Traveling wave analyzing and positioning method for small-current grounding fault

Abstract

The invention relates to the technical field of ground fault analysis, in particular to a small-current ground fault traveling wave analyzing and positioning method, which comprises the following steps of: extracting a line mode current traveling wave signal by adopting modal transformation when detecting sudden change of a current traveling wave, and recording a fault triggering moment; extracting three-phase voltage traveling wave data after the fault triggering moment, performing modal transformation to obtain a line mode voltage traveling wave signal, and identifying polarity characteristics of the line mode voltage traveling wave signal near the fault triggering moment; and judging a fault direction, determining a fault section in combination with a distribution line topological structure, and calculating the position of a fault point. Compared with a traditional mode based on phasor abrupt change or differential calculation, the method has higher signal abrupt change sensitivity and anti-noise capability, and is especially suitable for typical working conditions with weak traveling wave signals, such as high-resistance grounding or non-metal grounding.

Description

Technical Field

[0001] This invention relates to the field of grounding fault analysis technology, and in particular to a traveling wave analysis method for locating low-current grounding faults. Background Technology

[0002] In power distribution systems, low-current grounding faults, such as single-phase high-resistance grounding and non-metallic grounding, may not immediately cause system tripping, but if they are not located and handled in a timely and accurate manner, they can easily cause resonant overvoltage, equipment insulation breakdown, or even evolve into multi-phase faults, seriously threatening the safety of the power grid. Because the current amplitude of low-current grounding faults is extremely small, the duration is short, and the traveling wave signal is weak, the fault information is easily drowned out by background noise, making it difficult for traditional location methods based on power frequency quantities to achieve effective fault location in complex power distribution networks, especially in scenarios with ring networks, distributed access, or many redundant branches.

[0003] In recent years, the traveling wave method, as an analytical approach based on the extraction of instantaneous high-frequency information from faults, has been increasingly applied to fault identification in transmission lines and some distribution networks due to its high positioning accuracy and fast response speed. However, existing traveling wave methods mostly rely on synchronous measurements at both ends or require high-precision communication and time synchronization systems, resulting in high costs and complex deployments. Furthermore, in practical applications, abrupt changes in traveling wave signals can easily become extremely weak due to high-impedance grounding, making wavefront identification difficult. Traditional amplitude judgment or time derivative algorithms are also sensitive to noise, leading to low wavefront detection accuracy. In addition, existing methods often lack quantitative means of determining the polarity of voltage traveling waves, relying on manual observation of waveform direction or simple threshold discrimination, which lacks mathematical rigor and engineering feasibility.

[0004] More importantly, the current power distribution network structure is becoming increasingly complex, with multi-level topology, distributed power source access, and multi-point measurement control characteristics. Without clear directional constraints and structural diagram support, it is difficult to directly determine the fault section by relying solely on the traveling wave distance value, which can easily lead to positioning errors such as "correct distance measurement but misaligned section". Summary of the Invention

[0005] This invention provides a traveling wave analysis method for locating low-current grounding faults, which integrates polarity analysis, topology information and precise wavefront identification into a single-end fault location method, thereby improving the accuracy and real-time performance of low-current grounding fault identification.

[0006] A traveling wave analysis method for locating low-current grounding faults includes the following steps: S1. Initial Fault Traveling Wave Detection and Time Calibration: Monitor the three-phase current signal of the power distribution line. When a sudden change in the current traveling wave is detected, use mode transformation to extract the line mode current traveling wave signal and record the fault triggering time corresponding to the first time the line mode current traveling wave signal exceeds the preset amplitude threshold. S2. Voltage traveling wave polarity feature extraction: Based on the fault triggering time, extract the three-phase voltage traveling wave data after the fault triggering time, perform mode transformation to obtain the line-mode voltage traveling wave signal, and identify the polarity feature of the line-mode voltage traveling wave signal near the fault triggering time. S3. Fault Section Identification and Location: Determine the fault direction based on the polarity characteristics of the line-mode voltage traveling wave signal, determine the fault section in combination with the power distribution line topology, and calculate the location of the fault point using the time difference between the first and second wavefronts of the line-mode current traveling wave signal.

[0007] Optionally, S1 acquires three-phase current signals in real time through a traveling wave sensor installed at the beginning of the line; it analyzes the filtered three-phase current signals using continuous wavelet transform, and determines that a current traveling wave mutation has occurred when the wavelet coefficient modulus exceeds a preset start-up threshold.

[0008] Optionally, when analyzing the filtered three-phase current signal using continuous wavelet transform, the method further includes bandpass filtering of the acquired three-phase current signal to enhance the traveling wave component and suppress power frequency interference.

