform. The constructed detection criteria are highly targeted and can accurately distinguish different fault types, providing a more effective solution for distribution network fault detection.

[0177] In order to verify the effectiveness and adaptability of the above-mentioned distribution network high-resistance grounding fault detection method and device, this embodiment builds an IEEE-33 node distribution network simulation model based on the PSCAD / EMTDC simulation platform, as shown in the accompanying figure. Figure 8 shown.

[0178] The high-resistance grounding fault model of distribution network is based on Fig. 9 , the model consists of two DC sources V p 、V n and the corresponding diode D P , D n To form a positive and negative half-cycle current path. Two DC sources V p 、V n The voltage from the simulated arc, whose value depends on the voltage level and asymmetry modeling of the system, varies randomly and independently every 0.1ms. When the instantaneous value V ph >V p When V ph <V n When V n <V ph <V p During this time, no current flows. p and V n The size of R increases the randomness of asymmetric faults and the arc suppression time. To simulate the arc resistance that causes asymmetric current, R P and R h The values ​​are different and change randomly and independently every 0.1ms. The switch capacitance is 5 / 3kVar, the active power of the switch load is 15kW, and the reactive power is 5.5kVar.

[0179] Then, the process of the above-mentioned distribution network high-resistance grounding fault detection method is used to perform fault detection, as follows:

[0180] S1. The transient zero-sequence current of the distribution network is decomposed and calculated using the improved variational modal decomposition algorithm with adaptive characteristics to obtain its intrinsic mode function.

[0181] S2. Calculate the kurtosis of each transient zero-sequence current eigenmode function, and select the eigenmode function with the largest kurtosis value as the characteristic eigenmode function.

[0182] S2.1. Calculate the kurtosis of the intrinsic mode function;

[0183]

[0184] Where: γ, μ and σ are the kurtosis, mean and standard deviation of the signal x(n) respectively.

[0185] S2.2. Compare the kurtosis of each intrinsic mode function and select the eigenmode function with the largest kurtosis as the characteristic intrinsic mode function of the original transient zero-sequence current.

[0186] S3. Calculate the Teager energy operator of the characteristic intrinsic mode function, obtain a waveform diagram of the Teager energy operator of the characteristic intrinsic mode function, and draw its area histogram.

[0187] S3.1. Calculate the Teager energy operator waveform of the characteristic intrinsic mode function. Teager energy operator:

[0188] ψ[x(n)]=x 2 (n)-x(n+1)x(n-1)

[0189] Where: x(n-1), x(n) and x(n+1) are three adjacent sampling points.

[0190] S3.2. Normalize the Teager energy operator waveform.

[0191] S3.3. Divide the normalized Teager energy operator waveform into intervals, calculate the area enclosed by each interval and the X-axis, and draw an integral histogram using the area value of each interval as the height of the histogram.

[0192] Take the normalized Teager energy operator waveform of the first 0.4s simulation time and divide it into eight intervals. The area integral of each interval is calculated as shown in Table 1:

[0193] Table 1

[0194]

[0195] Then, Table 1 is plotted as an integral histogram of the Teager energy operator waveform, as shown in the attached figure. Figure 4 , Figure 5 and Figure 6 shown.

[0196] S4, calculating the kurtosis determination coefficient k1, the Shannon entropy determination coefficient k2 and the variance determination coefficient k3 of the Teager energy operator waveform integral histogram;

[0197] S4.1. Calculate the kurtosis of the integral histogram of the Teager energy operator waveform, denoted as k1.

[0198] S4.2. Calculate the Shannon entropy SE of each interval of the integral histogram i .

[0199] P m (Z l) indicates that the Teager energy operator waveform falls within the interval Z l The probability of falling on Z l The ratio of the area of ​​the histogram to the total area of ​​the histogram.

[0200] SE i =-P i (Z l )ln[P i (Z l )]

[0201] S4.3, sum the Shannon entropy of each interval of the integral histogram to obtain k2;

[0202] k2=∑SE i ,n=1,2,...,N

[0203] Where N is the number of bins in the integral histogram.

[0204] S4.4. Calculate the deviation between the maximum value and the mean value of the integral histogram of the Teager energy operator waveform, denoted as k3.

[0205] S5. Define the comprehensive determination coefficient k using the kurtosis determination coefficient k1, the Shannon entropy determination coefficient k2 and the variance determination coefficient k3, and set the determination threshold k set .

[0206] S5.1. Define the comprehensive determination coefficient as follows:

[0207] k=k2(1 / k1+1 / k3)

[0208] The values ​​of the determination coefficients of the integral histograms under high-resistance grounding faults, capacitor switching, and load switching in the distribution network are shown in Table 2:

[0209] Table 2

[0210]

[0211] S5.2. Setting the decision threshold k set is 0.3, when the comprehensive determination coefficient is greater than k set When the comprehensive determination coefficient is less than k set It is recognized as capacitor switching or load switching.

