A high-voltage transmission line fault detection and identification method

Abstract

The present application belongs to the technical field of fault detection, and relates to a high-voltage transmission line fault detection and identification method. The present application synchronously collects steady-state and transient traveling wave signals of key nodes of a power grid, divides candidate fault sections according to a topological structure; further utilizes a multi-feature fusion criterion to identify faults and extract traveling wave features; performs steady-state section discrimination based on pattern recognition and transient positioning based on multi-terminal ranging and polarity verification in parallel, to generate two types of positioning result sets; finally, intelligently fuses and judges the two types of results under topological constraints. The method solves the problems that a traditional single-point signal analysis method cannot determine a fault section, lacks spatial positioning capability and is insufficient in adaptability, realizes accurate positioning of a fault section, and significantly improves positioning accuracy, robustness and fault handling efficiency.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of fault detection and relates to a high-voltage transmission line fault detection and identification method. BACKGROUND

[0002] As an important part of the power system, high-voltage transmission lines have higher requirements for fault detection and positioning technology with the continuous expansion of the power system scale and the in-depth promotion of smart grid construction. Traditional methods mostly rely on analyzing the electrical signals such as current time series collected by a single monitoring node in the line and detecting the abnormal characteristics of the signals to realize fault monitoring.

[0003] A typical existing scheme, such as the high-voltage line fault monitoring system for electrical engineering proposed in Chinese patent CN119335299B, improves the abnormal detection algorithm based on a histogram by analyzing the surge characteristics, harmonic interference, and data mutation of the current signal to improve the identification accuracy of abnormal states of mine-used high-voltage lines.

[0004] However, this kind of method based on single-point signal analysis still has the following main defects: first, this method only evaluates the abnormality of the current time series data of a single monitoring node, and its judgment is whether there is a fault, not where the fault is. In an actual power grid with a multi-branch structure, since the topological connection relationship of the power grid is not integrated, the electrical quantity abnormality cannot be mapped to a specific physical line section, resulting in a lack of spatial pertinence for fault isolation operation.

[0005] Secondly, the core of this method is to extract features and score abnormalities of the electrical quantity of a single monitoring node, and it does not involve synchronous collection and joint analysis of signals generated by multiple monitoring nodes in transient processes. Therefore, it cannot realize distance estimation based on time-based signal propagation differences, nor can it determine the fault direction by means of spatial comparison of the initial polarity of the signal, and it is essentially difficult to have the ability to spatially locate the fault point.

[0006] Although this method introduces harmonic, surge and other features to improve abnormal scoring, it still belongs to a single-stage discrimination mode of feature-scoring, has the defects of fixed criteria and weak adaptability, and is difficult to adapt to the complex changes of different fault types and power grid operation modes. SUMMARY

[0007] In view of this, to solve the problems proposed in the background art, a high-voltage transmission line fault detection and identification method is proposed.

[0008] The object of the application can be achieved by the following technical solutions: A high-voltage transmission line fault detection and identification method, comprising: S1, acquiring steady-state electrical quantities and transient traveling wave signals synchronously collected by monitoring devices arranged at key nodes of a power grid topology, and recording the absolute arrival times of each signal.

[0009] S2, based on the steady-state electrical quantities, performing fault identification through multi-feature fusion criteria, and extracting time-frequency distribution features and wave head polarity features from the transient traveling wave signals after identifying the fault.

[0010] S3, dividing candidate fault sections according to the power grid topology, and performing the following positioning processing for each section: S31, extracting section steady-state feature quantities based on the current phasor of each node after the fault, and performing section fault state identification through a pre-trained pattern recognition model to generate a first positioning result set.

[0011] S32, based on the absolute arrival time, time-frequency distribution features and wave head polarity features, performing multi-terminal ranging and polarity verification to obtain effective ranging results, and cross-verifying the effective ranging results to generate a second positioning result set.

[0012] S4, under the constraint of the power grid topology, performing fusion judgment on the first positioning result set and the second positioning result set to output a final fault section.