[0009] Optionally, after the current traveling wave abruptly changes, a mode transformation based on Clarke transform is applied to the three-phase current signal to extract the line-mode current component as the line-mode current traveling wave signal; the amplitude of the line-mode current traveling wave signal is monitored in real time, and when the amplitude exceeds the preset amplitude threshold obtained based on historical normal operation data for the first time, the current moment is recorded as the fault trigger moment.

[0010] Optionally, in step S2, the three-phase voltage transient data within a preset time window are extracted, starting from the fault triggering time. The preset time window covers the complete initial traveling wave response process.

[0011] Optionally, S2 includes applying Clarke transform to the three-phase voltage transient data to calculate the line-mode voltage component as a line-mode voltage traveling wave signal; calculating the polarity integral value of the line-mode voltage traveling wave signal within the integral analysis interval starting at the fault triggering time; and determining the polarity characteristics based on the positive and negative characteristics of the polarity integral value.

[0012] Optionally, determining the polarity characteristic based on the positive or negative nature of the polarity integral value specifically includes: When the integral value is negative, the line-mode voltage traveling wave signal is determined to exhibit negative polarity characteristics. When the integral value is positive, the line-mode voltage traveling wave signal is determined to exhibit positive polarity characteristics.

[0013] Optionally, determining the fault direction based on the polarity characteristics of the line-mode voltage traveling wave signal specifically includes: When the polarity characteristic is negative, the fault is determined to be located in the positive direction of the measurement point; When the polarity characteristic is positive, the fault is determined to be located in the opposite direction of the measurement point.

[0014] Optionally, based on the judgment result of the fault direction, combined with the pre-stored power distribution line topology data, a graph theory traversal algorithm is used to determine a set of specific sections where the fault occurred, and a candidate fault section set is constructed.

[0015] Optionally, S3 further includes identifying the arrival times of the first and second wavefronts in the line-mode current traveling wave signal using the wavelet transform modulus maxima method; calculating the time difference between the arrival times of the first and second wavefronts; calculating the distance between the fault point and the measurement point using the single-end ranging method combined with the time difference; projecting the calculated distance value onto the cumulative length coordinate axis of the candidate fault segment set; determining the specific branch segment into which the distance value falls according to the distance matching principle; and determining the current branch segment as the actual location of the fault if it falls within the length range from the start to the end of a certain branch segment.

[0016] The beneficial effects of this invention are: This invention deploys traveling wave sensors at measurement points to acquire three-phase current signals in real time. After applying bandpass filtering to the signals in a specific frequency band, continuous wavelet transform is used for abrupt change detection. Then, Clarke transform is used to extract the traveling wave component of the line-mode current. By monitoring the amplitude of abrupt changes based on historical statistical thresholds, precise triggering of weak grounding traveling wave signals is achieved. Compared with traditional methods based on phasor abrupt changes or differential calculations, this invention has higher signal abrupt change sensitivity and noise immunity, and is especially suitable for typical working conditions with weak traveling wave signals, such as high-resistance grounding or non-metallic grounding.

[0017] This invention proposes an integral polarity criterion based on line-mode voltage traveling wave signals, transforming subjective waveform observation into a quantifiable, sign-consistent polarity judgment mechanism, and clearly establishing a one-to-one correspondence between polarity characteristics and fault direction. Based on this directional result, combined with the distribution network topology diagram, a graph theory traversal algorithm is used to limit the search range of fault sections, filtering only on directional matching paths. This improves the automation, structural adaptability, and computational efficiency of fault location algorithms in complex distributed distribution networks, avoids large-scale path redundancy and misjudgments, and possesses good engineering scalability.

[0018] This invention introduces the wavelet transform modulus maxima method to perform multi-scale analysis on the line mode current traveling wave signal, accurately extracting the arrival times of the first wave and reflected wave, and calculating the fault distance using the single-end traveling wave ranging formula. Combined with a set of directionally limited sections, it achieves consistent fusion positioning of ranging results and structural paths. Compared with traditional ranging methods that rely solely on distance values ​​for rough judgment, this solution improves the accuracy, error resistance, and traceability of ranging and positioning through bidirectional verification of "section constraints + precise distance calculation". Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this 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 for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the positioning method according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating fault segment identification and location according to an embodiment of the present invention. Detailed Implementation

[0021] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. For some well-known technologies, those skilled in the art may also use other alternative methods to implement the invention. Moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0022] like Figures 1-2 As shown, a traveling wave analysis method for locating a low-current grounding fault includes the following steps: S1. Initial Fault Traveling Wave Detection and Time Calibration: Monitor the three-phase current signal of the power distribution line. When a sudden change in the current traveling wave is detected, use mode transformation to extract the line mode current traveling wave signal and record the fault triggering time corresponding to the first time the line mode current traveling wave signal exceeds the preset amplitude threshold.