[0212] As can be seen from Table 2, the integral histogram of the Teager energy operator of the transient zero-sequence current characteristic intrinsic mode function of the distribution network high-resistance grounding fault, capacitor switching and load switching shows significantly different characteristics when calculating the comprehensive determination coefficient, among which the value of the comprehensive determination coefficient of the high-resistance grounding fault is much larger than that of the capacitor switching and the load switching, so it is easy to select the determination threshold. It can be seen that the method proposed in the present invention can accurately determine the high-resistance grounding fault and the capacitor (or load) switching of the distribution network.

[0213] Fig.10 is a schematic diagram of an electronic device provided by an embodiment of the present invention. Fig.10 The electronic device shown is a general data processing device, which includes a general computer hardware structure, which at least includes a processor 801 and a memory 802. The processor 801 and the memory 802 are connected via a bus 803. The memory 802 is suitable for storing one or more instructions or programs executable by the processor 801. The one or more instructions or programs are executed by the processor 801 to implement the steps in the above-mentioned distribution network high-resistance grounding fault detection method.

[0214] The processor 801 may be an independent microprocessor or a collection of one or more microprocessors. Thus, the processor 801 executes the commands stored in the memory 802 to execute the method flow of the embodiment of the present invention as described above to realize the processing of data and the control of other devices. The bus 803 connects the above-mentioned multiple components together, and at the same time connects the above-mentioned components to the display controller 804 and the display device and the input / output (I / O) device 805. The input / output (I / O) device 805 may be a mouse, a keyboard, a modem, a network interface, a touch input device, a somatosensory input device, a printer, and other devices known in the art. Typically, the input / output (I / O) device 805 is connected to the system via an input / output (I / O) controller 806.

[0215] The memory 802 may store software components such as an operating system, a communication module, an interaction module, and an application program. Each of the modules and applications described above corresponds to a set of executable program instructions that implement one or more functions and methods described in the embodiments of the invention.

[0216] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned distribution network high-resistance grounding fault detection method are implemented.

[0217] The method and device for detecting high-resistance grounding faults in distribution networks proposed by the present invention can more accurately obtain the intrinsic mode function of transient zero-sequence current, and select the characteristic intrinsic mode function by kurtosis, thereby improving the credibility and representativeness of the features, overcoming the shortcomings of the existing fixed-base feature extraction method, and eliminating the problems of mode aliasing and endpoint effect of the traditional adaptive decomposition method. In addition, the detection criteria constructed by the present application based on the Teager energy operator, integral histogram and comprehensive criterion have a small amount of calculation, do not require a complex training process, and can quickly and accurately determine the type of fault. It can still work effectively under 1dB strong noise interference, and has a high accuracy rate for judging high-resistance grounding faults, capacitor switching and load switching in distribution networks, effectively solving the problems of long detection time and high misjudgment rate of existing detection methods. Finally, the present application makes full use of the essential difference between the intermittent reignition and extinguishing characteristics of high-resistance grounding faults in distribution networks and capacitor switching and load switching in current waveforms. The constructed detection criteria are highly targeted and can accurately distinguish different fault types, providing a more effective solution for distribution network fault detection.

[0218] The preferred embodiments of the present invention are described above with reference to the accompanying drawings. Many features and advantages of these embodiments are clear from this detailed description, and therefore the claims are intended to cover all such features and advantages of these embodiments that fall within their true spirit and scope. In addition, since many modifications and changes are easily conceivable to those skilled in the art, it is not intended to limit the embodiments of the present invention to the precise structure and operation illustrated and described, but all suitable modifications and equivalents falling within its scope may be covered.

[0219] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0220] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1A device that provides the functions specified in a block or multiple blocks.

[0221] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0222] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0223] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for detecting high-resistance grounding faults in a distribution network, characterized in that: The method comprises: The transient zero-sequence current of the distribution network is decomposed by variational modal decomposition using an improved variational modal decomposition algorithm with adaptive characteristics, and several intrinsic mode functions are obtained. Obtaining the kurtosis of each of the intrinsic mode functions, and selecting the eigenmode function with the largest kurtosis as the characteristic eigenmode function; Obtaining a Teager energy operator of the characteristic intrinsic mode function, and converting the Teager energy operator into an integral histogram; Based on the integral histogram, obtaining the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient; Obtaining a comprehensive determination coefficient according to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient; It is determined whether the comprehensive determination coefficient is greater than a preset threshold value. If so, it is determined to be a high-resistance grounding fault; otherwise, it is determined to be a capacitor switching or a load switching.