[0013] Compared with the prior art, the application has the following advantages: (1) The application divides candidate fault sections according to the power grid topology, and analyzes the multi-source steady-state and transient electrical quantities synchronously collected by key nodes to accurately associate electrical signal abnormalities to specific line sections. This method solves the defect that traditional schemes can only determine whether there is a fault but cannot determine where the fault is, realizes section-level positioning of the fault occurrence position, provides a clear operation target for subsequent rapid fault isolation and power restoration, and significantly improves the pertinence and efficiency of fault handling.

[0014] (2) The application synchronously collects transient traveling wave signals of multiple key nodes, and comprehensively utilizes the absolute arrival time difference, wave head polarity relationship and time-frequency distribution features to perform multi-terminal ranging and cross verification. This method overcomes the limitations of relying only on single-point time sequence signal analysis, utilizes the propagation characteristics and direction information of signals in space, realizes spatial distance estimation and direction judgment of the fault point, and thus endows the fault detection system with essential positioning accuracy, greatly reducing the fault line inspection range.

[0015] (3) The application identifies faults by constructing a criterion of fusion of multi-dimensional features such as instantaneous change rate, harmonic, and zero sequence component, and introduces a pattern recognition model trained based on historical data to intelligently identify the section steady-state characteristics. This method solves the problem of single feature of traditional methods and cannot flexibly cope with complex fault types and operation mode changes. Through multi-level and multi-criterion fusion analysis, the accuracy and robustness of fault identification and section identification are significantly improved, and the false alarm and missed alarm rates are reduced. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the application, the following will briefly introduce the drawings needed to be used in the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0017] Figure 1 A high-voltage transmission line fault detection and identification method steps chart in the application.

[0018] Figure 2 A specific method flow chart of S2 in the application.

[0019] Figure 3 A specific method flow chart of S31 in the application. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the application will be described clearly and completely in the following with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only some embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the application.

[0021] Please refer to Figure 1 As shown in the drawings, the application provides a high-voltage transmission line fault detection and identification method, comprising: S1, acquiring steady-state electrical quantities and transient traveling wave signals synchronously collected by monitoring devices arranged at key nodes of a power grid topology, and recording the absolute arrival time of each signal.

[0022] When a fault occurs in a high-voltage transmission line, the symmetry and stability of its electrical system are destroyed, and this change will directly reflect on the mutation of the amplitude, phase, harmonic, and sequence component of the steady-state electrical quantities such as voltage and current.

[0023] Therefore, by deploying a synchronous phasor measurement unit at key nodes of the power grid such as a transformer substation bus, line ends, and T-junctions, three-phase current waveform data and three-phase voltage waveform data are synchronously collected as steady-state electrical quantities.

[0024] At the same time, the electromagnetic transient process generated at the moment of fault occurrence will propagate in the form of traveling wave to both ends of the line. The arrival time of the wave head of the transient traveling wave signal and the waveform characteristics carry the accurate position information of the fault point. Therefore, from the above three-phase current waveform data and three-phase voltage waveform data, the transient traveling wave signal is captured through signal processing technology.

[0025] All monitoring devices access a unified time reference source and perform time synchronization calibration processing to eliminate the inherent time deviation between devices.

[0026] In a specific embodiment, the specific method of S1 is to synchronously collect three-phase current waveform data and three-phase voltage waveform data at a unified sampling rate (such as 1 MHz) through monitoring devices arranged at key nodes of the power grid topology, and to perform time synchronization calibration processing. The processed three-phase current waveform data and three-phase voltage waveform data are used as steady-state electrical quantities.

[0027] The transient traveling wave signal is captured from the steady-state electrical quantities through wavelet transform technology. The wave head detection algorithm is applied to the transient traveling wave signal of each node, and the corresponding synchronous absolute time reference value at that moment is read, which is recorded as the absolute arrival time of the node.

