Single-phase grounding fault intelligent line selection method for small current grounding system of power distribution network

Abstract

The present application relates to the technical field of power system grounding fault detection, in particular to a single-phase grounding fault intelligent line selection method for a small-current grounding system of a distribution network, which locks a transient signal interception window through a zero-sequence voltage instantaneous value, accurately extracts a zero-sequence current high-frequency component signal corresponding to a high-frequency frequency band level by using discrete wavelet packet decomposition, and determines a morphological difference value based on an energy distribution deviation of a feeder compared with other feeders, so as to convert absolute magnitude discrimination into a comparison of frequency band distribution forms, and eliminate the interference of high-frequency transformation errors of underlying mutual inductors on a decision result; in combination with a total energy weight representing a physical scale of a signal, a morphological distribution artifact generated by weak white noise is identified and suppressed in a targeted manner; by fusing the morphological difference value and the total energy weight, deep mining of a fault branch distribution structure feature and adaptive shielding of noise interference are realized, so that the single-phase grounding fault intelligent line selection for the small-current grounding system of the distribution network is more accurate.

Description

Technical Field

[0001] This invention relates to the field of power system grounding fault detection technology, specifically to an intelligent fault location method for single-phase grounding faults in low-current grounding systems of distribution networks. Background Technology

[0002] Medium-voltage distribution networks widely use low-current grounding systems. The core of single-phase grounding fault location lies in capturing the transient high-frequency zero-sequence current signal generated instantaneously during a fault. Existing technologies typically use the absolute amplitude characteristics of transient waveforms for single-phase grounding fault location. This involves directly extracting the high-frequency absolute amplitude of the transient zero-sequence current collected from each parallel feeder and performing algebraic addition and subtraction to find the faulty branch with unbalanced residual current.

[0003] However, existing technologies do not take into account the high-frequency transmission error that is common in current transformers configured in the data acquisition channels at the bottom of substations. This makes it impossible to achieve theoretical equation closure when the absolute current value with measurement ratio error is directly summed algebraically. It cannot effectively eliminate the interference of hardware measurement error on the characteristic absolute magnitude, and is very easy to cause residual current illusion due to weak white noise. As a result, the accuracy of existing technologies in selecting single-phase grounding faults based on the absolute amplitude characteristics of transient waveforms is low. Summary of the Invention

[0004] To address the low accuracy of existing technologies that rely on the absolute amplitude characteristics of transient waveforms for single-phase grounding fault location, this invention aims to provide an intelligent fault location method for single-phase grounding faults in low-current grounding systems of distribution networks. The specific technical solution adopted is as follows: The first aspect of this invention provides an intelligent fault location method for single-phase grounding faults in a low-current grounding system of a distribution network, comprising: Obtain the instantaneous value of zero-sequence voltage of the substation bus at each sampling time; determine the transient signal interception window based on the numerical limit characteristics of the instantaneous value of zero-sequence voltage; extract the original waveform signal of zero-sequence current of each feeder from the transient signal interception window; Discrete wavelet packet decomposition is performed on the original waveform signal of the zero-sequence current to determine all high-frequency band levels and their corresponding high-frequency components of the zero-sequence current for each feeder; based on the high-frequency components of the zero-sequence current, the morphological difference value of each feeder is determined according to the energy distribution deviation of the high-frequency band levels of each feeder compared with all other feeders. Based on the overall energy magnitude of all zero-sequence current high-frequency component signals corresponding to each feeder, the corresponding total energy weight is determined; the total energy weight and the morphological difference value are fused to determine the single-phase grounding fault characteristic value of each feeder; and the single-phase grounding fault line is selected based on the single-phase grounding fault characteristic value.

[0005] Furthermore, the process of acquiring the transient signal interception window includes: The sampling time corresponding to the first instantaneous value of the zero-sequence voltage that is greater than the preset voltage start-up threshold is taken as the transient signal interception time. The sampling time corresponding to the preset buffer duration is traced back from the moment the transient signal is captured, and this is taken as the start time of the transient signal. The transient signal capture window is determined by extending the preset capture time length from the start time of the transient signal.

[0006] Furthermore, the process of performing discrete wavelet packet decomposition on the original zero-sequence current waveform signal to determine all high-frequency band levels and their corresponding zero-sequence current high-frequency component signals for each feeder includes: Discrete wavelet packet decomposition is performed on the original waveform signal of the zero-sequence current to determine the zero-sequence current component signal corresponding to each frequency band level in the effective characteristic frequency band range. All frequency band levels within the effective characteristic frequency band range are arranged in descending order of their corresponding frequencies to determine the frequency band level sequence; the first preset number of frequency band levels in the frequency band level sequence are selected as high-frequency frequency band levels; and the zero-sequence current component signal corresponding to each high-frequency frequency band level is selected as the zero-sequence current high-frequency component signal.

