A method and system for detecting high-resistivity grounding faults based on multi-band variable data windows

Abstract

A method and system for detecting high-resistance grounding faults based on multi-band variable data windows are disclosed. The method includes: initiating a two-cycle fault data buffer based on the effective value and abrupt change of the zero-sequence voltage; then performing multi-band filtering and decomposition on the fault data; subsequently analyzing the similarity and volt-ampere characteristics of zero-sequence voltage and zero-sequence current data at different frequency bands; performing secondary verification by changing the data window; and finally detecting whether a high-resistance fault has occurred in the line. This invention is applicable to high-resistance fault detection in low-current systems. By performing multi-band analysis of zero-sequence voltage and zero-sequence current data, it more accurately extracts frequency bands with obvious fault characteristics. Combined with variable data window calculation and verification, it avoids misjudgments caused by random fluctuations, demonstrating significant engineering application value.

Description

Technical Field

[0001] This invention belongs to the technical field of power distribution network fault handling methods, and more specifically, relates to a method and system for detecting high-resistance grounding faults based on multi-band variable data windows. Background Technology

[0002] The power distribution network has a complex and variable structure with numerous branch lines, making it susceptible to severe weather such as typhoons and lightning strikes. When a line comes into contact with high-impedance media such as tree branches, grass, or ponds, the fault transition resistance can reach thousands of ohms or even greater, and the fault electrical quantity is weak, with the fault current sometimes being less than 1A. This poses significant challenges to fault signal extraction and identification.

[0003] Currently, the algorithms for grounding fault detection in low-current systems are mainly divided into steady-state component method and transient component method. Among them, the transient component method has better performance in engineering practice. The transient zero-sequence power method, the third and fifth harmonic methods are usually used. The methods for processing electrical quantities usually include finite pulse filtering, wavelet transform, Hilbert-Huang transform, etc.

[0004] Each algorithm has its own advantages and disadvantages. They all have relatively high accuracy in detecting low-resistance grounding faults. However, for high-resistance grounding faults with a transition resistance of 3000 ohms or more, especially under severe noise interference, the detection effect of various algorithms still needs to be improved.

[0005] For example, but not limited to the solution in the prior art (1), the ground fault line selection is based on VMD (Variational Mode Decomposition). Its shortcomings are that VMD is affected by the number of decomposition layers and the penalty factor, which will cause mode mixing problem and affect the transient frequency domain line selection effect. In addition, the algorithm has a large amount of computation. At the same time, it does not consider that the high resistance fault of the resonant grounding system is affected by harmonic interference, and the zero-sequence voltage and zero-sequence current harmonics are opposite in direction, while the non-zero-sequence voltage leads the zero-sequence current by 90°.

[0006] For example, but not limited to the solution in the existing technology (2), ground fault line selection is based on EMD (Empirical Mode Decomposition). Its shortcomings are that EMD has the disadvantages of mode mixing and endpoint effect problems and large computational load, and it is easy to have non-convergence problems when decomposing high impedance fault signals.

[0007] The core challenge of high-resistance grounding faults lies in the weak electrical quantities, susceptibility to noise interference, and the highly variable fault characteristics. Arc-induced high-resistance grounding faults, in particular, are difficult to identify using a single, standardized approach due to the influence of nonlinear arcs and noise. Therefore, technical issues related to signal processing, the sensitivity of detection methods, and reliability of high-resistance grounding fault detection urgently need to be addressed. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this invention provides a method and system for detecting high-impedance grounding faults based on multi-band variable data windows. The method decomposes weak high-impedance signals into multi-band signals to reduce noise interference, while extracting frequency bands with obvious fault characteristics. In addition, a variable data window is used for secondary verification to reduce the probability of false faults. Finally, the method integrates multi-band similarity judgment and current-voltage characteristic results to balance sensitivity and reliability.

[0009] The present invention adopts the following technical solution.

[0010] The first aspect of the present invention provides a method for detecting high-resistivity grounding faults based on multi-band variable data windows, comprising the following steps:

[0011] The zero-sequence voltage and zero-sequence current are set to a cycle fault data buffer based on the effective value or sudden change of the zero-sequence voltage.

