Power distribution network overhead line fault detection method and system based on double-end traveling waves

Abstract

The invention relates to the technical field of overhead line fault detection, and particularly discloses a power distribution network overhead line fault detection method and system based on double-end traveling waves, and the method comprises the steps: obtaining the traveling wave signal characteristics of a current traveling wave signal and a voltage traveling wave signal of a line end point through a sensor group; preprocessing and time sequence segmentation are carried out based on the traveling wave signal features to construct a traveling wave sample set, traveling wave sample time sequence features are captured according to the traveling wave sample set, and a one-dimensional traveling wave feature row vector is generated to construct a traveling wave feature sample set; respectively extracting a first correlation feature and a second correlation feature according to the traveling wave feature sample set, and determining whether the traveling wave generated by the switching operation affects the traveling wave of the fault line based on fusion discrimination of the two correlation features; according to the method, high-precision, rapid and robust fault detection and positioning can be realized in a complex power distribution network environment, switch operation interference and real faults can be distinguished, and solid data support is provided for safe operation and intelligent management of the power distribution network.

Description

Technical Field

[0001] This invention relates to the field of overhead line fault detection technology, and more specifically, to a method and system for detecting faults in overhead lines of distribution networks based on double-ended traveling waves. Background Technology

[0002] In the operation of overhead power distribution lines, fault location and detection are crucial for ensuring power supply reliability and rapid restoration. Traveling wave-based fault detection methods have gradually become a research and engineering application hotspot due to their fast response speed and high location accuracy. Among them, the dual-end traveling wave ranging technology collects the arrival times of traveling waves at both ends of the line when a fault occurs, and then uses the time difference to calculate the location of the fault point.

[0003] In actual power distribution networks, the opening and closing of equipment such as line switches, disconnectors, and circuit breakers also generate electromagnetic traveling wave signals on the conductors. These interference traveling waves caused by switch operations share certain similarities in waveform characteristics and propagation patterns with the initial fault traveling wave. For example, they may be confused in terms of high-frequency components, arrival timing, and wavefront abrupt changes, thus interfering with the identification and extraction of fault traveling waves. Relying solely on traditional traveling wave feature extraction methods often makes it difficult to distinguish between fault traveling waves and operational interference traveling waves, easily leading to misjudgments or decreased location accuracy.

[0004] Therefore, it is necessary to provide a method and system for fault detection of overhead lines in distribution networks based on double-ended traveling waves to solve the above-mentioned technical problems. In order to solve the above problems, a technical solution is provided. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, this invention provides a method and system for detecting faults in overhead lines of distribution networks based on double-ended traveling waves. This addresses the problem that existing technologies, which rely solely on traditional traveling wave feature extraction methods, often struggle to distinguish between fault traveling waves and operational interference traveling waves, leading to misjudgments or decreased positioning accuracy.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for fault detection of overhead lines in a distribution network based on double-ended traveling waves includes the following steps: The current traveling wave signal and voltage traveling wave signal at the line endpoints are acquired by a sensor array, and the traveling wave signal characteristics of the current traveling wave signal and voltage traveling wave signal are extracted respectively. Based on the characteristics of the traveling wave signal, a traveling wave sample set is constructed by preprocessing and time-series segmentation. The time-series characteristics of the traveling wave samples are captured based on the traveling wave sample set, and a one-dimensional traveling wave feature row vector is generated to construct the traveling wave feature sample set. The first and second associated features are extracted from the traveling wave feature sample set, and the traveling wave generated by the switch operation is determined to affect the traveling wave of the faulty line based on the fusion and discrimination of the two types of associated features.

[0007] As a further aspect of the present invention, the current traveling wave signal and the voltage traveling wave signal at the line endpoint are acquired by a sensor group, and the traveling wave signal features of the current traveling wave signal and the voltage traveling wave signal are extracted respectively; the traveling wave signal features include a first traveling wave signal feature, a second traveling wave signal feature and a third traveling wave signal feature.