[0023] S11. Install a three-phase traveling wave sensor at the beginning of the power distribution line to collect three-phase traveling wave current signals in real time. .

[0024] S12. Bandpass Filtering Enhances Traveling Wave Characteristics: A bandpass filter with a specific frequency band is applied to the acquired three-phase traveling wave current signal to obtain an enhanced signal. It is used to suppress power frequency components and low-frequency interference, and improve the detectability of traveling wave abrupt changes. The filter bandwidth is selected as the typical traveling wave frequency band.

[0025] In normal operation, the three-phase signals in a power distribution line are mainly 50Hz power frequency signals. However, when a ground fault occurs, high-frequency traveling wave components are excited in the system. These traveling wave components propagate rapidly and change drastically at the moment of the fault, making them suitable for rapid location. However, in addition to the useful high-frequency traveling wave components, the original current signal also contains a large number of power frequency fundamental waves, power frequency harmonics, environmental noise, and low-frequency oscillation components caused by capacitance and inductance distribution effects. If left unprocessed, these irrelevant components will severely mask or interfere with the identification of traveling wave abrupt changes.

[0026] By designing a bandpass filter adapted to the frequency range of traveling waves, the main energy of the traveling wave signal can be preserved. Traveling waves are generally concentrated between 5kHz and 30kHz (the specific range depends on system parameters and sensor sampling rate). The filter is set to only pass signals in this frequency band to ensure that the traveling wave components are not attenuated. The center frequency and bandwidth of the filter are selected as follows: based on system simulation and field experiments, a center frequency of around 15kHz and a bandwidth of 10–25kHz are chosen to cover the main traveling wave components while avoiding harmonic interference. The filter type can be an IIR or FIR filter, depending on the balance between real-time performance and amplitude-frequency response. S13. Continuous Wavelet Transform Detection of Abrupt Changes: A continuous wavelet transform (CWT) is performed on the filtered three-phase signal. The Morlet function is selected as the wavelet function, expressed as: , ;in, Indicates the first Phase signal in scale With translation The wavelet coefficients below, For the mother wavelet function, Indicates conjugate; The above formula means that for each phase current signal (phase A, phase B, phase C), it is first obtained after bandpass filtering. Then, the signal is subjected to continuous wavelet transform to obtain the result at the scaling factor. With translation factor Transformation coefficients under The core of the transformation is to convolve and match the signal with a scaled and shifted mother wavelet function, and the output is... This indicates the degree of similarity of signals at that scale and location. It is the mother wavelet The conjugate form, first scaled, with a scale of Then translate it again, and the position is The integration operation matches the signal with the wavelet over the entire time range to obtain local features; in layman's terms, this process is like sliding a wavelet template of variable shape across the signal to find the location and time point where the shape matches. If the wavelet coefficients... A sudden increase indicates a significant abrupt change or frequency component shift in the signal at that moment.

[0027] When there exists a coefficient modulus of a certain phase Exceeding the preset startup threshold When this occurs, it is determined to be a sudden change in the current traveling wave, which is represented as: A traveling wave abrupt change was detected; this formula indicates that there exists at least one phase (phase A, phase B, or phase C) such that the magnitude of its continuous wavelet transform coefficients changes at a certain scale. and location If the current exceeds the preset trigger threshold, it is determined that a traveling wave abrupt change event has occurred in the current signal. Preset trigger threshold This is a criterion used for sudden change identification, based on historical statistics. Its setting must be able to sensitively respond to sudden fault changes without being falsely triggered by normal fluctuations. This is achieved by collecting multiple sets of three-phase current signals from different time periods under fault-free conditions, performing wavelet transform processing on them in the same manner as the main process, and statistically analyzing the modulus of all wavelet coefficients. The maximum value, mean, and standard deviation are set, and a preset threshold is set to the mean plus a certain multiple of the standard deviation: ;in, This represents the mean value of the wavelet modulus under normal operating conditions. Standard deviation, This is an empirical coefficient, ranging from 3 to 5.

[0028] S14. Modal transformation to extract line-mode current traveling wave signal: at the moment of detection of abrupt change. The original three-phase current signal is subjected to Clarke transform to extract its line-mode component: ; in, This represents the traveling wave signal component of the line-mode current. It represents the zero-modulus component.

[0029] The main purpose of the Clarke transform is to transform the current signal in a three-phase symmetrical system. Transform it into two mutually orthogonal components: The α-axis projection (similar to the x-axis direction) represents the linear modulus component; The β-axis projection (similar to the y-axis direction) represents the linear mode component; This represents the common component of the three-phase current, namely the zero-mode component.