[0007] Furthermore, the process of obtaining the morphological difference value includes: The zero-sequence current high-frequency component signal is integrally squared to determine the local frequency band energy value of each feeder at each high-frequency band level; the local frequency band energy values ​​of each feeder at all high-frequency band levels are superimposed to determine the corresponding total feeder energy value. At each high-frequency band level, the local frequency band energy values ​​corresponding to all other feeders outside each feeder are superimposed to determine the corresponding total background frequency band energy value. The total energy value of all feeders other than each feeder is superimposed to determine the corresponding total energy value of the background feeders. The local frequency band energy value is standardized by the total energy value of the feeder to determine the feeder energy distribution ratio of each feeder at each high frequency band level. The total energy value of the background frequency band is standardized by the total energy value of the background feeder to determine the background energy distribution ratio of each feeder in each high-frequency band level. For each feeder, the corresponding morphological difference value is determined based on the overall deviation between the feeder energy distribution ratio and the background energy distribution ratio corresponding to each high-frequency band level.

[0008] Furthermore, the process of obtaining the total energy weight includes: Each feeder is then used as the target feeder in turn; When the total energy value of the target feeder is greater than or equal to the preset energy baseline threshold, the corresponding total energy weight is set to the preset weight value. When the total energy value of the target feeder is less than the preset energy baseline threshold, the total energy value of the feeder is negatively correlated and mapped to determine the corresponding total energy weight; the total energy weight is greater than or equal to the preset weight value.

[0009] Furthermore, the process of obtaining the single-phase ground fault characteristic value includes: The characteristic value of a single-phase grounding fault is determined by multiplying the total energy weight with the morphological difference value.

[0010] Furthermore, the process of selecting the line for a single-phase ground fault based on the single-phase ground fault characteristic value includes: The single-phase grounding fault characteristic values ​​are sequentially subjected to negative correlation mapping and normalization to determine the fault probability of each feeder; based on the fault probability, the final selected feeder is selected; and the single-phase grounding fault line selection is performed based on the final selected feeder.

[0011] Furthermore, the process of obtaining the final selected feeder includes: The feeder with the highest probability of failure will be selected as the final feeder.

[0012] Furthermore, the preset weight value is set to 1.

[0013] Furthermore, the effective characteristic frequency band range is set to 1000Hz to 1200Hz.

[0014] Secondly, the present invention provides an intelligent fault location system for single-phase grounding faults in a low-current grounding system of a power distribution network, the system comprising: The data acquisition and preprocessing module is used to acquire the instantaneous value of zero-sequence voltage of the substation bus at each sampling time; determine the transient signal interception window based on the numerical limit characteristics of the instantaneous value of zero-sequence voltage; and extract the original waveform signal of zero-sequence current of each feeder from the transient signal interception window. The morphological difference determination module is used to perform discrete wavelet packet decomposition on the original waveform signal of the zero-sequence current to determine all high-frequency band levels and their corresponding zero-sequence current high-frequency component signals for each feeder; based on the zero-sequence current high-frequency component signals, the morphological difference value corresponding to each feeder is determined according to the energy distribution deviation of the high-frequency band levels of each feeder compared with all other feeders. The fault selection module is used to determine the corresponding total energy weight based on the overall energy magnitude of all zero-sequence current high-frequency component signals corresponding to each feeder; to fuse the total energy weight with the morphological difference value to determine the single-phase grounding fault characteristic value of each feeder; and to perform single-phase grounding fault selection based on the single-phase grounding fault characteristic value.

[0015] Thirdly, the present invention provides a computer device including a memory and a processor. The memory is used to store computer program code, and the processor is used to call and run the computer program code from the memory to perform the method as described in the first aspect or any embodiment of the first aspect of the present invention.

[0016] Fourthly, the present invention provides a computer program product comprising computer program code, which, when executed, performs the method as described in the first aspect or any embodiment of the first aspect of the present invention.

[0017] Fifthly, the present invention provides a computer-readable storage medium storing computer program code that, when executed, performs the method as described in the first aspect or any embodiment of the first aspect of the present invention.

[0018] The present invention has the following beneficial effects: This invention locks the transient signal extraction window by using the instantaneous value of zero-sequence voltage, accurately extracts the high-frequency component signal of zero-sequence current corresponding to the high-frequency band level using discrete wavelet packet decomposition, and determines the morphological difference value based on the energy distribution deviation of the feeder compared to other feeders. This transforms the absolute magnitude discrimination into a comparison of frequency band distribution morphology, effectively eliminating the interference of high-frequency transmission errors of the underlying transformers on the decision results. Combined with the total energy weight representing the physical scale of the signal, it can specifically identify and suppress morphological distribution artifacts caused by weak white noise, blocking the secondary transmission of measurement errors and noise fluctuations to the final feature value. By fusing the morphological difference value and the total energy weight, it achieves in-depth mining of the structural characteristics of the fault branch distribution and adaptive shielding of noise interference, resulting in higher accuracy in intelligent fault selection for single-phase grounding faults in low-current grounding systems of distribution networks. Attached Figure Description