[0008] As a further aspect of the present invention, the traveling wave signal characteristics of the current traveling wave signal and the voltage traveling wave signal when switching operation is performed under fault-free conditions are obtained as the first traveling wave signal characteristics; the traveling wave signal characteristics of the current traveling wave signal and the voltage traveling wave signal when switching operation is performed under fault conditions are obtained as the second traveling wave signal characteristics; and the traveling wave signal characteristics of the current traveling wave signal and the voltage traveling wave signal when switching operation is performed under fault conditions are obtained as the third traveling wave signal characteristics.

[0009] As a further aspect of the present invention, a traveling wave sample set is constructed by preprocessing and time-series segmentation based on the traveling wave signal features. The time-series features of the traveling wave samples are captured from the traveling wave sample set, and a one-dimensional traveling wave feature row vector is generated to construct a traveling wave feature sample set. The specific steps are as follows: Traveling wave signal features are acquired, and the traveling wave signal features are preprocessed and time-series segmented to obtain a single-window traveling wave feature set. ,in, The characteristics of the traveling wave signal at time t are... To construct a traveling wave sample set based on the single-window traveling wave characteristics, the sample set is defined with a single window length. , For the i-th single-window traveling wave feature set, The number of features in a single-window traveling wave feature set; Based on traveling wave sample set A time series feature analysis model for traveling wave samples is constructed to capture the time series features of traveling wave samples corresponding to the single-window traveling wave feature set. One-dimensional traveling wave feature row vectors are generated based on the temporal characteristics of the traveling wave samples to form a traveling wave feature sample set.

[0010] As a further aspect of the present invention, the traveling wave feature sample set includes a traveling wave feature sample set composed of the traveling wave sample time-series features obtained based on the first traveling wave signal features, and is labeled as a switching operation sample reference set. ,in, Let i be the set of single-window traveling wave features in the switch operation sample reference set. This is the nth single-window traveling wave feature set in the switch operation sample reference set. The number of traveling wave feature sets in a single window; the traveling wave feature sample set composed of the traveling wave sample time-series features obtained based on the second traveling wave signal features is labeled as the fault sample reference set. ,in, Let i be the set of single-window traveling wave features in the fault sample reference set. The nth single-window traveling wave feature set in the fault sample reference set; the traveling wave feature sample set composed of the traveling wave sample time series features obtained based on the third traveling wave signal features is labeled as the mixed sample reference set. ,in, Let i be the set of traveling wave features in the i-th single-window sample reference set. Let n be the nth single-window traveling wave feature set in the mixed sample reference set.

[0011] As a further aspect of the present invention, by constructing a switch operation sample reference set, a fault sample reference set, and a mixed sample reference set, first correlation features and second correlation features are extracted respectively. Based on the fusion and discrimination of the two types of correlation features, it is determined whether the traveling wave generated by the switch operation affects the traveling wave of the faulty line. The specific steps are as follows: The switching operation sample reference set, the fault sample reference set, and the mixed sample reference set are extracted from the traveling wave feature sample set respectively; A first traveling wave signal correlation analysis model is constructed by using a switch operation sample reference set and a mixed sample reference set, and the first correlation feature is obtained by fusion analysis. A second traveling wave signal correlation analysis model is constructed by using a fault sample reference set and a mixed sample reference set, and the second correlation feature is obtained by fusion analysis. Based on the first and second correlation features, it is determined whether the traveling wave generated by the switch operation affects the traveling wave of the faulty line.