[0012] The cached fault data is decomposed by multi-band filtering to obtain zero-sequence voltage and zero-sequence current data in different frequency bands;

[0013] For each frequency band, select a characteristic data window for zero-sequence voltage and zero-sequence current data, calculate the zero-sequence parameter similarity and volt-ampere characteristic index under the same frequency band; and under the same frequency band, change the characteristic data window and calculate the zero-sequence parameter similarity and volt-ampere characteristic index again.

[0014] For each frequency band, a consistency check is performed based on the two calculation results. If the consistency check passes, the zero-sequence parameter similarity and volt-ampere characteristic index of that frequency band are substituted into the criterion. The high-impedance fault is judged according to the interval to which the zero-sequence parameter similarity and volt-ampere characteristic index belong, or the frequency of occurrence.

[0015] Preferably, the cached fault data includes cached fault data with a total length of two cycles: 1.5 cycles before the fault and 0.5 cycles after the fault.

[0016] Preferably, the step of performing multi-band filtering and decomposition on the cached fault data to obtain zero-sequence voltage and zero-sequence current data in different frequency bands includes:

[0017] IIR filters are used to perform multi-frequency filtering decomposition on the buffered fault data of zero-sequence voltage and zero-sequence current to obtain zero-sequence voltage and zero-sequence current data in different frequency bands.

[0018] Based on the obtained zero-sequence voltage and zero-sequence current data for each frequency band, the first-order difference of the zero-sequence voltage is calculated for each frequency band.

[0019] Preferably, the step of selecting a feature data window for zero-sequence voltage and zero-sequence current data for each frequency band, and calculating the zero-sequence parameter similarity and volt-ampere characteristic index under the same frequency band includes:

[0020] For the k-th frequency band, the segment between the minimum point S1 and the maximum point S2 after zero-sequence voltage filtering is selected as the feature data window.

[0021] Based on the selected feature data window, the similarity of zero-sequence voltage and zero-sequence current, as well as the similarity of zero-sequence voltage difference and zero-sequence current under the k-th frequency band, are calculated, and the smaller of the two similarity values ​​is selected as the zero-sequence parameter similarity.

[0022] Calculate the volt-ampere characteristics of zero-sequence voltage and zero-sequence current in the k-th frequency band.

[0023] Preferably, before changing the feature data window, the similarity of the zero-sequence parameter under the k-th frequency band is checked. If it is less than the set threshold set, the check passes, the feature data window is changed, and the similarity of the zero-sequence parameter and the current-voltage characteristic index under the k-th frequency band are calculated again for consistency verification of the two calculation results; otherwise, the similarity of the zero-sequence parameter and the current-voltage characteristic index under the k-th frequency band are no longer used, the frequency band value is updated with k=k+1, and the calculation of the similarity of the zero-sequence parameter and the current-voltage characteristic index under the next frequency band begins.

[0024] Preferably, the step of changing the feature data window and recalculating the zero-sequence parameter similarity and current-voltage characteristic index under the same frequency band includes:

[0025] Set the translation step size ΔS to move the feature data window backward;

[0026] Based on the shifted feature data window, the similarity of zero-sequence voltage and zero-sequence current, the similarity of zero-sequence voltage difference and zero-sequence current, and the volt-ampere characteristic index of zero-sequence voltage and zero-sequence current are calculated again in the k-th frequency band, and then the similarity of zero-sequence parameters after changing the feature data window in the k-th frequency band is obtained.

[0027] Preferably, for each frequency band, the consistency check based on the two calculation results includes:

[0028] If the similarity of the zero-sequence parameters before and after changing the feature data window in the k-th frequency band is less than the set threshold set, then the consistency check is passed. The similarity of the zero-sequence parameters and the current-voltage characteristic index in the k-th frequency band can be used as a criterion to determine whether a high-impedance fault has occurred. If the consistency check is not passed, the similarity of the zero-sequence parameters and the current-voltage characteristic index in the k-th frequency band will no longer be used. The frequency band value will be updated with k=k+1, and the calculation of the similarity of the zero-sequence parameters and the current-voltage characteristic index in the next frequency band will begin.