[0019] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1This is a flowchart of an intelligent fault location method for a single-phase grounding fault in a low-current grounding system of a power distribution network, provided by an embodiment of the present invention. Figure 2 This is a structural diagram of a single-phase grounding fault intelligent fault location system for a low-current grounding system in a power distribution network, provided in one embodiment of the present invention. Figure 3 This is a schematic diagram of a computer device structure provided in one embodiment of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] This invention provides an intelligent fault location method for single-phase grounding faults in a low-current grounding system of a power distribution network. Please refer to [link / reference]. Figure 1 The diagram illustrates a flowchart of an intelligent fault location method for a single-phase grounding fault in a low-current grounding system of a distribution network, according to an embodiment of the present invention. The method includes: Step S101: Obtain the instantaneous value of zero-sequence voltage of the substation bus at each sampling time; determine the transient signal interception window based on the numerical over-limit characteristics of the instantaneous value of zero-sequence voltage; extract the original waveform signal of zero-sequence current of each feeder from the transient signal interception window.

[0023] In the actual operation environment of a distribution network substation, this embodiment of the invention collects the instantaneous value of zero-sequence voltage at each sampling moment through zero-sequence voltage transformers configured on the substation bus side, and collects the real-time signal of zero-sequence current corresponding to each feeder through zero-sequence current transformers installed on each parallel feeder. In order to effectively capture the high-frequency components during transient discharge and avoid signal misalignment caused by inherent delays in the hardware of each acquisition terminal, this embodiment of the invention performs synchronous discrete sampling of the signal through a preset high-frequency sampling frequency. This embodiment of the invention sets the preset high-frequency sampling frequency to 3kHz, and connects all data acquisition channels to a unified in-station clock source (such as IRIG-B or PPS time synchronization system) for centralized time synchronization control to ensure absolute time synchronization of the entire network signal. It should be noted that, considering that the number of feeders is less than or equal to 2, it will affect the calculation of subsequent morphological difference values, therefore, this embodiment of the invention only analyzes the case where the number of feeders is greater than 2.

[0024] Considering that the bus zero-sequence voltage will rise sharply when a single-phase ground fault occurs, and that the entire transient discharge process is extremely short and highly concentrated, this embodiment of the invention further determines the transient signal interception window based on the numerical limit-crossing characteristics of the instantaneous zero-sequence voltage value. Specifically: the sampling time corresponding to the first instantaneous zero-sequence voltage value that is greater than the preset voltage start threshold is taken as the transient signal interception time; since the high-frequency resonance characteristics generated at the moment of a single-phase ground fault have significant leading and rapid decay physical properties, in order to prevent the loss of key information before the voltage exceeds the limit, the system sends a read command to the ring buffer of the underlying data acquisition device, and backtracks the sampling time corresponding to the preset buffer duration from the transient signal interception time as the transient signal start time; the transient signal interception window is determined by extending the preset interception time length from the transient signal start time.

[0025] In one specific implementation of this invention, the preset voltage initiation threshold is set to 15% of the rated phase voltage of the busbar, used to characterize the significant potential fluctuations during a single-phase ground fault. Its specific value can be adaptively fine-tuned by maintenance personnel based on the voltage deviation level of the busbar under normal operation and the intensity of electromagnetic interference on site. The preset buffer duration can be set to 5 milliseconds to cover the steady-state and transient transition period before fault initiation; the preset truncation time is set to 20 milliseconds to cover the waveform components with the most complete release of transient energy, thereby providing complete data samples for subsequent frequency band characteristic analysis. The preset buffer duration and preset truncation time can be flexibly configured according to the distributed capacitance of the power grid and the attenuation constant of the transient process.

[0026] Further, the original waveform signal of the zero-sequence current of each feeder is extracted in the transient signal extraction window. Specifically, the real-time zero-sequence current signal of each feeder retained in the ring buffer is synchronously extracted directly according to the transient signal extraction window to obtain the original waveform signal of the zero-sequence current of each feeder in the transient signal extraction window.

[0027] Step S102: Perform discrete wavelet packet decomposition on the original waveform signal of zero-sequence current to determine all high-frequency band levels and their corresponding high-frequency components of zero-sequence current for each feeder; based on the high-frequency components of zero-sequence current, determine the morphological difference value for each feeder according to the energy distribution deviation of the high-frequency band levels of each feeder compared to all other feeders.

[0028] After extracting the original waveform signal of the zero-sequence current for each feeder, considering that the single-phase grounding transient signal is a typical non-stationary random signal, its rich high-frequency resonance features have extremely strong localized transient characteristics in both the time and frequency domains. However, the traditional Fourier transform cannot take into account the joint time-frequency resolution, and the conventional wavelet transform lacks the ability to further refine the high-frequency components. Therefore, this embodiment of the invention further performs discrete wavelet packet decomposition on the original waveform signal of the zero-sequence current to determine all high-frequency band levels corresponding to each feeder and their corresponding high-frequency components of the zero-sequence current. By dividing the signal into multi-scale uniform divisions across the entire frequency band, the transient features of each frequency band hidden in the complex time-domain waveform can be independently decoupled, laying a high-fidelity frequency domain data foundation for subsequently obtaining the dimensionless energy distribution ratio and morphological difference values.