[0012] As a further aspect of the present invention, a first traveling wave signal correlation analysis model is constructed by using a switching operation sample reference set and a mixed sample reference set, and the first correlation feature is obtained through fusion analysis. The specific steps are as follows: Obtain a reference set of switch operation samples ,in, Let i be the set of single-window traveling wave features in the switch operation sample reference set. For the nth single-window traveling wave feature set in the switch operation sample reference set, extract the switch operation sample reference set. With mixed sample reference set The time series characteristics of all traveling wave samples in the data, among which, Let i be the set of traveling wave features in the i-th single-window sample reference set. For the nth single-window traveling wave feature set in the mixed sample reference set, calculate the minimum absolute difference based on the time series characteristics of the traveling wave samples in the two sample reference sets, determine the single-window traveling wave feature set corresponding to the minimum absolute difference, construct the first traveling wave signal correlation analysis model, and obtain the first correlation feature through fusion analysis.

[0013] As a further aspect of the present invention, the formula for the first traveling wave signal correlation analysis model is as follows: ; In the formula: As the first associated feature, The time window length, Let be the traveling wave signal characteristics at time t in the j-th single-window traveling wave feature set of the switch operation sample reference set. Let be the traveling wave signal feature at time t in the j-th single-window traveling wave feature set of the mixed sample reference set.

[0014] As a further aspect of the present invention, the traveling wave generated by the switch operation is determined based on the first correlation feature and the second correlation feature to determine whether it affects the traveling wave of the faulty line. The specific steps are as follows: Calculate the mixed sample reference set Standard deviation ,in, Let i be the set of traveling wave features in the i-th single-window sample reference set. For the nth single-window traveling wave feature set in the mixed sample reference set, based on standard deviation Set the distinction threshold , For adjustment coefficients; A second traveling wave signal correlation analysis model is constructed using a fault sample reference set and a mixed sample reference set, and the second correlation features are obtained through fusion analysis. ; The first associated feature With the second associated feature The difference Compared with the discrimination threshold, if the first associated feature With the second associated feature The difference If the value is greater than or equal to the distinguishing threshold, the traveling wave generated by the switching operation will affect the traveling wave of the faulty line; if the first associated feature... With the second associated feature The difference If the value is less than the distinguishing threshold, the traveling wave generated by the switching operation will not affect the traveling wave of the faulty line.

[0015] A fault detection system for overhead lines in a distribution network based on a double-ended traveling wave, the system comprising: a traveling wave signal feature extraction module, a traveling wave feature sample aggregation module, and a traveling wave feature influence discrimination module; The traveling wave signal feature extraction module is used to acquire the current traveling wave signal and voltage traveling wave signal at the line endpoints through the sensor group, and extract the traveling wave signal features of the current traveling wave signal and voltage traveling wave signal respectively. The traveling wave feature sample aggregation module performs preprocessing and time-series segmentation based on the traveling wave signal features to construct a traveling wave sample set. It captures the time-series features of the traveling wave samples based on the traveling wave sample set and generates a one-dimensional traveling wave feature row vector to construct the traveling wave feature sample set. The traveling wave feature impact discrimination module extracts the first and second associated features from the traveling wave feature sample set, and determines whether the traveling wave generated by the switch operation affects the traveling wave of the fault line based on the fusion discrimination of the two types of associated features.

[0016] The technical effects and advantages of this invention, a method and system for detecting faults in overhead lines of a distribution network based on double-ended traveling waves, are as follows: This invention acquires current traveling wave signals and voltage traveling wave signals at the line endpoints using a sensor array, and extracts the traveling wave signal features of the current and voltage traveling wave signals respectively; based on the traveling wave signal features, preprocessing and time-series segmentation are performed to construct a traveling wave sample set; the time-series features of the traveling wave samples are captured according to the traveling wave sample set, and a one-dimensional traveling wave feature row vector is generated to construct a traveling wave feature sample set; based on the traveling wave feature sample set, a first correlation feature and a second correlation feature are extracted respectively, and based on the fusion and discrimination of the two types of correlation features, it is determined whether the traveling wave generated by the switching operation affects the traveling wave of the faulty line.