[0030] This transformation can remove the redundancy of the three-phase signal, making the analysis process simpler and easier to detect characteristic changes. It is particularly suitable for handling the rapid changes in traveling wave signals.

[0031] The specific transformations are as follows: Three-phase current Space vector mapping: The three-phase current is regarded as three vectors, pointing to three axes at 120° angles respectively; the combination of these axes can be used to construct a two-dimensional coordinate system.

[0032] Projection Formation Axis: Through mathematical mapping, the combination of these three vectors is "projected" onto a two-dimensional plane, becoming a pair of mutually perpendicular α and β components; that is, the three oblique forces are decomposed into the resultant forces in the horizontal and vertical directions.

[0033] Redundancy removal and feature extraction: The original three-phase signal contains redundancy (the sum of the three-phase currents is zero). The Clarke transform preserves the main dynamic components (α, β) and extracts features through the zero-mode component. Characterizes symmetry or imbalance.

[0034] In the fault detection process of this invention, the Clarke transform can extract the traveling wave signal of the line-mode current, i.e. and These two components represent the components of the three-phase current that truly change rapidly over time and reflect abrupt changes. They assist in subsequent amplitude monitoring and wavefront identification, including the calculation of the line mode amplitude. , as a mutation marker.

[0035] S15. Fault Trigger Timing Calibration: After completing the mode transformation of the three-phase current, two components of the line-mode current are obtained: the component of the line-mode current in the α direction. The component of the line-mode current in the β direction To monitor for any sudden changes in the current traveling wave, these two components are combined into a single overall line-mode current amplitude function. The line-mode current amplitude function represents the intensity of the overall line-mode current at the current moment. This amplitude function remains at a low level when there is no fault, but when a ground fault occurs in the system and a traveling wave is generated, It will rise rapidly at a certain point in time. To determine the exact time of the mutation, it is necessary to... Perform real-time monitoring and define an amplitude threshold. When the amplitude first exceeds the threshold, the corresponding moment is identified as the fault trigger moment. This refers to the point in time when the line-mode current first shows a significant anomaly, serving as a reference point for subsequent voltage polarity analysis; specifically as follows: [The text then abruptly shifts to a different topic:] The line-mode current amplitude function... Conduct real-time monitoring; ;in, This indicates that a preset amplitude threshold is obtained based on statistical analysis of historical normal operation data. A large amount of three-phase current data is collected from the system under normal operation, and after Clarke transformation, its line mode amplitude sequence is calculated. Calculate the mean of this set of data. and standard deviation Set a threshold value equal to the mean plus a certain number of standard deviations to cover the normal fluctuation range: ;in This is an empirical constant, ranging from 3 to 5, representing the degree of outlier. The threshold is defined as the fault trigger moment, which is the time when the line-mode current first exceeds the threshold.

[0036] S2. Voltage traveling wave polarity feature extraction: Based on the fault triggering time, extract the three-phase voltage traveling wave data after the fault triggering time, perform mode transformation to obtain the line-mode voltage traveling wave signal, and identify the polarity feature of the line-mode voltage traveling wave signal near the fault triggering time.

[0037] S21. Extract transient voltage data, including data from the fault trigger time. Starting from this point, extract the three-phase voltage signal within a preset time window of 200 microseconds, and denote it as: , .

[0038] When a ground fault occurs, a rapidly propagating voltage traveling wave signal is generated in the power distribution line. The initial response of these traveling waves typically occurs within tens to hundreds of microseconds after the fault is triggered. Choosing a 200-microsecond time window can fully cover the arrival of the initial traveling wavefront, the main polarity change segment, and the main response portion before initial reflection or scattering has a significant impact. This helps extract the most original and representative polarity information. As the voltage traveling wave propagates in the power distribution line, it encounters impedance discontinuities such as branches, switches, and grounding points, causing reflection and refraction. Generally, after 200 microseconds, these reflected waves gradually superimpose. If the time window is too long, it will introduce a large amount of secondary wave interference, affecting the accuracy of polarity determination.

[0039] S22. Mode transformation to extract line-mode voltage components: The above three-phase voltage signals should be subjected to Clarke transformation to obtain the line-mode voltage components. ,Right now: ; in, It is a line-mode voltage traveling wave signal. The α component represents the line-mode voltage. This represents the β component of the line-mode voltage. It is the zero-modulus component.