[0029] Preferably, in some possible implementations of the embodiments of the present invention, the process of performing discrete wavelet packet decomposition on the original waveform signal of the zero-sequence current to determine all high-frequency band levels corresponding to each feeder and its corresponding high-frequency component signal of the zero-sequence current includes: Discrete wavelet packet decomposition is performed on the original waveform signal of zero-sequence current to determine the zero-sequence current component signal corresponding to each frequency band level within the effective characteristic frequency band range. Compared with adaptive decomposition algorithms such as empirical mode decomposition or local mean decomposition, discrete wavelet packet decomposition uses a fixed decomposition tree structure and scaling function calculation mechanism, which can ensure that for the original waveforms of different feeders across the entire network, the decomposed frequency bands maintain absolute strict correspondence and consistency on the physical frequency axis, completely avoiding the problem of misalignment of co-frequency features caused by mode mixing. In a specific implementation of this invention, the effective characteristic frequency band range is set to 1000Hz to 1200Hz. This range typically covers the dominant high-frequency resonant center frequency caused by line distribution parameters when a single-phase ground fault occurs in a medium-voltage distribution network, which can avoid 50Hz power frequency fundamental interference and extremely high frequency random white noise to the greatest extent. In one specific implementation of this invention, the number of frequency band decomposition layers is set to 6. The system analysis spectrum is proportionally divided into multiple sub-bands with fixed bandwidths using a wavelet packet binary tree decomposition structure. Each frequency band layer essentially corresponds to a sub-band within a specific frequency range, with the corresponding frequency specifically referring to the center frequency value of that sub-band, thereby ensuring that the extracted features of the same level from the entire network are absolutely aligned in physical bandwidth. It should be noted that discrete wavelet packet decomposition is a technique well-known to those skilled in the art and will not be further limited or elaborated upon here.

[0030] All frequency band levels within the effective characteristic frequency range are arranged in descending order of their corresponding frequencies to determine the frequency band level sequence. The first preset number of frequency band levels in the frequency band level sequence are designated as high-frequency frequency band levels. The zero-sequence current component signal corresponding to each high-frequency frequency band level is designated as the zero-sequence current high-frequency component signal. In a specific implementation of this invention, the preset number of high-frequency screening is set to 5, which is used to accurately extract the core frequency band set where transient discharge energy is most concentrated after completely stripping away power frequency and low-frequency components. The specific value can be adjusted according to the fineness requirements of frequency domain feature resolution and the computing load capacity of the underlying computer in the specific implementation environment.

[0031] After extracting the original waveform signal of the zero-sequence current for each feeder, considering the complex operating conditions at the substation, the differences in the physical transmission characteristics of current transformers and the dynamic changes in the distribution network topology often lead to the absolute amplitude of the zero-sequence current between feeders not satisfying the theoretical closure condition of algebraic addition and subtraction. Directly using amplitude characteristics for line selection is prone to misjudgment. Therefore, this embodiment of the invention, based on the high-frequency component signal of the zero-sequence current, determines the morphological difference value corresponding to each feeder according to the high-frequency band level energy distribution deviation of each feeder compared to all other feeders. By introducing a relative morphological comparison mechanism, the line selection criterion is transformed from relying on unstable absolute physical quantities to relying on robust frequency band distribution morphology, laying a reliable feature benchmark for subsequent elimination of hardware error interference and achieving consistency discrimination across feeders.

[0032] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the morphological difference value includes: The local band energy value of each feeder at each high-frequency band level is determined by performing square integration on the high-frequency component signal of the zero-sequence current. This involves squaring all signal amplitudes in the high-frequency component signal of the zero-sequence current and then integrating the results to determine the local band energy value. Since the high-frequency resonant signal exhibits alternating positive and negative polarity oscillations within the transient signal interception window, directly integrating the high-frequency component signal of the zero-sequence current would cause the positive and negative half-waves to cancel each other out, failing to accurately reflect the cumulative disturbance scale released in that frequency band. Therefore, this embodiment of the invention calculates based on the square of the signal amplitude, so that a larger local band energy value indicates more intense discharge activity occurring on that feeder within that specific frequency range.

[0033] The total energy value of the feeder is determined by superimposing the local frequency band energy values ​​of each feeder at all high-frequency band levels. In this embodiment of the invention, the total energy value of the feeder is determined by accumulating the local frequency band energy values ​​of each feeder at all high-frequency band levels, so that the total energy value of the feeder can characterize the overall disturbance intensity of the line during the fault transient process.