[0029] Preferably, the step of determining whether a high-resistance fault has occurred based on the similarity of zero-order parameters and the range of current-voltage characteristic indicators, or the frequency of occurrence, includes:

[0030] If, among all the frequency bands K of a certain line, any one frequency band value k appears, and satisfies the zero-sequence parameter similarity P(k) < -0.8 and the volt-ampere characteristic index Q(k) < 0, then this line is a faulty line; otherwise, continue to perform subsequent judgments;

[0031] For this line, count the frequency of occurrence of the state where -0.8 < P(k) < -0.5 and Q(k) < 0. If the frequency is greater than 3, this line is a faulty line; otherwise, this line is a non-faulty line.

[0032] The second aspect of the present invention provides a high-resistance grounding fault detection system based on multi-band variable data windows, which operates according to a high-resistance grounding fault detection method based on multi-band variable data windows described in the first aspect, including:

[0033] A zero-sequence voltage and zero-sequence current fault data caching module for obtaining discrete data sequences of zero-sequence voltage and current after a fault;

[0034] A multi-band filtering module for performing multi-band filtering decomposition on the cached fault data to obtain zero-sequence voltage and zero-sequence current data in different frequency bands;

[0035] A similarity and volt-ampere characteristic analysis module for, for each frequency band, selecting characteristic data windows of zero-sequence voltage and zero-sequence current data, and calculating the zero-sequence parameter similarity and volt-ampere characteristic index in the same frequency band;

[0036] A variable data window calculation module for changing the characteristic data window and recalculating the zero-sequence parameter similarity and volt-ampere characteristic index;

[0037] A line fault fusion judgment module for comprehensively judging whether the line is a faulty line based on the calculation results of similarity and volt-ampere characteristics.

[0038] The third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the computer program, when loaded onto the processor, implements a high-resistance grounding fault detection method based on multi-band variable data windows described in the first aspect.

[0039] The fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program, when executed by a processor, implements a high-resistance grounding fault detection method based on multi-band variable data windows described in the first aspect.

[0040] Compared with the prior art, the beneficial effects of the present invention include at least the following: The present invention provides a method and system for high-impedance grounding fault detection based on multi-band variable data window, which decomposes the high-impedance weak signal into multi-band signals to reduce noise interference, and at the same time compares and extracts the frequency bands with obvious fault characteristics. In addition, a variable data window is used for secondary verification to reduce the probability of fault misjudgment. Finally, the results of multi-band similarity judgment and volt-ampere characteristic are integrated to balance sensitivity and reliability. Attached Figure Description

[0041] Figure 1 This is a flowchart of a high-resistance grounding fault detection method based on multi-band variable data window provided in Embodiment 1 of the present invention;

[0042] Figure 2 This is a schematic diagram of a high-resistance grounding fault detection system based on a multi-band variable data window, according to Embodiment 2 of the present invention;

[0043] Figure 3 This is a waveform diagram of a grounding fault at the site;

[0044] Figure 4 It is a multi-band decomposition diagram of zero-sequence voltage and zero-sequence current waveforms;

[0045] Figure 5 This is the volt-ampere characteristic diagram of the faulty circuit;

[0046] Figure 6 This is the volt-ampere characteristic diagram of a non-faulty circuit. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this invention.

[0048] like Figure 1 As shown, Embodiment 1 of the present invention provides a method for detecting high-resistance grounding faults based on multi-band variable data windows, comprising the following steps:

[0049] Step 1: Real-time cache of zero-sequence voltage, zero-sequence current, and self-generated zero-sequence voltage, and calculate the effective value and mutation amount of zero-sequence voltage. Start the zero-sequence voltage and zero-sequence current two-cycle fault data cache based on the effective value or mutation amount of zero-sequence voltage.