[0028] Please refer to Figure 2 As shown in the figure, S2, based on the steady-state electrical quantities, performs fault identification through multi-feature fusion criteria, and extracts time-frequency distribution features and wave head polarity features from the transient traveling wave signal after identifying the fault.

[0029] Considering that fault identification based on a single electrical quantity feature may not be able to comprehensively cover different types of line faults such as short circuit and ground due to noise interference or working condition changes, and transient traveling wave analysis needs to be started after confirming the occurrence of the fault to save computing resources.

[0030] During the propagation of the traveling wave, frequency dispersion usually occurs, resulting in a change in wave speed with frequency. Therefore, the time-frequency distribution features that can reflect the energy concentration frequency band and time-frequency distribution characteristics of the traveling wave are extracted as the basis for subsequent wave speed correction.

[0031] At the same time, since the direction information of the fault point relative to each monitoring node is the core of distinguishing the fault section and verifying the physical reasonableness of the ranging result, the wave head polarity feature that can represent the direction relationship of the initial traveling wave wave head of each monitoring node is extracted for preliminary identification of the fault section and physical verification of the ranging result.

[0032] In a specific embodiment, the specific method of S2 is to perform real-time analysis on the steady-state electrical quantities, calculate the instantaneous change rate, harmonic content of each phase current and voltage, and the amplitude and phase of the zero-sequence current and voltage, and combine them into a multi-dimensional feature vector.

[0033] The preset fault feature condition includes a short-circuit fault feature sub-condition and a ground fault feature sub-condition.

[0034] For example, the short-circuit fault feature sub-condition can include that at least two-phase current instantaneous values exceed a set multiple of rated current, corresponding phase voltage drops significantly, and harmonic total content increases sharply; and the ground fault feature sub-condition can include that zero-sequence current amplitude exceeds a normal operation threshold, zero-sequence voltage rises simultaneously, and the phase relationship between the zero-sequence current and the zero-sequence voltage meets a ground fault feature angle.

[0035] In the judgment, when the multi-dimensional feature vector meets the short-circuit fault feature sub-condition or the ground fault feature sub-condition, it is determined that a line fault of a corresponding type occurs.

[0036] After determining that a line fault occurs, time-frequency analysis is performed on the transient traveling wave signal, and an energy distribution entropy value of the transient traveling wave signal in a preset frequency band range is extracted as a time-frequency distribution feature.

[0037] Specifically, the time-frequency distribution feature acquisition method is that a time-frequency analysis technology such as a short-time Fourier transform is used to process the transient traveling wave signal, and energy distribution of the signal in the time-frequency domain is obtained, and the preset frequency band range generally covers a main high-frequency band in which the energy of the traveling wave signal is concentrated, for example, 5 kHz to 200 kHz.

[0038] Subsequently, in the preset frequency band range, the frequency band is uniformly or non-uniformly divided into a plurality of sub-frequency bands, and the proportion of the energy value of each sub-frequency band in the total energy of the preset frequency band is calculated.

[0039] The proportion of the energy of each sub-frequency band is regarded as a probability value, the probability entropy of the energy proportions of all sub-frequency bands is calculated, a scalar value representing the uniformity of the energy distribution is obtained, the value is the energy distribution entropy value, and the energy distribution entropy value is used as the time-frequency distribution feature.

[0040] The polarity signs of the initial current traveling wave wave head at each monitoring node are obtained, and the wave head polarity feature is constructed based on the comparison relationship of the wave head polarity signs between different nodes.

[0041] Further, the specific construction method of the wave head polarity feature is that the polarity sign of the traveling wave wave head is determined by detecting the change direction of the current instantaneous value at the starting point of the traveling wave according to the initial current traveling wave waveform at each monitoring node. For example, the rising edge is marked as “+”, and the falling edge is marked as “-”.

[0042] For any two adjacent monitoring nodes in the power grid topology, the wave head polarity signs of the two nodes are compared: when the wave head polarity signs of the two nodes are the same, it is marked as a same polarity relationship; and when the wave head polarity signs of the two nodes are opposite, it is marked as an opposite polarity relationship.