[0034] Furthermore, considering the fundamental difference in charge release mechanisms between faulty and non-faulty feeders during a single-phase ground fault in the power grid (the faulty feeder typically bears the transient capacitor discharge current of the entire network), relying solely on the energy value of a single feeder is insufficient to eliminate the influence of environmental noise or hardware measurement deviations. Therefore, in this embodiment of the invention, when analyzing any target feeder, all other feeders are considered as a whole set. At each high-frequency band level, the local frequency band energy values ​​of all feeders other than the target feeder are accumulated to determine the corresponding total background frequency band energy value; and further, the total feeder energy values ​​of all feeders within this set are accumulated to determine the corresponding total background feeder energy value. By constructing such a reference benchmark based on the overall network energy distribution, subsequent comparisons can reflect the characteristics of a single feeder deviating from the overall network distribution, thereby eliminating local environmental interference through group consistency.

[0035] Further, the local frequency band energy value is standardized using the total feeder energy value to determine the feeder energy distribution ratio of each feeder at each high-frequency band level. Specifically, the feeder energy distribution ratio is determined based on the ratio between the local frequency band energy value of each feeder at each high-frequency band level and the total feeder energy value of the corresponding feeder. Similarly, the background frequency band total energy value is standardized using the background feeder total energy value to determine the background energy distribution ratio of each feeder at each high-frequency band level. Specifically, the background energy distribution ratio is determined based on the ratio between the background frequency band total energy value of each feeder at each high-frequency band level and the background feeder total energy value of the corresponding feeder.

[0036] This standardization process transforms the absolute physical energy values ​​affected by hardware transmission errors into relative proportional characteristics that characterize the energy distribution structure of frequency bands. The core function of this normalization is that even if there are gain differences in the current transformer ratios of different outgoing line cabinets, as long as the gain is linear, the energy proportion structure between frequency bands will remain constant after normalization. This eliminates the absolute magnitude bias in the hardware measurement process, highlighting subtle distortion differences in the frequency domain morphology of different lines and enhancing the robustness of the line selection process to inconsistencies in hardware parameters.

[0037] It should be noted that, considering the possibility of completely unloaded or disconnected idle feeders in the power grid, when the total energy value of a feeder is 0, it indicates that the corresponding feeder has not experienced a transient energy response. In this case, the energy distribution ratio of the feeder is directly set to 0 to avoid a denominator of 0 and to ensure that the feeder does not participate in subsequent morphological difference calculations. Similarly, when the total energy value of the background feeder is 0, it indicates that the entire network is extremely stable or that there is signal loss at the acquisition end. In this case, the background energy distribution ratio is directly set to 0 to avoid division by zero anomalies in mathematical calculations, thereby ensuring the numerical stability of the calculation process.

[0038] For each feeder, a corresponding morphological difference value is determined based on the overall deviation between the feeder energy distribution ratio and the background energy distribution ratio corresponding to each high-frequency band level. In a specific implementation of this invention, for each feeder, a corresponding distribution deviation factor is determined based on the absolute value of the difference between the background energy distribution ratio and the corresponding feeder energy distribution ratio corresponding to each high-frequency band level; then, the corresponding morphological difference value is determined based on the cumulative value of the distribution deviation factors for all high-frequency band levels under each feeder.

[0039] The feeder energy distribution ratio characterizes the structural proportion of energy released by a single line in a specific frequency band, while the background energy distribution ratio characterizes the baseline distribution level of energy released by the remaining lines in the entire network in that frequency band. Therefore, the larger the cumulative value of the distribution deviation factor, the more serious the deviation between the energy distribution ratio of the line in each high-frequency band and the common distribution characteristics of the entire network. Consequently, the morphological difference value representing the degree of distribution structure distortion in the corresponding single-phase grounding fault characteristic value is also greater, and the less it matches the characteristics of the faulty line in terms of feature similarity. This logic based on the accumulation of absolute difference values ​​can keenly capture the distribution morphological deviation characteristics of non-faulty lines caused by the lack of a dominant resonant frequency band, thus achieving preliminary screening of faulty lines solely through morphological characteristics without relying on absolute amplitude.

[0040] Step S103: Based on the overall energy magnitude of all zero-sequence current high-frequency component signals corresponding to each feeder, determine the corresponding total energy weight; fuse the total energy weight and the morphological difference value to determine the single-phase grounding fault characteristic value of each feeder; select the single-phase grounding fault line based on the single-phase grounding fault characteristic value.

[0041] While the standardization process eliminates transformer ratio errors, extremely weak background white noise or electromagnetic interference, due to its random and uniform distribution across the entire frequency band, may occasionally exhibit a highly homogeneous shape that closely matches the transient distribution of the entire network after standardization. This results in minimal calculated morphological differences, potentially leading to fatal misjudgments. Since real grounding transient signals should possess a significant absolute advantage in energy scale, this embodiment further determines the corresponding total energy weight based on the overall energy magnitude of all zero-sequence current high-frequency component signals corresponding to each feeder (i.e., the total energy value of the feeder). By introducing an absolute scale characterizing the physical energy scale as a secondary verification benchmark, it is possible to effectively identify and suppress false feature similarities induced by weak signals. This leverages the amplification effect of the energy weight to eliminate noise interference, ensuring the physical rationality and high accuracy of the line selection results.