[0050] Preferably, but not limitingly, in step 1, the zero-sequence voltage and current of the line are discretely sampled and cyclically cached. The cache termination logic is activated by the effective value and sudden change of the zero-sequence voltage to obtain cached fault data with a length of two cycles: 1.5 cycles before the fault and 0.5 cycles after the fault. The zero-sequence voltage and current of the line are expressed by the following formula:

[0051] u 0 (n)=u 0_P (n)+u 0_T (n)

[0052] i 0 (n)=i 0_P (n)+i 0_T (n)

[0053] In the formula:

[0054] u 0 (n) The first in the zero-sequence voltage buffer data n +1 sampling point, u 0_P (n) The first power frequency component data of zero-sequence voltage fault n +1 sampling point, u 0_T (n) The first of the transient component data of zero-sequence voltage fault n +1 sampling point; n =1,2,…,2 N -1;

[0055] i 0 (n) The first in the zero-sequence current buffer data n +1 sampling point, i 0_P (n) The first power frequency component data of zero-sequence current fault n +1 sampling point, i 0_T (n) The first of the transient components of zero-sequence current fault data n +1 sampling point;

[0056] N The length of the sampled cycle data.

[0057] It is worth noting that the cyclic buffer stores the zero-sequence voltage and current of the line, including data before and after the fault, ensuring that the transient decomposition of the fault can identify the section with the most obvious ground fault characteristics. The high-resistance ground fault detection method only requires half-cycle data at the beginning of the fault, but the fault initiation is related to the set value. To ensure that as much data as possible is processed, including the fault initiation segment, more data needs to be stored before the initiation. Therefore, it is preferred, but not limited to, obtaining buffered fault data for two cycles: 1.5 cycles before the fault and 0.5 cycles after the fault.

[0058] Step 2: Perform multi-band filtering decomposition on the cached fault data to obtain zero-sequence voltage and zero-sequence current data in different frequency bands.

[0059] Preferably, but not limitingly, step 2 specifically includes:

[0060] Step 2.1: Use an IIR filter (Infinite Impulse Response Filter) to perform multi-frequency filtering decomposition on the buffered fault data of zero-sequence voltage and zero-sequence current, as shown in the following formula:

[0061]

[0062]

[0063] In the formula:

[0064] u 0_iir_k ( n ), i 0_iir_k ( n The zero-sequence voltage and zero-sequence current are respectively in the first... k Filtered data for each frequency band; k =1,2,…, K , K This represents the total number of frequency bands decomposed by the multi-frequency filter.

[0065] u 0( n ), i 0( n ) are the first and second zero-sequence voltage and zero-sequence current buffer data, respectively. n +1 sampling point;

[0066] M k , N k The first k The feedforward order and feedback order for each frequency band;

[0067] iira j_k( n ), iirb i_k ( n ) are respectively the first k The coefficients of each frequency band, i =0,1,2,…, M k , j =0,1,2,…, N k .

[0068] Step 2.2: Based on the multi-frequency filter decomposition results obtained in Step 2.1, calculate the first-order difference of the zero-sequence voltage, expressed by the following formula:

[0069] u 0_iir_diff_k ( n )= u 0_iir_k ( n +1)- u 0_iir_k ( n )

[0070] In the formula:

[0071] u 0_iir_diff_k ( n ) is the zero-sequence voltage in the first... k Differential data across frequency bands.

[0072] Step 3: Perform similarity and volt-ampere characteristic analysis on zero-sequence voltage and zero-sequence current data and zero-sequence voltage differential and zero-sequence current under different frequency bands.

[0073] Preferably, but not limitingly, step 3 specifically includes:

[0074] Step 3.1: Set frequency band values k Initialize, let k =1.

[0075] Step 3.2: For the first k For each frequency band, select the location of the minimum value point after zero-sequence voltage filtering. S 1. Location of the maximum value point after zero-sequence voltage filtering S The segment between 2 is used as the feature data window.

[0076] Step 3.3: Based on the feature data window selected in Step 3.2, calculate the first... k The similarity of zero-sequence voltage and zero-sequence current in each frequency band is expressed by the following formula:

[0077]

[0078] In the formula:

[0079] P 1( k ) is the first k Similarity of zero-sequence voltage and zero-sequence current across different frequency bands.