[0043] According to the connection order of the power grid topology, polar relationship of all adjacent node pairs is arranged in turn to form a polar contrast sequence, and a wave head polarity feature matrix is constructed based on the distribution mode of the same polarity relationship and the opposite polarity relationship in the polar contrast sequence.

[0044] S3, according to the power grid topology structure, the candidate fault section is divided, and positioning processing is performed on each section.

[0045] Considering the complexity of high-voltage transmission line topology, direct global positioning has low efficiency and poor accuracy, and single-dimensional positioning results have insufficient reliability.

[0046] Therefore, the candidate fault section is divided by the power grid topology information, and then steady-state feature positioning and transient traveling wave positioning are performed on each section respectively to generate a first positioning result set and a second positioning result set.

[0047] Among them, the division of the candidate fault section takes the key nodes where the monitoring devices are arranged as the section boundary nodes according to the power grid topology structure, and divides the line part between adjacent key nodes as a unit to divide into multiple candidate fault sections.

[0048] Please refer to Figure 3 As shown in the figure, S31, based on the current phasor of each node after the fault, the steady-state feature quantity of the section is extracted, and the section fault state is identified by the pre-trained pattern recognition model to generate the first positioning result set.

[0049] In a specific embodiment, the specific method of S31 is: after determining that a line fault occurs, considering that the current phasor relationship inside and outside the fault section follows Kirchhoff's current law, there is an essential difference, so the current phasor corresponding to the boundary nodes at both ends of each candidate fault section is obtained, the vector sum of the current phasors at both ends of the section is calculated as the differential current phasor, and the vector difference is calculated as the braking current phasor.

[0050] Among them, the differential current phasor reflects the total fault current flowing into the candidate fault section, and the braking current phasor reflects the current imbalance degree at both ends of the section, which is larger when the section is externally faulty.

[0051] The differential current phasor and the braking current phasor are respectively subjected to polar coordinate transformation, and the amplitude ratio and the phase difference are extracted as the steady-state feature quantity of the section, which quantitatively describes the electrical response mode of the section under fault state.

[0052] The steady-state feature quantity of the section is input into the pattern recognition model trained based on historical fault data for identification.

[0053] The mode recognition model divides a feature space into an internal fault clustering area and an external fault clustering area. Specifically, historical line fault data is collected to extract historical steady-state feature quantities, and according to actual fault records and on-site verification information, each historical sample is marked to indicate whether a corresponding candidate fault section is a real fault section, to form a sample label.

[0054] For example, if a historical fault occurs in section A according to line inspection, all steady-state feature quantity samples corresponding to section A extracted from this fault recording data are marked as internal faults (positive samples), and steady-state feature quantity samples corresponding to other non-fault sections (such as sections B and C) at the same time are marked as external faults (negative samples).

[0055] The samples marked as real fault sections are used to construct an internal fault sample set, and the samples marked as non-fault sections are used to construct an external fault sample set.

[0056] A supervised machine learning algorithm, such as a support vector machine, a random forest, or a gradient boosting decision tree, is used to train the internal fault sample set and the external fault sample set, with historical steady-state feature quantities as input features and sample labels as training targets.

[0057] After training, the model forms a classification boundary in the feature space. The region formed by the coordinate points judged to be faults located inside the section is defined as the internal fault clustering area, and vice versa.

[0058] If the projection point of the section steady-state feature quantity in the feature space falls into the internal fault clustering area, it indicates that the electrical response mode of the section is highly similar to the historical internal fault mode, and it is determined that the fault is located inside the candidate section; otherwise, it indicates that the mode is closer to the external fault or normal operation, and it is determined that the fault is not in the candidate section.

[0059] Finally, the discrimination results of all candidate fault sections are summarized, and the sections determined to have faults inside are included in the first positioning result set.

[0060] S32, based on the absolute arrival time, time-frequency distribution characteristics and wave head polarity characteristics, multi-terminal ranging and polarity verification are performed to obtain effective ranging results, and cross verification is performed on the effective ranging results to generate a second positioning result set.