[0040] S23. Calculate the polarity integral of the line-mode voltage traveling wave, including: S231. The selected interval for integral analysis is: ; The integration analysis interval is set from 10 microseconds before the fault trigger to 50 microseconds after the fault. The first 10 microseconds serve as a baseline reference before the fault. Before the fault occurs, the voltage traveling wave signal is generally in a relatively stable or low-amplitude disturbance state. Setting the first 10 microseconds before the trigger as the integration starting point helps to introduce a reference segment for comparing subsequent abrupt changes. The introduction of this baseline segment can improve the sensitivity of the integration judgment to "directional abrupt changes" and help highlight the sign characteristics of the wavefront polarity change. The last 50 microseconds cover the main traveling wavefront response. After a fault occurs, the voltage traveling wave will propagate on the line at an extremely fast speed. The wavefront generally arrives at the measurement point within tens of microseconds. Setting the integration endpoint 50 microseconds after the fault trigger can completely cover the first voltage traveling wavefront and its polarity change process. This time range can avoid the mixing of subsequent reflected waves and ensure that the integration result reflects the most original polarity response. At the same time, for a typical 1MHz sampling rate system, the 60 sampling points from the first 10 microseconds to the last 50 microseconds are sufficient to perform integration calculation and trend discrimination.

[0041] S232, Calculate the polarity integral value of the combined amplitude of the line-mode voltage. : express The sign function of the component: when The time is +1. When it is -1, This represents the combined amplitude of the line-mode voltage (i.e., the magnitude of the space vector). Or, to simplify the form, considering only the principal axis, it can be represented as: ; If we take The main direction (the circuit structure makes this component the most sensitive) can be directly integrated; if robustness is required, a weighted integral of the synthesized amplitude plus the sign function can be used.

[0042] The core of the S23 scheme is to identify the polarity characteristics of the voltage traveling wave in the early stage of the fault, so as to determine the relative direction (positive or negative) of the ground fault at the measurement point. Its main idea is to select a high-confidence time interval near the fault triggering time, perform integral calculation on the line-mode voltage signal, and quantify its polarity by the positive or negative value of the integral.

[0043] The three-phase voltage signal is transformed by Clarke to obtain two line-mode components, which constitute a traveling wave vector in two-dimensional space, reflecting the true voltage propagation characteristics. Within this time interval, the signal is integrated to quantify its overall trend. By determining the sign of the integral value, the wavefront is identified as either "positive" or "negative," thus aiding in subsequent determination of the fault's direction at the measurement point (e.g., whether it propagates forward or to the end of the line).

[0044] Two integration methods: Method 1: Integrate the sum of the amplitude and the sign of the α component, first calculate the modulus of the line-mode voltage in two-dimensional space (α-β) as the actual energy amplitude; then use... The sign of the is used as a directional weighting factor, and the integral is performed over the entire synthesized amplitude. Its directionality is more explicit, thanks to... It strengthens the dominant direction, has good anti-interference capabilities, and makes comprehensive use of resources. Two-axis information is less susceptible to interference from noise in a single component.

[0045] Method 2: Directly integrate only the α component, directly... Integrating is performed, assuming that the direction is sufficient to represent the main trend of the traveling wave. The calculation is fast and applicable to scenarios with a simple line structure and a clear dominant direction. A clear dominant direction means that in power lines, the main energy of the fault traveling wave signal is concentrated in a certain direction or component during the propagation process, and this direction can stably reflect the actual propagation trend of the traveling wave under most operating conditions.

[0046] S24. Define the polarity characteristic judgment criterion, based on the polarity integral value. Determining the polarity of a line-mode voltage traveling wave signal by its positive and negative characteristics: .

[0047] When a small-current ground fault occurs in a power distribution line, a voltage traveling wave propagates in a certain direction from the fault point to both sides. According to traveling wave theory: If the fault point is upstream of the measuring point, the first traveling wave front received at the measuring point will usually show a negative polarity jump. If the fault point is downstream of the measuring point, the first traveling wave head will usually exhibit a positive polarity jump.

[0048] Therefore, the polarity of the first wave of the traveling wave received at the measuring point can reflect the position and direction of the fault relative to the measuring point, and is one of the core bases for judging the fault section.

[0049] Traditional methods rely on manual observation of waveform fluctuations to determine polarity, which is subjective. However, in S2, by integrating the voltage traveling wave signal over a certain time interval, the "directional change" of the waveform can be represented by an integral value with a positive or negative sign. To quantify: If the waveform jumps upwards overall within the integration interval (stronger energy in the positive direction), then the integration result is positive; If the overall waveform drops sharply downwards (negative energy dominates), then the integral value is negative.

[0050] This design transforms the original subjective judgment process of "observing waveforms" into a programmable and quantifiable mathematical process, making it suitable for embedded fault criterion implementation.

[0051] Compared with direct sampling point comparison or instantaneous slope discrimination, the integral operation has smoothing characteristics and can effectively filter high-frequency random noise, multiple small fluctuations, and instantaneous disturbances caused by non-fault factors. Therefore, using the positive or negative value of the integral as the basis for polarity judgment not only reflects the essence of the trend, but also enhances the anti-interference ability and stability.