[0080] Step 3.4: Based on the feature data window selected in Step 3.2, calculate the first... k The similarity of zero-sequence voltage differential and zero-sequence current in each frequency band is expressed by the following formula:

[0081]

[0082] In the formula:

[0083] P 2( k ) for the first k Similarity of zero-sequence voltage difference and zero-sequence current in each frequency band.

[0084] Step 3.5: In the... k In each frequency band, the similarity between zero-sequence voltage and zero-sequence current is compared with the similarity between zero-sequence voltage difference and zero-sequence current. The smaller of these is selected as the zero-sequence parameter similarity, expressed by the following formula:

[0085]

[0086] In the formula:

[0087] P ( k ) for the first k The zero-order parameter similarity in the nth frequency band, i.e., the similarity of the nth frequency band. k Similarity of zero-sequence voltage and zero-sequence current in different frequency bands P 1( k Similarity to zero-sequence voltage differential and zero-sequence current P 2( k The smaller value between )

[0088] Step 3.6: Calculate the... k The zero-sequence voltage and zero-sequence current voltage-current characteristics at each frequency band are expressed by the following formula:

[0089]

[0090] In the formula:

[0091] S 1 represents the location of the minimum value point after zero-sequence voltage filtering. S 2 represents the location of the maximum value point after zero-sequence voltage filtering;

[0092] Q ( k ) for the firstk The volt-ampere characteristics of zero-sequence voltage and zero-sequence current in each frequency band.

[0093] It is worth noting that, as one of the prominent substantive features of this invention, firstly, the initial reverse polarity of the zero-mode voltage and current of the faulty line, if taken as... x The axis is the current. y With voltage as the axis, it is easy to see that the set of points formed by zero-sequence current and voltage rotates clockwise, while non-faulty points rotate counterclockwise. In order to express clockwise and counterclockwise rotations mathematically, we start with the cross product of vectors. A positive result of the phasor cross product represents counterclockwise rotation, and a negative result represents clockwise rotation. To optimize the calculation, we introduce the phasor dot product, thus obtaining the calculation formulas for the volt-ampere characteristics of zero-sequence voltage and zero-sequence current in different frequency bands.

[0094] Step 3.7: Place the first k Zero-order parameter similarity in each frequency band and a set threshold set If the value is less than the set threshold, then... set If the value is not less than the set threshold, the verification passes and proceed to step 4; if the value is not less than the set threshold, the verification continues. set If the verification fails, the process will not continue. k Fault diagnosis is performed using zero-sequence parameter similarity and volt-ampere characteristic indices across multiple frequency bands. k = k +1 Update the frequency band value and return to step 3.2.

[0095] Step 4: Change the data window to perform a secondary verification of the similarity and volt-ampere characteristics of the zero-sequence voltage, zero-sequence current data, zero-sequence voltage difference, and zero-sequence current.

[0096] Preferably, but not limitingly, step 4 specifically includes:

[0097] Step 4.1: Set the translation step size ΔS With translation step size ΔS Shift the feature data window backward; specifically, with S 1+ ΔS renew S 1. With S 2+ ΔS renew S 2. Shift the data window between the minimum and maximum values ​​of the zero-sequence voltage after filtering to the right; further preferably, but not restrictively, the shift step size... ΔS There are 3-5 data points.

[0098] Step 4.2: Based on the shifted feature data window, recalculate the value at the [missing information - likely a specific point or time frame]. k Similarity of zero-sequence voltage and zero-sequence current in different frequency bands P 1_chg ( kSimilarity between zero-sequence voltage differential and zero-sequence current P 2_chg ( k ) and the current-voltage characteristic index of zero-sequence voltage and zero-sequence current Q _chg ( k ), and thus obtain the first k Zero-order parameter similarity after changing the feature data window in each frequency band P _chg ( k The calculation method used is the same as in steps 3.3 to 3.6.

[0099] Step 4.3: Place the first k Zero-order parameter similarity after changing the feature data window in each frequency band P _chg ( k ) and set threshold set In comparison, if it is less than the set threshold set If the value is not less than the set threshold, then proceed to step 5; set Then k = k +1 Update the frequency band value and return to step 3.2.