[0042] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the total energy weight includes: Each feeder is then used as the target feeder in turn; When the total energy value of the target feeder is greater than or equal to the preset energy baseline threshold, the corresponding total energy weight is set to the preset weight value. When the total energy value of the target feeder is less than the preset energy baseline threshold, the total energy value of the feeder is negatively correlated and the corresponding total energy weight is determined; the total energy weight is greater than or equal to the preset weight value.

[0043] The aforementioned weighting logic aims to establish a physical barrier based on energy scale. By setting a preset energy threshold, the system can distinguish between genuine transient discharge signals and system background noise. For genuine signals with sufficient energy, a constant preset weight value is assigned to maintain their original morphological characteristics; while for suspected noise signals with energy scales below the threshold, a larger weight value is assigned through negative correlation mapping, aiming to amplify the morphological differences caused by possible false similarities, thereby effectively filtering out weak noise artifacts.

[0044] In one specific implementation of this invention, the preset energy baseline threshold is set to be 1.5 to 3 times the average background noise energy under normal substation operation. The specific value can be adjusted according to the electromagnetic interference intensity in the specific implementation environment. In this embodiment, it is preferably set to 2 times this average. When there are many nonlinear loads such as frequency converters connected in the implementation environment, and the environmental electromagnetic interference is strong, the preset energy baseline threshold should be set larger to improve the system's fault tolerance to high-level background noise. Conversely, if the substation operating conditions are relatively clean, the threshold can be appropriately lowered to retain the detection sensitivity for extremely weak grounding faults. The average background noise energy can be determined by capturing the steady-state zero-sequence current signals of all feeders in the entire network under multiple preset capture time lengths when the substation is in a state of normal and stable grid operation without transient disturbances, calculating the corresponding total feeder energy value, and obtaining the arithmetic mean of the entire network. The preset weight value is set to 1, used as the reference gain for normal signals, ensuring that the original physical distribution attributes of the morphological difference value are not changed when the fault energy is sufficient, thereby ensuring the objectivity of the line selection result.

[0045] In one specific implementation of this invention, the process of obtaining the total energy weight when the total energy value of the target feeder is less than a preset energy baseline threshold is expressed by the following formula: ;in, For target feeder Total energy weight; The preset energy baseline threshold; For target feeder The total energy value of the feeder. It should be noted that if the target feeder... If the total energy value of the feeder is 0, it means that the feeder is in a completely silent or shut-down state with no induced current flowing through it, and does not have the physical basis for a grounding fault. In this case, we will directly abandon the subsequent processing of the target feeder and assume that the feeder has no abnormality, so as to avoid the situation where the denominator is 0.

[0046] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the characteristic value of a single-phase ground fault includes: The single-phase grounding fault characteristic value is determined by multiplying the total energy weight with the morphological difference value. The total energy weight characterizes the degree of deviation or scarcity of the feeder transient signal energy level relative to the system background noise level; essentially, it is a reliability adjustment factor for signal strength. Therefore, by multiplying and weighting the morphological difference value with the total energy weight, background noise or occasional interference signals that may have false morphological similarity in frequency band distribution due to accidental factors but whose actual physical energy volume is severely insufficient can have their final single-phase grounding fault characteristic value forcibly pushed up to a numerical range representing a non-fault state through the amplifying effect of the weighting coefficient.

[0047] Preferably, in some possible implementations of the embodiments of the present invention, the process of selecting a single-phase ground fault line based on the characteristic value of a single-phase ground fault includes: The characteristic values ​​of single-phase grounding faults are sequentially subjected to negative correlation mapping and normalization to determine the fault probability of each feeder; based on the fault probability, the final selected feeder is selected; and the single-phase grounding fault line selection is performed based on the final selected feeder.

[0048] In one specific implementation of this invention, the process of obtaining the fault probability is expressed by the formula: ;in, For feeder The probability of failure; For feeder Characteristic values ​​of single-phase grounding faults; It is an exponential function with the natural constant as its base; For feeder The negative correlation mapping results of the characteristic values ​​of single-phase grounding faults; This represents the total number of feeders; For the first Characteristic values ​​of single-phase grounding faults in each feeder.

[0049] In one specific implementation of this invention, if it is possible to determine in advance that a fault exists in the implementation environment but the faulty line cannot be identified, the feeder corresponding to the highest fault probability is selected as the final feeder.

[0050] The faulty line most closely resembles the background distribution in terms of frequency band distribution, and its transient energy scale is much larger than the noise floor threshold. Therefore, its corresponding single-phase grounding fault characteristic value is at its minimum level. By introducing a negative correlation mapping, this tiny characteristic value can be transformed into the largest fault tendency scalar in the entire network, thus occupying the largest probability weight in the subsequent normalization summation process, significantly differentiating the faulty line from the non-faulty line at the probability output level. Therefore, the feeder with the highest fault probability is finally selected as the feeder.