2. The method for detecting a high-resistance grounding fault in a distribution network according to claim 1, characterized in that: The improved variational modal decomposition algorithm with adaptive characteristics is used to perform variational modal decomposition calculation on the transient zero-sequence current of the distribution network, and several intrinsic modal functions are obtained, including: Step 1: convert the modal signal into an analytical signal by Hilbert transform and obtain a single-sided spectrum, and then convert the spectrum of the modal signal into a baseband region by exponential mixing tuned to the respective estimated center frequencies; Step 2, estimating the bandwidth of the modal signal by using the Gaussian smoothness of the demodulated signal; Step 3, after obtaining the bandwidth of each modal signal, the modal signal and the center frequency are iteratively updated by a variational optimization method to minimize the bandwidth, and the variational problem in the variational optimization method is converted into an unconstrained optimization problem by introducing a quadratic penalty term and a Lagrange multiplier, and is solved by an alternating direction multiplier method; Step 4: Set the iteration termination condition. If the condition is met, the iteration is stopped and the modal components are output; Step 5: Use the sparrow algorithm to find the optimal solution of the variational mode decomposition parameters by taking the sample entropy as the fitness function; Step 6, repeating steps 1 to 5 until the sample entropy is minimum, outputting the variational modal decomposition parameters at this moment, and using the output parameters to perform variational modal decomposition on the transient zero-sequence current, decomposing the transient zero-sequence current into a number of intrinsic modal functions.

3. The method for detecting high-resistance grounding fault in a distribution network according to claim 1, characterized in that: The step of obtaining the kurtosis of each of the intrinsic mode functions and selecting the eigenmode function with the largest kurtosis as the characteristic intrinsic mode function comprises: The kurtosis of each of the intrinsic mode functions is calculated using the following formula: In the above formula: γ, μ and σ are the kurtosis, mean value and standard deviation of the intrinsic mode function signal x(n), respectively, and E is the expected value; The kurtosis of each of the intrinsic mode functions is compared, and the eigenmode function with the largest kurtosis is selected as the characteristic intrinsic mode function of the transient zero-sequence current.

4. The method for detecting high-resistance grounding fault in a distribution network according to claim 1, characterized in that: The obtaining of the Teager energy operator of the characteristic intrinsic mode function and converting the Teager energy operator into an integral histogram comprises: The Teager energy operator waveform of the characteristic intrinsic mode function is calculated using the following formula: ψ[x(n)]=x 2 (n)-x(n+1)x(n-1)? In the above formula: ψ[x(n)]= is the Teager energy operator, x(n-1), x(n) and x(n+1) are three adjacent sampling points; Normalizing the Teager energy operator waveform; The normalized Teager energy operator waveform is divided into intervals, the area enclosed by each interval and the X-axis is calculated, and an integral histogram is drawn with the area value of each interval as the height of the histogram.

5. The method for detecting high-resistance grounding fault in a distribution network according to claim 1, characterized in that: The step of obtaining the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient based on the integral histogram includes: Calculate the kurtosis of the integral histogram of the Teager energy operator waveform, denoted as the kurtosis determination coefficient k1; The Shannon entropy value SE of each interval of the integral histogram is calculated using the following formula: i : SE i =-P i (FROM l )ln[P i (FROM l )]; In the above formula, P m (Z l ) indicates that the Teager energy operator waveform falls within the interval Z l The probability of falling on Z l The ratio of the area of ​​the histogram to the total area of ​​the histogram; The Shannon entropy determination coefficient k2 is obtained by summing the Shannon entropy of each interval of the integral histogram: k2=∑SE i ,n=1,2,...,N In the above formula, N is the number of intervals of the integral histogram; Calculate the deviation between the maximum value and the mean value of the Teager energy operator waveform integral histogram, and record it as the deviation determination coefficient k3.

6. The method for detecting high-resistance grounding fault in a distribution network according to claim 5, characterized in that: The comprehensive determination coefficient obtained according to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient includes: According to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient, the comprehensive determination coefficient is obtained using the following formula: k=k2(1 / k1+1 / k3).