[0052] Based on this, clear judgment criteria were designed: Positive polarity; therefore, the fault is located downstream of the measuring point. If the polarity is negative, the fault is located upstream of the measuring point.

[0053] S3. Fault Section Identification and Location: Determine the fault direction based on the polarity characteristics of the line-mode voltage traveling wave signal, determine the fault section in combination with the power distribution line topology, and calculate the location of the fault point using the time difference between the first and second wavefronts of the line-mode current traveling wave signal.

[0054] S31. Fault Direction Determination: Based on the polarity characteristics of the line-mode voltage traveling wave signal identified in S2, determine the fault direction: When the polarity characteristic is negative, the fault is located in the positive direction of the measurement point (the direction of the end of the line). When the polarity characteristic is positive, the fault is located in the opposite direction of the measurement point (the direction of the line source end). Note: This directionality is used to constrain fault search paths and improve section identification efficiency. In complex distribution networks, a single measurement point typically connects to multiple branch lines. Without direction determination, it would be necessary to traverse all possible directions, increasing the number of path combinations and computational load, and easily introducing interference from irrelevant sections. Introducing "fault directionality" can immediately eliminate paths that do not conform to the direction, limiting the search scope. Only half or fewer network graph nodes and branches need to be traversed, reducing computational complexity.

[0055] S32. Topology-assisted localization: Based on the above fault direction judgment results, call the pre-stored power distribution line topology data, including nodes, branches, cable segments, and branch relationships. Starting from the measurement point, execute the graph theory traversal algorithm under the direction constraint to identify candidate line segments that match the fault direction. Output: The set of candidate fault sections where the fault is most likely to occur. Each segment has a clear start and end point and electrical parameters. This set is used for subsequent matching of ranging results and is the key to limiting the search space and improving positioning accuracy.

[0056] S32, through directional and topological analysis, eliminates irrelevant branches, resulting in... It is a set of candidate fault segments, representing "where the fault might occur." However, since directionality can only indicate "ahead" or "behind," and cannot provide precise distance, it is not yet possible to determine the specific segment. Subsequently, S33 calculates an actual physical distance through time difference analysis of the traveling wavefront. It needs to be combined with this in the future. The length and topological location of each segment are matched one by one to ultimately determine the segment that actually experienced the fault from the "possible segments". Without the segment filtering in S32, the distance calculated in S33 would be... Matching across dozens of segments in the entire network would be time-consuming and prone to errors. By using S32's directional path constraints, distance matching can be performed only in one direction and within a limited number of branch lines. (Segment set) This provides a candidate pool for comparison in subsequent ranging measurements, and the resulting distance values... It is a point value, by... Projecting the cumulative length coordinates onto each candidate segment determines which branch it falls into, and that branch is the finally located fault segment.

[0057] The pre-stored power distribution line topology data is a structured model of the entire power distribution network, stored in a graph structure, and mainly includes the following fields: 1. Node Information: Each node represents a physical connection point, such as a switch, transformer, branch point, measurement point, etc. Typical fields include: Node ID (unique identifier); Node types (trunk node, branch node, terminal node); Feeder number; Coordinates.

[0058] 2. Branch Information: Each edge (branch) represents a segment of actual conductor, cable or equipment connection. Typical fields include: starting node ID, ending node ID, cable segment number, length, electrical parameters, whether it is energized / whether it is passable.

[0059] 3. Measurement point mapping information: includes the number and name of the node where the current / voltage sensor is installed, used to determine the starting point of fault analysis.

[0060] The graph-based traversal algorithm, based on directional constraints, selects candidate fault segments by starting with the fault measurement point and combining the fault directionality information (forward or reverse). Step 1: Initialization: Read the measurement point node ID: ; The candidate path set is initialized to empty; Set up a traversal queue or stack; Determine the directionality (obtained from voltage polarity): If it is a forward fault, traverse the downstream of the measuring point; if it is a reverse fault, traverse the upstream of the measuring point.

[0061] Step 2: Graph structure orientation modeling: The topology graph can be represented as a directed graph or a bidirectional graph; if the direction is not specified, the current flow direction or topological hierarchy relationship can be manually set as a guide for traversal direction; each edge is marked with a direction attribute.

[0062] Step 3: Perform a traversal, using breadth-first search, starting from... start: a joins the team Marked as visited; b leaves the current node Find the next-hop node that is connected to it and meets the direction requirements. ; c. If the cable segment or branch corresponding to the edge is a passable path, then add it to the candidate segment set; d will Join the queue and continue the traversal; The traversal depth is set to an upper limit, which is to set the maximum allowed transmission distance to avoid redundancy, based on the traveling wave propagation speed. The maximum time window that wave head can recognize Estimate the maximum allowable transmission distance .