[0100] Step 5: By combining zero-sequence voltage and zero-sequence current data under multiple different frequency bands and different data window lengths, the minimum similarity of zero-sequence voltage difference and zero-sequence current, and the volt-ampere characteristics, it is determined whether a high-resistance fault has occurred in the line.

[0101] Preferably, but not limitingly, step 5 specifically includes:

[0102] Step 5.1: If the total number of frequency bands of a certain line... K In, any frequency band value appears k Satisfying zero-order parameter similarity P ( k The voltage-current characteristic index of zero-sequence voltage and zero-sequence current is less than -0.8. Q ( k If the value is less than 0, then the line is a faulty line; otherwise, continue to step 5.2.

[0103] Step 5.2: For this line, calculate -0.8 < P ( k )<-0.5 and Q ( k The frequency of the state < 0 is considered. If the frequency is greater than 3, the line is a faulty line; otherwise, the line is a non-faulty line.

[0104] Embodiment 2 of the present invention provides a high-resistance grounding fault detection system based on a multi-band variable data window, which is based on the aforementioned method for detecting high-resistance grounding faults based on a multi-band variable data window, such as... Figure 2 As shown, it includes:

[0105] Zero-sequence voltage and zero-sequence current fault data caching module, multi-band filtering module, similarity and volt-ampere characteristic analysis module, variable data window calculation module, and line fault fusion judgment module;

[0106] The zero-sequence voltage and zero-sequence current fault data caching module is used to acquire discrete zero-sequence voltage and current data after a fault; the sequence multi-band filtering module is used to perform infinite pulse filtering on the cached data at different frequency bands; the similarity and volt-ampere characteristic analysis module is used to process the mathematical relationship calculation of zero-sequence voltage and current after filtering; the variable data window calculation module is used to change the similarity and volt-ampere characteristic analysis data segment and perform calculation again; the line fault fusion judgment module is used to determine whether the line belongs to the faulty line by comprehensively considering the similarity and volt-ampere characteristic calculation results.

[0107] Embodiment 3 of the present invention provides an electronic device based on a multi-band variable data window high-resistance grounding fault detection method, which runs the aforementioned multi-band variable data window high-resistance grounding fault detection method:

[0108] It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

[0109] Memory, used to store computer programs;

[0110] The processor is used to execute the program stored in the memory to implement the steps of the method for detecting high-resistance grounding faults based on multi-band variable data windows.

[0111] Embodiment 4 of the present invention provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned method for detecting high-resistance grounding faults based on multi-band variable data windows.

[0112] To more clearly illustrate the outstanding substantive features of this invention and the significant progress it brings to the prior art, an application example of implementing this invention is described below.

[0113] Embodiment 5 of the present invention performs multi-band decomposition of zero-sequence voltage and zero-sequence current data using the infinite pulse filter proposed in Embodiment 1. The infinite pulse filter adopts a Chebyshev Type I filter, and the multi-band filter coefficients are as follows:

[0114]

[0115] The waveform of the grounding fault in the field is processed by using multi-band filtering coefficients. The similarity of zero-sequence voltage and zero-sequence current and the volt-ampere characteristics in Example 1 are analyzed, and the fault result is judged by combining the results of variable data window. Figure 3 This is the waveform of the on-site fault. Figure 4 The results of multi-band processing show that zero-sequence voltage and zero-sequence current exhibit different similarity characteristics in different frequency bands. Figure 5 The typical volt-ampere characteristics of a fault waveform are shown; the volt-ampere characteristics of a faulty line exhibit a clockwise rotation. Figure 6 The typical characteristics of a non-fault waveform are shown, and the volt-ampere characteristics of a non-fault line exhibit a counterclockwise rotation.

[0116] The method of this invention performs multi-band decomposition on high-resistance grounding waveforms, identifies faults by using zero-sequence voltage, zero-sequence current similarity and volt-ampere characteristics, and reduces the interference of noise on fault misjudgment by combining the results of variable data window calculation. The effectiveness of the algorithm is verified by a large number of waveforms.