[0061] In one specific embodiment, the method for obtaining the effective ranging result is as follows: based on the absolute arrival time of the transient traveling wave signal recorded by each monitoring node, the propagation time difference between any node pair is calculated, and the distance estimation value of the fault occurrence point to each node is calculated in combination with the traveling wave propagation speed.

[0062] Considering the frequency dispersion effect of the traveling wave in the transmission line, i.e. the propagation speed of different frequency components is different, the fixed wave speed will cause ranging error, therefore, the wave speed correction is needed by using the time-frequency distribution characteristics.

[0063] The specific method is: by analyzing the peak value of energy distribution entropy, the dominant frequency band with the most concentrated energy in the transient fault traveling wave signal is identified, the corresponding relationship between frequency and wave speed, i.e. the preset dispersion curve, is queried and the optimal propagation speed corresponding to the dominant frequency band is determined, the preliminary distance estimation value is recalculated and corrected by using the optimal propagation speed, and the corrected distance estimation value is obtained.

[0064] By using the corrected distance estimation value, geometric positioning can be performed by using the double-end or three-end ranging formula, and the possible positional relationship of the fault point relative to each monitoring node can be inferred, for example, the fault point is located on the line segment between node A and node B.

[0065] Based on the traveling wave propagation theory, when the fault point is located between two monitoring nodes, the initial current traveling wave generated by the fault will propagate to the two nodes in opposite directions, resulting in opposite polarities of the initial wave head detected by the two nodes. Therefore, a polarity criterion is established based on the theory to verify the physical rationality of the ranging result.

[0066] The specific verification method is: comparing the fault position relationship inferred by geometric positioning with the extracted wave head polarity feature. For all node pairs whose fault points are inferred to be located therebetween, check whether the polarity relationship recorded in the wave head polarity feature of the node pair is an opposite polarity relationship.

[0067] If the inferred position and polarity relationship are consistent, i.e. meet the opposite polarity, the ranging result is physically self-consistent and should be retained; if they are inconsistent, for example, the fault is inferred to be between two points, but the polarities of the two points are the same, it indicates that the preliminary ranging result may be unreliable due to wave head recognition error, reflected wave interference, etc. and should be excluded. The ranging result retained through this verification is the effective ranging result.

[0068] Further, since multiple effective ranging results are obtained from multiple pairs of different nodes, they are independent of each other and may have slight deviations, and need to be fused to obtain the optimal positioning estimation value.

[0069] Specifically, the cross-verification method is: for all effective ranging results, i.e. the sum or difference of the distances from the fault point to two nodes, a topological correlation graph is constructed with the monitoring nodes as the vertices and the distance estimation values as the edge weights, and a graph theory optimization algorithm such as an optimization algorithm based on least square estimation is applied to minimize the overall variance between all edge weights (observations) in the graph and the theoretical distance value calculated from a hypothetical single fault point as the target, and iterative solution is performed.

[0070] This process can screen out the optimal combination of node pairs that can make all observations self-consistent, and simultaneously estimate the distance from the fault point to each node.

[0071] In the optimization solving process, the confidence of the node distance estimate is calculated. The specific method of calculating the confidence is: according to the optimal position of the fault point obtained by solving the optimization model, the theoretical value of each effective ranging result is calculated by back calculation. The residual of each effective ranging result is defined as the absolute value of the difference between the measured value and the theoretical value.

[0072] The confidence is set to be inversely proportional to the square of the residual, that is, the ranging result with smaller residual has higher confidence. The effective ranging results are weighted and fused according to the confidence, and the final fault position estimate value in each candidate fault section is generated.

[0073] Specifically, the confidence of each effective ranging result calculated is normalized, and the normalized confidence value is directly used as the weight coefficient corresponding to each result in the weighted fusion. Then, multiple effective ranging results located in the same candidate section are weighted and averaged according to their weights, and the result is the final fault position estimate value of the section.