[0063] Step 4: Construct a candidate segment set Each time a branch that meets the directional requirements and has normal connections is passed, it is recorded as a candidate fault section, and the following information is saved: Start and end node numbers; Cable number, electrical parameters; Relative topological distance to the measuring point; Are they downstream of the same feeder? You can also select by measurement point distance or network level. Prioritize and optimize subsequent fault distance matching.

[0064] S33. Traveling wave front identification: Line mode current components are obtained in S1 through Clarke transform. However, these only convert the three-phase current signal into an orthogonal line-mode coordinate system; essentially, it is still a time-domain signal. Therefore, it is necessary to analyze the traveling wave component of the line-mode current at the measurement point. Wavelet transform analysis was performed using the Modulus Maxima method to simultaneously identify true wavefront abrupt change points at multiple scales, eliminating spurious peaks. This included identifying the arrival times of the first obvious abrupt change point and the first subsequent reflection abrupt change point. The arrival time of the first wavefront; The arrival time of the second wavefront (reflected wave); The wavelet mode maxima are abrupt change points that occur simultaneously on multiple scales, representing the wavefront position of the traveling wave signal of the line mode current. It is a feature that has already removed the three-phase redundancy, while retaining the main abrupt change components. Performing wavelet transform on them is more focused and efficient than directly processing the original three-phase signal.

[0065] The first wavefront: indicates that the traveling wave arrives at the measurement point for the first time from the fault point, carrying the "original waveform information" of the fault; The second wavefront: is the waveform reflected back from the fault point and reaches the measurement point again; The time difference between the two reflects the "going-back" propagation time of the traveling wave and can be used for distance measurement.

[0066] The modulus maxima method observes a time-domain signal at different scales, which is equivalent to "scanning" the signal with magnifying glasses of different resolutions to capture abrupt changes. The smaller the scale, the higher the resolution. The output is "wavelet coefficients", and the magnitude (absolute value) of the coefficients reflects the degree of abrupt changes in the signal. At the abrupt changes in the signal (such as a sharp change in current or voltage), the wavelet modulus will show significant peaks at multiple scales. Therefore, the co-occurring maxima at multiple scales are the high-confidence traveling wave front positions, which can be considered as the abrupt change points being confirmed by multiple resolutions.

[0067] The specific steps are as follows: Step 1: To Perform wavelet transform to obtain a set of wavelet coefficients that vary with time and scale. Select the mother wavelet to adapt to the characteristics of sudden current changes.

[0068] Step 2: Find the modulus maxima: Along the time axis, find the local peaks of the modulus curve at different scales; mark the maxima that coincide or are close in position at multiple scales to form a "modulus maxima chain", and take the starting time of the first modulus maxima chain as the arrival time of the first wavefront; continue searching within a few microseconds thereafter to find the next stable modulus maxima, which is defined as the arrival time of the second wavefront (reflected wave).

[0069] Step 3: Remove spurious maxima: Set a modulus intensity threshold to eliminate small peaks caused by noise. Maximum values ​​must appear simultaneously on at least three scales; if the positional difference between scales exceeds a certain range, they are discarded. The wavelet modulus is normalized, and the modulus intensity threshold is set to 0.2–0.3. A threshold that is too low will misjudge many small fluctuations caused by noise as wavefronts, leading to misidentification or repeated calibration; a threshold that is too high may miss weak traveling wave signals such as high-impedance grounding, making wavefront extraction impossible.

[0070] S34. Calculate the propagation time difference between wavefronts: This time difference reflects the total time it takes for the traveling wave to propagate from the measurement point to the fault point and back.

[0071] S35. Calculate the physical distance between the fault point and the measurement point using the single-end ranging method. ;in, This indicates the distance from the measuring point to the fault point. This indicates the propagation speed of the traveling wave in the circuit. The time difference between the arrival times of the two wavefronts is represented by the "2" in the denominator, indicating that the traveling wave propagation process is a round-trip path of "going + returning".

[0072] When a ground fault occurs in a power distribution line, it will generate transient high-frequency traveling waves of current or voltage. These traveling waves will travel at a speed of It propagates along the route to both ends.

[0073] The traveling wave propagates from the fault point to both sides; at a distance from the fault point... A traveling wave sensor is installed at a certain measurement point (at the substation end); the time it takes for the traveling wave to propagate from the fault point to the measurement point is... Upon reaching the measuring point, the wave will be reflected at the busbar side or switch node, forming a reflected wave. The reflected wave will then propagate back to the fault point and be reflected back, forming a second wavefront, which will then reach the measuring point again.