[0117] The beneficial effects of this invention are that, compared with the prior art, this invention provides a high-resistivity grounding fault detection method and system based on multi-band variable data window. By performing multi-band analysis on zero-sequence voltage and zero-sequence current data, it can more accurately extract the frequency bands with obvious fault characteristics. Combined with variable data window calculation and verification, it avoids misjudgment caused by random fluctuations and has great engineering application value.

[0118] Specifically, compared with existing solutions that use VMD or EMD for ground fault line selection, this invention employs infinite pulse filtering with low computational load, allows for free selection of the frequency band combination range, optimizes the frequency band based on actual engineering faults, is compatible with ground fault characteristics of both high-current and low-current systems, and considers the similarity and volt-ampere characteristics of zero-sequence voltage and zero-sequence current in each frequency band, as well as the difference in zero-sequence voltage and the similarity of zero-sequence current in each frequency band. It has high sensitivity and reliability in identifying high-impedance faults, and the transient decomposition frequency band range is easily optimized, with controllable filtering convergence.

[0119] It is worth noting that in the embodiments of the present invention, "steps + numbers" is only an expression for clearly describing the specific implementation of the method for detecting high-resistance grounding faults based on multi-band variable data windows, and is not an absolute restriction on the order of the steps. Under the guidance of the core concept of the present invention, changing the order of these steps to obtain the same or similar technical effects all fall within the scope of the present invention.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for detecting high-resistivity grounding faults based on multi-band variable data windows, characterized in that, Includes the following steps: The zero-sequence voltage and zero-sequence current are set to buffer two-cycle fault data based on the effective value or sudden change of the zero-sequence voltage. The cached fault data is decomposed by multi-band filtering to obtain zero-sequence voltage and zero-sequence current data in different frequency bands; For each frequency band, select a characteristic data window for zero-sequence voltage and zero-sequence current data, calculate the zero-sequence parameter similarity and volt-ampere characteristic index under the same frequency band; and under the same frequency band, change the characteristic data window and calculate the zero-sequence parameter similarity and volt-ampere characteristic index again. For each frequency band, a consistency check is performed based on the two calculation results. If the consistency check passes, the zero-sequence parameter similarity and volt-ampere characteristic index of that frequency band are substituted into the criterion. The high-impedance fault is judged according to the interval to which the zero-sequence parameter similarity and volt-ampere characteristic index belong, or the frequency of occurrence.

2. The method for detecting high-resistivity grounding faults based on multi-band variable data windows according to claim 1, characterized in that: The cached fault data includes two cycles of cached fault data: 1.5 cycles before the fault and 0.5 cycles after the fault.

3. A method for detecting high-resistance grounding faults based on multi-band variable data windows according to claim 1 or 2, characterized in that: The multi-band filtering and decomposition of the cached fault data to obtain zero-sequence voltage and zero-sequence current data in different frequency bands includes: IIR filters are used to perform multi-frequency filtering decomposition on the buffered fault data of zero-sequence voltage and zero-sequence current to obtain zero-sequence voltage and zero-sequence current data in different frequency bands. Based on the obtained zero-sequence voltage and zero-sequence current data for each frequency band, the first-order difference of the zero-sequence voltage is calculated for each frequency band.

4. The method for detecting high-resistivity grounding faults based on multi-band variable data windows according to claim 3, characterized in that: The process of selecting a feature data window for zero-sequence voltage and zero-sequence current data for each frequency band, and calculating the zero-sequence parameter similarity and volt-ampere characteristic indices for the same frequency band, includes: For the k-th frequency band, the segment between the minimum point S1 and the maximum point S2 after zero-sequence voltage filtering is selected as the feature data window. Based on the selected feature data window, the similarity of zero-sequence voltage and zero-sequence current, as well as the similarity of zero-sequence voltage difference and zero-sequence current under the k-th frequency band, are calculated, and the smaller of the two similarity values ​​is selected as the zero-sequence parameter similarity. Calculate the volt-ampere characteristics of zero-sequence voltage and zero-sequence current in the k-th frequency band.