[0074] All candidate sections are traversed, and those whose final fault position estimate value falls within their physical length range and are generated by data fusion verified by polarity are determined as potential fault sections. All these potential fault sections and their internal estimate positions are summarized to form a second positioning result set. This result set provides transient traveling wave-based positioning information that has been physically verified and data fused.

[0075] It should be noted that if the signal-to-noise ratio of the transient traveling wave signal is lower than a preset threshold (such as 10 dB), S32 is skipped, and the first positioning result set is directly used as the final output.

[0076] S4, under the constraint of the power grid topology, the first positioning result set and the second positioning result set are fused to output the final fault section.

[0077] In a specific embodiment, the specific method of S4 is: under the constraint of the power grid topology, the first positioning result set and the second positioning result set are fused to output the final fault section.

[0078] (ii) If there is no common section, the candidate fault sections with the highest confidence from the first and second positioning result sets are denoted as section A and section B, respectively, and the distance between the two sections is calculated based on the grid topology, such as the node connection graph, for example, in terms of the number of intermediate interval line sections or physical distance.

[0079] If the distance is less than a preset topological proximity threshold, it indicates that although the specific sections located by the two methods are not exactly the same, the indicated possible fault locations are adjacent in the electrical topology, and the fault is likely to have actually occurred in the adjacent area or near the connection point of the two sections.

[0080] wherein the topological proximity threshold is set according to actual conditions, and is usually set to be adjacent or not more than two sections apart.

[0081] In this case, section A and section B and all sections between them are merged into a continuous extended fault section, and output as the final fault section.

[0082] Otherwise, it indicates that the most likely fault locations of the two methods are far apart in the electrical topology, and the automatic fusion is determined to fail, an indication of positioning failure and request for manual intervention is output, and the final decision is left to the operator.

[0083] It should be noted that if both the first and second positioning result sets are empty sets, the region with the highest similarity based on the similarity of the feature vector of the fault identified in S2 and the historical fault mode is output as the final fault section; if there is no region that meets the conditions, an indication of positioning failure and request for manual intervention is output.

[0084] In summary, the present application first acquires steady-state electrical quantities and transient traveling wave signals through synchronous monitoring; then, the multi-feature fusion criterion of steady-state electrical quantities is used to reliably identify faults and trigger subsequent analysis, while time-frequency features and polarity features for precise positioning are extracted from transient traveling waves.

[0085] Next, candidate sections are divided according to the grid topology, and the section state discrimination based on the steady-state differential principle and pattern recognition, as well as the transient positioning based on multi-terminal traveling wave distance measurement, polarity verification and data fusion, are performed in parallel to generate preliminary positioning result sets.

[0086] Finally, under the physical constraints of the grid topology structure, the two result sets are intelligently fused to determine the unique and reliable final fault section, or request for manual intervention when the conclusions conflict.

[0087] This method realizes the full-process automation and high reliability from fault detection, feature extraction, parallel positioning to intelligent decision-making by fusing multi-dimensional information such as steady-state and transient, electrical quantities and topological relationships.

[0088] The above-described embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented in software, the above-described embodiments can be implemented in the form of a computer program product.

[0089] Those skilled in the art can realize that the algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be realized in electronic hardware or in a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0090] In addition, each functional module in each embodiment of the present application can be integrated in one processing module, or each module can exist physically independently, or two or more modules can be integrated in one module.

[0091] The above is merely specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0092] Finally, the above is merely preferred embodiments of the present application, and is not intended to limit the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for detecting and identifying faults in high-voltage transmission lines, characterized in that, include: S1. Acquire steady-state electrical quantities and transient traveling wave signals synchronously collected by monitoring devices deployed at key nodes of the power grid topology, and record the absolute arrival time of each signal. S2. Based on steady-state electrical quantities, fault identification is performed through multi-feature fusion criteria, and after fault identification, time-frequency distribution features and wavefront polarity features are extracted from transient traveling wave signals. S3. Based on the power grid topology, divide the candidate fault sections and perform the following location processing on each section: S31. Extract steady-state feature quantities of the section based on the current phasors of each node after the fault, and identify the fault state of the section through a pre-trained pattern recognition model to generate the first localization result set. S32. Based on the absolute time of arrival, time-frequency distribution characteristics, and wavefront polarity characteristics, perform multi-terminal ranging and polarity verification to obtain effective ranging results, and perform cross-validation to generate a second positioning result set; S4. Under the constraints of the power grid topology, perform a fusion judgment on the first location result set and the second location result set to output the final fault section.