[0051] In another specific implementation of this invention, if it is impossible to determine whether a fault exists in the implementation environment, a fault probability threshold is set, and feeders with fault probabilities greater than the fault probability threshold are selected as the final feeders. In this embodiment, the fault probability threshold is set to twice the reciprocal of the number of feeders. Considering that when there are no faulty feeders, the fault probabilities of all feeders are usually quite close in value, and since the sum of the fault probabilities of all feeders is 1, feeders with fault probabilities greater than the fault probability threshold are usually not found. When a feeder with a fault probability greater than the fault probability threshold appears, it indicates that the fault probability of the corresponding feeder is relatively prominent, which matches the characteristics of a faulty line. Therefore, feeders with fault probabilities greater than the fault probability threshold are selected as the final feeders.

[0052] In summary, a method for intelligent fault location in a low-current grounding distribution network system locks the transient signal extraction window using the instantaneous value of zero-sequence voltage, accurately extracts the high-frequency component signal of zero-sequence current corresponding to the high-frequency band level using discrete wavelet packet decomposition, and determines the morphological difference value based on the energy distribution deviation of the feeder compared to other feeders. This transforms the absolute magnitude discrimination into a comparison of frequency band distribution morphology, effectively eliminating the interference of high-frequency transmission errors of the underlying instrument transformers on the decision results. Combined with the total energy weight representing the physical scale of the signal, it can specifically identify and suppress morphological distribution artifacts caused by weak white noise, blocking the secondary transmission of measurement errors and noise fluctuations to the final characteristic value. By fusing the morphological difference value and the total energy weight, it achieves in-depth mining of the fault branch distribution structure characteristics and adaptive shielding of noise interference, resulting in higher accuracy for intelligent fault location in a low-current grounding distribution network system.

[0053] This invention also provides an intelligent fault location system for single-phase grounding faults in a low-current grounding system of a distribution network. Please refer to [link / reference]. Figure 2 The diagram shows a structural diagram of a single-phase grounding fault intelligent fault location system for a low-current grounding system in a power distribution network according to an embodiment of the present invention. The system includes: a data acquisition and preprocessing module 201, a morphological difference determination module 202, and a fault location module 203.

[0054] The data acquisition and preprocessing module 201 is used to acquire the instantaneous value of the zero-sequence voltage of the substation bus at each sampling time; determine the transient signal interception window based on the numerical limit characteristics of the instantaneous value of the zero-sequence voltage; and extract the original waveform signal of the zero-sequence current of each feeder from the transient signal interception window. The morphological difference determination module 202 is used to perform discrete wavelet packet decomposition on the original waveform signal of zero-sequence current to determine all high-frequency band levels and their corresponding high-frequency component signals of zero-sequence current for each feeder; based on the high-frequency component signals of zero-sequence current, the morphological difference value corresponding to each feeder is determined according to the energy distribution deviation of the high-frequency band levels of each feeder compared with all other feeders. The fault selection module 203 is used to determine the corresponding total energy weight based on the overall energy magnitude of all zero-sequence current high-frequency component signals corresponding to each feeder; to determine the single-phase grounding fault characteristic value of each feeder by fusing the total energy weight and the morphological difference value; and to select the single-phase grounding fault based on the single-phase grounding fault characteristic value.

[0055] It should be noted that the system provided in the above embodiments is only an example of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer equipment can be divided into different functional modules to complete all or part of the functions described above. In addition, the intelligent line selection system for single-phase grounding faults in a low-current grounding system of a distribution network and the intelligent line selection method for single-phase grounding faults in a low-current grounding system of a distribution network provided in the above embodiments belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0056] This invention also provides a computer device; please refer to [link / reference]. Figure 3 The illustration shows a schematic diagram of a computer device structure provided by an embodiment of the present invention. The computer device includes a memory 301, a processor 302, and a computer program 303 stored in the memory 301 and running on the processor 302. When the processor 302 executes the computer program 303, the computer device can execute any of the aforementioned intelligent fault location methods for single-phase grounding faults in a low-current grounding system of a power distribution network.

[0057] This invention also provides a computer program product that, when run on a computer device, enables the computer device to execute any of the aforementioned intelligent fault location methods for single-phase grounding faults in low-current grounding systems of power distribution networks.

[0058] This invention also provides a computer-readable storage medium storing computer program code. When the computer program code is run on a computer device, the computer device can execute any of the aforementioned intelligent fault location methods for single-phase grounding faults in a low-current grounding system of a power distribution network.

[0059] In the embodiments provided by the present invention, it should be understood that the computer device, computer program product and computer-readable storage medium provided are all used to execute the corresponding methods provided above, and therefore the beneficial effects they can achieve can be referred to the beneficial effects of the methods provided above, which will not be repeated here.