5. The method for detecting high-resistivity grounding faults based on multi-band variable data windows according to claim 4, characterized in that: Before changing the feature data window, the zero-sequence parameter similarity under the k-th frequency band is checked. If it is less than the set threshold set, the check passes, the feature data window is changed, and the zero-sequence parameter similarity and volt-ampere characteristic index under the k-th frequency band are recalculated for consistency verification of the two calculation results; otherwise, the zero-sequence parameter similarity and volt-ampere characteristic index under the k-th frequency band are no longer used, the frequency band value is updated with k=k+1, and the calculation of the zero-sequence parameter similarity and volt-ampere characteristic index under the next frequency band begins.

6. A multi - band variable data window high - impedance grounding fault detection method according to claim 4 or 5, characterized in that: Under the same frequency band, changing the characteristic data window and recalculating the zero - sequence parameter similarity and volt - ampere characteristic index includes: Setting a translation step ΔS, and shifting the characteristic data window backward by the translation step ΔS; Based on the shifted characteristic data window, recalculating the similarity of zero - sequence voltage and zero - sequence current, the similarity of zero - sequence voltage difference and zero - sequence current, and the volt - ampere characteristic index of zero - sequence voltage and zero - sequence current at the k - th frequency band, and further obtaining the zero - sequence parameter similarity after changing the characteristic data window at the k - th frequency band.

7. A multi - band variable data window high - impedance grounding fault detection method according to claim 4 or 5, characterized in that: For each frequency band, performing consistency verification based on the two calculation results includes: If the zero - sequence parameter similarities before and after changing the characteristic data window at the k - th frequency band are both less than the set threshold set, then the consistency verification is passed, and the zero - sequence parameter similarity and volt - ampere characteristic index at the k - th frequency band can be used to substitute into the criterion for judging whether a high - impedance fault occurs; if the consistency verification fails, the zero - sequence parameter similarity and volt - ampere characteristic index at the k - th frequency band are no longer used, and the frequency band value is updated with k = k + 1, and the calculation of the zero - sequence parameter similarity and volt - ampere characteristic index at the next frequency band is entered.

8. A multi - band variable data window high - impedance grounding fault detection method according to claim 4 or 5, characterized in that: Judging whether a high - impedance fault occurs according to the interval to which the zero - sequence parameter similarity and volt - ampere characteristic index belong, or the occurrence frequency includes: If, among all the frequency band numbers K of a certain line, there appears any frequency band value k such that the zero - sequence parameter similarity P(k) < - 0.8 and the volt - ampere characteristic index Q(k) < 0, then this line is a fault line; otherwise, continue to perform the subsequent judgment; For this line, counting the occurrence frequency of the state where - 0.8 < P(k) < - 0.5 and Q(k) < 0. If the frequency is greater than 3, this line is a fault line; otherwise, this line is a non - fault line.

9. A high-resistance grounding fault detection system based on a multi-band variable data window, operating the high-resistance grounding fault detection method based on a multi-band variable data window as described in any one of claims 1 to 8, characterized in that, Including: A zero - sequence voltage and zero - sequence current fault data caching module for obtaining the discrete data sequences of zero - sequence voltage and current after a fault; A multi - band filtering module for performing multi - band filtering decomposition on the cached fault data to obtain zero - sequence voltage and zero - sequence current data at different frequency bands; A similarity and volt - ampere characteristic analysis module for, for each frequency band, selecting the characteristic data window of zero - sequence voltage and zero - sequence current data and calculating the zero - sequence parameter similarity and volt - ampere characteristic index at the same frequency band; A variable data window calculation module for changing the characteristic data window and recalculating the zero - sequence parameter similarity and volt - ampere characteristic index; A line fault fusion judgment module for comprehensively judging whether the line is a fault line according to the calculation results of similarity and volt - ampere characteristics.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The computer program, when loaded into the processor, implements a multi - band variable data window high - impedance grounding fault detection method according to any one of claims 1 to 8.

11. A computer-readable storage medium storing a computer program, characterized in that, The computer program, when executed by the processor, implements a multi - band variable data window high - impedance grounding fault detection method according to any one of claims 1 to 8.

Images