2. The high-voltage transmission line fault detection and identification method as described in claim 1, characterized in that, The specific method of S1 is as follows: Three-phase current waveform data and three-phase voltage waveform data are collected synchronously by monitoring devices deployed at key nodes of the power grid topology, and time synchronization calibration is performed. The processed three-phase current waveform data and three-phase voltage waveform data are used as steady-state electrical quantities; Transient traveling wave signals can be captured from steady-state electrical quantities using wavelet transform technology. The time it takes for the transient traveling wave signal to arrive at each monitoring node is recorded as the absolute arrival time.

3. The high-voltage transmission line fault detection and identification method as described in claim 1, characterized in that, The specific method of S2 is as follows: Real-time analysis of steady-state electrical quantities is performed to calculate the instantaneous rate of change of current and voltage in each phase, harmonic content, and the amplitude and phase of zero-sequence current and zero-sequence voltage. The instantaneous rate of change, harmonic content, zero-sequence current amplitude, zero-sequence current phase, zero-sequence voltage amplitude, and zero-sequence voltage phase are combined into a multi-dimensional feature vector; Based on preset fault characteristic conditions, a line fault is determined to have occurred when the multidimensional feature vector satisfies the short-circuit fault characteristic sub-condition or the ground fault characteristic sub-condition. After determining that a line fault has occurred, time-frequency analysis is performed on the transient traveling wave signal to extract its energy distribution entropy value within a preset frequency band as time-frequency distribution characteristics. The polarity sign of the initial current traveling wavefront at each monitoring node is obtained, and the wavefront polarity feature is constructed based on the comparison relationship of wavefront polarity signs between different nodes.

4. The high-voltage transmission line fault detection and identification method as described in claim 3, characterized in that, The specific method for constructing the wavehead polarity feature is as follows: Based on the initial current traveling wave waveform at each monitoring node, the polarity sign of the traveling wave front is determined by detecting the direction of change of the instantaneous current value at the starting point of the traveling wave. For any two adjacent monitoring nodes in the power grid topology, compare their wavefront polarity signs: if the wavefront polarity signs of the two nodes are the same, they are marked as having the same polarity; if the wavefront polarity signs of the two nodes are opposite, they are marked as having opposite polarity. According to the power grid topology connection order, the polarity relationship of all adjacent node pairs is arranged in sequence to form a polarity comparison sequence; Based on the distribution patterns of same-polarity and opposite-polarity relationships in polarity contrast sequences, a matrix reflecting wavefront polarity characteristics is constructed.

5. The high-voltage transmission line fault detection and identification method as described in claim 1, characterized in that, The specific method for dividing candidate fault sections according to the power grid topology is as follows: Based on the power grid topology, the key nodes where monitoring devices are deployed are taken as section boundary nodes, and the line sections between adjacent key nodes are taken as units to divide the area into multiple candidate fault sections.

6. The high-voltage transmission line fault detection and identification method as described in claim 1, characterized in that, The specific method of S31 is as follows: After determining that a line fault has occurred, the synchronously acquired current phasors corresponding to the boundary nodes at both ends of each candidate fault section are obtained. The vector sum of the current phasors at both ends of the section is calculated as the differential current phasor, and the vector difference is calculated as the braking current phasor. Polar coordinate transformations were performed on the differential current phasor and the braking current phasor respectively, and their amplitude ratio and phase difference were extracted as steady-state characteristic quantities of the section. The steady-state features of the section are input into a pattern recognition model trained based on historical fault data for discrimination. The pattern recognition model divides the feature space into internal fault clustering regions and external fault clustering regions. If the projection point of the steady-state characteristic quantity of a segment in the feature space falls into the internal fault clustering region, then the fault is determined to be located inside the candidate segment; otherwise, the fault is determined not to be within the candidate segment. The judgment results of all candidate fault segments are summarized to generate the first localization result set.