7. A distribution network high-resistance grounding fault detection device, characterized in that: The device comprises: A variational mode decomposition unit is used to perform variational mode decomposition calculation on the transient zero-sequence current of the distribution network using an improved variational mode decomposition algorithm with adaptive characteristics to obtain a number of intrinsic mode functions; A characteristic mode selection unit, used for obtaining the kurtosis of each of the intrinsic mode functions, and selecting the eigenmode function with the largest kurtosis as the characteristic eigenmode function; A histogram acquisition unit, used for acquiring the Teager energy operator of the characteristic intrinsic mode function, and converting the Teager energy operator into an integral histogram; A coefficient acquisition unit, used for acquiring a kurtosis determination coefficient, a Shannon entropy determination coefficient and a deviation determination coefficient based on the integral histogram; A comprehensive coefficient acquisition unit, used for obtaining a comprehensive determination coefficient according to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient; The fault detection unit is used to determine whether the comprehensive determination coefficient is greater than a preset threshold value. If so, it is determined to be a high-resistance grounding fault; otherwise, it is determined to be a capacitor switching or a load switching.

8. The distribution network high-resistance grounding fault detection device according to claim 7, characterized in that: The variational mode decomposition unit comprises: A conversion module, used to convert the modal signal into an analytical signal by Hilbert transform and obtain a single-sided spectrum, and then convert the spectrum of the modal signal into a baseband region by exponential mixing tuned to the respective estimated center frequencies; A bandwidth estimation module, used for estimating the bandwidth of the modal signal by using the Gaussian smoothness of the demodulated signal; A variational optimization module, which is used to obtain the bandwidth of each modal signal and then iteratively update the modal signal and the center frequency through a variational optimization method to minimize the bandwidth. The variational problem in the variational optimization method is converted into an unconstrained optimization problem by introducing a quadratic penalty term and a Lagrange multiplier, and is solved by an alternating direction multiplier method; Iteration termination module, used to set the iteration termination condition, when the condition is met, the iteration is stopped and the modal component is output; The parameter optimization module is used to optimize using the sparrow algorithm and use sample entropy as the fitness function to find the optimal solution for the variational mode decomposition parameters; A parameter output module, used for outputting the variational mode decomposition parameters at the moment when the sample entropy is minimum; The variable mode decomposition module is used to perform variable mode decomposition on the transient zero-sequence current using the output parameters, and decompose the transient zero-sequence current into a number of intrinsic mode functions.

9. The distribution network high-resistance grounding fault detection device according to claim 7, characterized in that: The characteristic mode selection unit is specifically used for: The kurtosis of each of the intrinsic mode functions is calculated using the following formula: In the above formula: γ, μ and σ are the kurtosis, mean value and standard deviation of the intrinsic mode function signal x(n), respectively, and E is the expected value; The kurtosis of each of the intrinsic mode functions is compared, and the eigenmode function with the largest kurtosis is selected as the characteristic intrinsic mode function of the transient zero-sequence current.

10. The distribution network high-resistance grounding fault detection device according to claim 7, characterized in that: The histogram acquisition unit comprises: The waveform calculation module is used to calculate the Teager energy operator waveform of the characteristic intrinsic mode function using the following formula: ψ[x(n)]=x 2 (n)-x(n+1)x(n-1)? In the above formula: ψ[x(n)]= is the Teager energy operator, x(n-1), x(n) and x(n+1) are three adjacent sampling points; A normalization processing module, used for normalizing the Teager energy operator waveform diagram; The histogram drawing module is used to divide the normalized Teager energy operator waveform into intervals, calculate the area enclosed by each interval and the X-axis, and draw an integral histogram using the area value of each interval as the height of the histogram.

11. The distribution network high-resistance grounding fault detection device according to claim 7, characterized in that: The coefficient acquisition unit comprises: The kurtosis coefficient acquisition module is used to calculate the kurtosis of the integral histogram of the Teager energy operator waveform, which is recorded as the kurtosis determination coefficient k1; The Shannon coefficient acquisition module is used to calculate the Shannon entropy value SE of each interval of the integral histogram using the following formula: i : SE i =-P i (FROM l )ln[P i (FROM l )]; In the above formula, P m (Z l ) indicates that the Teager energy operator waveform falls within the interval Z l The probability of falling on Z l The ratio of the area of ​​the histogram to the total area of ​​the histogram; The Shannon entropy determination coefficient k2 is obtained by summing the Shannon entropy of each interval of the integral histogram: k2=∑SE i ,n=1,2,...,N In the above formula, N is the number of intervals of the integral histogram; The deviation coefficient acquisition module is used to calculate the deviation between the maximum value and the mean value of the Teager energy operator waveform integral histogram, which is recorded as the deviation determination coefficient k3.

12. The distribution network high-resistance grounding fault detection device according to claim 11, characterized in that: The comprehensive coefficient acquisition unit is specifically used for: According to the kurtosis determination coefficient, the Shannon entropy determination coefficient and the deviation determination coefficient, the comprehensive determination coefficient is obtained using the following formula: k=k2(1 / k1+1 / k3).

13. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

14. 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.

Images