Claims

1. A method for intelligent fault location of a single-phase grounding fault in a low-current grounding system of a power distribution network, characterized in that, include: Obtain the instantaneous value of zero-sequence voltage of the substation bus at each sampling time; Based on the numerical limit-crossing characteristics of the instantaneous value of the zero-sequence voltage, the transient signal interception window is determined; Extract the original waveform signal of the zero-sequence current of each feeder from the transient signal capture window; Discrete wavelet packet decomposition is performed on the original waveform signal of the zero-sequence current to determine all high-frequency band levels and their corresponding high-frequency components of the zero-sequence current for each feeder; based on the high-frequency components of the zero-sequence current, the morphological difference value of each feeder is determined according to the energy distribution deviation of the high-frequency band levels of each feeder compared with all other feeders. The total energy weight is determined based on the overall energy of all zero-sequence current high-frequency component signals corresponding to each feeder. By combining the total energy weight with the morphological difference value, the single-phase ground fault characteristic value of each feeder is determined; Single-phase grounding fault line selection is performed based on the single-phase grounding fault characteristic values; The process of obtaining the morphological difference value includes: The zero-sequence current high-frequency component signal is integrally squared to determine the local frequency band energy value of each feeder at each high-frequency band level; the local frequency band energy values ​​of each feeder at all high-frequency band levels are superimposed to determine the corresponding total feeder energy value. At each high-frequency band level, the local frequency band energy values ​​corresponding to all other feeders outside each feeder are superimposed to determine the corresponding total background frequency band energy value. The total energy value of all feeders other than each feeder is superimposed to determine the corresponding total energy value of the background feeders. The local frequency band energy value is standardized by the total energy value of the feeder to determine the feeder energy distribution ratio of each feeder at each high frequency band level. The total energy value of the background frequency band is standardized by the total energy value of the background feeder to determine the background energy distribution ratio of each feeder in each high-frequency band level. For each feeder, the corresponding morphological difference value is determined based on the overall deviation between the feeder energy distribution ratio and the background energy distribution ratio corresponding to each high-frequency band level.

2. The intelligent fault location method for single-phase grounding faults in a low-current grounding system of a distribution network according to claim 1, characterized in that, The process of acquiring the transient signal interception window includes: The sampling time corresponding to the first instantaneous value of the zero-sequence voltage that is greater than the preset voltage start-up threshold is taken as the transient signal interception time. The sampling time corresponding to the preset buffer duration is traced back from the moment the transient signal is captured, and this is taken as the start time of the transient signal. The transient signal capture window is determined by extending the preset capture time length from the start time of the transient signal.

3. The intelligent fault location method for single-phase grounding faults in a low-current grounding system of a distribution network according to claim 1, characterized in that, The process of performing discrete wavelet packet decomposition on the original waveform signal of the zero-sequence current to determine all high-frequency band levels and their corresponding high-frequency components of the zero-sequence current for each feeder includes: Discrete wavelet packet decomposition is performed on the original waveform signal of the zero-sequence current to determine the zero-sequence current component signal corresponding to each frequency band level in the effective characteristic frequency band range. All frequency band levels within the effective characteristic frequency band range are arranged in descending order of their corresponding frequencies to determine the frequency band level sequence; the first preset number of frequency band levels in the frequency band level sequence are selected as high-frequency frequency band levels; and the zero-sequence current component signal corresponding to each high-frequency frequency band level is selected as the zero-sequence current high-frequency component signal.

4. The intelligent fault location method for single-phase grounding faults in a low-current grounding system of a distribution network according to claim 1, characterized in that, The process of obtaining the total energy weight includes: Each feeder is then used as the target feeder in turn; When the total energy value of the target feeder is greater than or equal to the preset energy baseline threshold, the corresponding total energy weight is set to the preset weight value. When the total energy value of the target feeder is less than the preset energy baseline threshold, the total energy value of the feeder is negatively correlated and mapped to determine the corresponding total energy weight; the total energy weight is greater than or equal to the preset weight value.

5. The intelligent fault location method for single-phase grounding faults in a low-current grounding system of a distribution network according to claim 1, characterized in that, The process of obtaining the characteristic values ​​of the single-phase ground fault includes: The characteristic value of a single-phase grounding fault is determined by multiplying the total energy weight with the morphological difference value.

6. The intelligent fault location method for single-phase grounding faults in a low-current grounding system of a distribution network according to claim 1, characterized in that, The process of selecting a single-phase ground fault line based on the single-phase ground fault characteristic value includes: The single-phase grounding fault characteristic values ​​are sequentially subjected to negative correlation mapping and normalization to determine the fault probability of each feeder; based on the fault probability, the final selected feeder is selected; and the single-phase grounding fault line selection is performed based on the final selected feeder.

7. The intelligent fault location method for single-phase grounding faults in a low-current grounding system of a distribution network according to claim 6, characterized in that, The process of obtaining the final selected feeder includes: The feeder with the highest probability of failure will be selected as the final feeder.

8. The intelligent fault location method for single-phase grounding faults in a low-current grounding system of a distribution network according to claim 4, characterized in that, The preset weight value is set to 1.

9. The intelligent fault location method for single-phase grounding faults in a low-current grounding system of a distribution network according to claim 3, characterized in that, The effective characteristic frequency band range is set to 1000Hz to 1200Hz.

Images