7. The high-voltage transmission line fault detection and identification method as described in claim 6, characterized in that, The pattern recognition model divides the feature space into internal fault clustering regions and external fault clustering regions, specifically: Historical line fault data is collected to extract historical steady-state features. Based on actual fault records and on-site verification information, each historical sample is labeled to determine whether its corresponding candidate fault section is a real fault section, thus forming a sample label. An internal fault sample set is constructed using samples labeled as actual fault segments, and an external fault sample set is constructed using samples labeled as non-fault segments. A supervised machine learning algorithm is used, with historical steady-state features as input features and sample labels as training targets, to train on internal and external fault sample sets. After training is completed, the region consisting of coordinate points that are determined to be faults located within the segment is defined as an internal fault cluster region; otherwise, it is defined as an external fault cluster region.

8. The high-voltage transmission line fault detection and identification method as described in claim 1, characterized in that, The method for obtaining the effective ranging result is as follows: Based on the absolute arrival time of the transient traveling wave signal recorded by each monitoring node, the propagation time difference between any pair of nodes is calculated, and the distance estimate from the fault location to each node is calculated in combination with the traveling wave propagation speed. By utilizing the time-frequency distribution characteristics, the dominant frequency band of transient traveling wave signal energy is identified by analyzing the peak value of energy distribution entropy. The optimal propagation speed corresponding to the dominant frequency band is determined according to the preset dispersion curve, and the distance estimate is corrected using the optimal propagation speed. Using the corrected distance estimates, the positional relationship between the fault location and each monitoring node is inferred through geometric positioning, and a polarity criterion is established based on traveling wave propagation theory, as follows: If the fault location is inferred to be on the line segment between two monitoring nodes based on the corrected distance estimate, then theoretically, the node should have an opposite polarity relationship with the polarity of the traveling wave front of the detected initial current. The inferred fault location relationship is compared with the extracted wavefront polarity features. For all node pairs inferred to be where the fault point is located, their wavefront polarity features are checked to see if they conform to the opposite polarity relationship. If the result does not meet the requirements, the corresponding distance measurement result will be discarded, and the distance measurement result that has passed the verification will be retained as the valid distance measurement result.

9. The high-voltage transmission line fault detection and identification method as described in claim 1, characterized in that, The specific method for cross-validation is as follows: For all valid distance measurement results, a topological association graph is constructed with monitoring nodes as vertices and distance estimates as edge weights. Graph theory algorithms are applied to minimize the overall variance of all edge weights in the graph as the optimization objective to select the optimal combination of node pairs and calculate the confidence level of the distance estimates of each node. The effective ranging results are weighted and fused according to the confidence level to generate the final fault location estimate in each candidate fault segment; The final fault location estimates of each candidate segment are summarized, and the segments whose estimated values ​​are within their physical range and pass the above polarity verification are identified as potential fault segments, forming a second location result set.

10. The high-voltage transmission line fault detection and identification method as described in claim 1, characterized in that, The specific method of S4 is as follows: Based on the constraints of the power grid topology, the following fusion judgment logic is executed using the first and second positioning result sets: (i) If there is a candidate fault segment that is included in both the first location result set and the second location result set, then the segment is directly identified and output as a fault segment; (ii) If there is no common section, the candidate fault section with the highest confidence is selected from the first location result set and the second location result set respectively, and the distance between the two sections is calculated based on the power grid topology. If the distance is less than the preset topological proximity threshold, the two segments are merged into a single continuous fault segment and output as the final fault segment. Otherwise, output an instruction indicating that the location has failed and requesting manual intervention.

Images