[0074] Therefore, the time difference between the two wave heads In fact, it includes traveling waves: measuring point Fault point (one way) and fault point The total time spent in the two stages of measuring the point (reflection back): ; Due to the propagation speed is According to the formula for uniform linear propagation: The above distance measurement formula is obtained.

[0075] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0076] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A traveling wave analysis method for locating low-current grounding faults, characterized in that, Includes the following steps: S1. Monitor the three-phase current signal of the power distribution line. When a sudden change in the current traveling wave is detected, use mode transformation to extract the line mode current traveling wave signal and record the fault triggering time corresponding to the first time the line mode current traveling wave signal exceeds the preset amplitude threshold. S2. Based on the fault triggering time, extract the three-phase voltage traveling wave data after the fault triggering time, perform mode transformation to obtain the line-mode voltage traveling wave signal, and identify the polarity characteristics of the line-mode voltage traveling wave signal near the fault triggering time. S3. Determine the fault direction based on the polarity characteristics of the line-mode voltage traveling wave signal, determine the fault section in combination with the power distribution line topology, and calculate the location of the fault point using the time difference between the first and second wavefronts of the line-mode current traveling wave signal.

2. The traveling wave analysis and location method for low-current grounding faults according to claim 1, characterized in that, The S1 acquires three-phase current signals in real time through a traveling wave sensor installed at the beginning of the line; The filtered three-phase current signal is analyzed by continuous wavelet transform. When the wavelet coefficient magnitude exceeds the preset start threshold, it is determined that a sudden change in the current traveling wave has occurred.

3. The traveling wave analysis and location method for low-current grounding faults according to claim 2, characterized in that, When analyzing the filtered three-phase current signal using continuous wavelet transform, the method also includes bandpass filtering of the acquired three-phase current signal to enhance the traveling wave component and suppress power frequency interference.

4. The traveling wave analysis and location method for low-current grounding faults according to claim 2, characterized in that, After the current traveling wave abruptly changes, a mode transformation based on Clarke transform is applied to the three-phase current signal to extract the line-mode current component as the line-mode current traveling wave signal; the amplitude of the line-mode current traveling wave signal is monitored in real time, and when the amplitude exceeds the preset amplitude threshold obtained based on historical normal operation data for the first time, the current moment is recorded as the fault trigger moment.

5. The traveling wave analysis and location method for low-current grounding faults according to claim 1, characterized in that, In step S2, the three-phase voltage transient data within a preset time window are extracted starting from the fault triggering time. The preset time window covers the complete initial traveling wave response process.

6. The traveling wave analysis and location method for low-current grounding faults according to claim 5, characterized in that, S2 includes applying Clarke transform to the three-phase voltage transient data to calculate the line-mode voltage component as the line-mode voltage traveling wave signal; calculating the polarity integral value of the line-mode voltage traveling wave signal within the integral analysis interval starting at the fault triggering time; and determining the polarity characteristics based on the positive and negative characteristics of the polarity integral value.

7. The traveling wave analysis and location method for low-current grounding faults according to claim 6, characterized in that, The determination of polarity characteristics based on the positive or negative nature of the polarity integral value specifically includes: When the integral value is negative, the line-mode voltage traveling wave signal is determined to exhibit negative polarity characteristics. When the integral value is positive, the line-mode voltage traveling wave signal is determined to exhibit positive polarity characteristics.

8. The traveling wave analysis and location method for low-current grounding faults according to claim 1, characterized in that, The method of determining the fault direction based on the polarity characteristics of the line-mode voltage traveling wave signal specifically includes: When the polarity characteristic is negative, the fault is determined to be located in the positive direction of the measurement point; When the polarity characteristic is positive, the fault is determined to be located in the opposite direction of the measurement point.

9. The traveling wave analysis and location method for low-current grounding faults according to claim 8, characterized in that, Based on the fault direction determination result, combined with the pre-stored power distribution line topology data, a graph theory traversal algorithm is used to determine a specific set of sections where the fault occurred, and a candidate fault section set is constructed.

10. The traveling wave analysis and location method for low-current grounding faults according to claim 9, characterized in that, S3 further includes identifying the arrival times of the first and second wavefronts in the traveling wave signal of the line-mode current using the wavelet transform modulus maxima method; calculating the time difference between the arrival times of the first and second wavefronts; calculating the distance between the fault point and the measurement point using the single-end ranging method combined with the time difference; projecting the calculated distance value onto the cumulative length coordinate axis of the candidate fault segment set; determining the specific branch segment into which the distance value falls according to the distance matching principle; and determining the current branch segment as the actual location of the fault if it falls within the length range from the start to the end of a certain branch segment.

Images