[0017] This invention acquires current and voltage traveling wave signals by deploying high-precision sensor groups at both ends of the line and extracts their traveling wave features. This allows the method to capture the high-frequency transient characteristics generated by the line at the moment of a fault, enabling rapid response to fault events. Compared to traditional methods based on steady-state quantities or single-end measurements, this approach can acquire fault information within milliseconds, improving the timeliness of fault detection. Preprocessing and time-series segmentation based on traveling wave signal features construct a traveling wave sample set and capture its temporal features. A one-dimensional traveling wave feature row vector is then generated to construct the traveling wave feature sample set. This allows the method to systematically characterize the dynamic changes of the traveling wave over time, effectively reducing background noise interference and enhancing the accuracy and robustness of fault feature extraction. By extracting the first and second correlation features of switch operation samples and fault samples respectively, and fusing these two types of correlation features for discrimination, the method can identify the difference between the interference traveling wave generated by the switch operation and the actual fault traveling wave, thereby eliminating interference components. This not only improves the accuracy of fault initial wave identification but also avoids false positives and false negatives, improving the reliability of fault location. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a fault detection method for overhead lines in a distribution network based on a double-ended traveling wave, provided in an embodiment of the present invention; Figure 2 This is a system block diagram of a fault detection system for overhead power distribution lines based on double-ended traveling waves, provided as an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described technical solutions are only a part of this invention, and not all of it. All other technical solutions obtained by those skilled in the art based on the technical solutions of this invention without inventive effort are within the scope of protection of this invention.

[0020] like Figure 1 The diagram shown is a flowchart of a fault detection method for overhead lines in a distribution network based on a double-ended traveling wave, provided by an embodiment of the present invention. Figure 1 The execution entity of the method shown can be a software and / or hardware device. The execution entity of this application can include, but is not limited to, at least one of the following: user equipment, network equipment, etc. User equipment can include, but is not limited to, computers, smartphones, personal digital assistants (PDAs), and the aforementioned electronic devices. Network equipment can include, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers. Cloud computing is a type of distributed computing, consisting of a super virtual computer composed of a group of loosely coupled computers. This embodiment does not limit this. Steps S1 to S3 are detailed as follows: S1, acquire the current traveling wave signal and voltage traveling wave signal at the line endpoint through the sensor group, and extract the traveling wave signal characteristics of the current traveling wave signal and voltage traveling wave signal respectively; This invention employs high-precision sensor arrays, including high-speed current transformers and voltage transformers, installed at both ends of overhead power lines in a distribution network to acquire current and voltage traveling wave signals at the line endpoints. The sampling rate of the sensor arrays needs to be sufficiently high, typically in the range of hundreds of kHz to 1 MHz, to ensure the rapid changes in the traveling wave signals are captured. The acquired raw current and voltage signals are first bandpass filtered by a pre-filter to remove power frequency and high-frequency noise, and then detrended to eliminate DC bias and baseline drift. Subsequently, the processed current and voltage signals are divided into fixed-length time windows, forming single-window traveling wave signal sets, where the data within each window represents the traveling wave characteristics of that time period. For each window's current and voltage traveling wave signals, characteristic parameters are further extracted, such as peak amplitude, rise slope, local maxima and minima, waveform energy, and spectral characteristics, to construct current and voltage traveling wave feature sets, providing fundamental data for subsequent traveling wave sample time-series analysis and fault criterion construction. In this process, to ensure comparability and time alignment between different types of traveling wave characteristics, the start time of all windows must be unified to ensure that current and voltage traveling wave characteristics are analyzed and processed on the same time scale.

[0021] S2, based on the characteristics of the traveling wave signal, preprocess and time-series segmentation are performed to construct a traveling wave sample set. The time-series characteristics of the traveling wave samples are captured according to the traveling wave sample set, and a one-dimensional traveling wave feature row vector is generated to construct the traveling wave feature sample set. This invention first preprocesses the acquired current and voltage traveling wave signals, including removing DC bias and bandpass filtering to suppress noise and abnormal fluctuations. Simultaneously, the signals are normalized to ensure comparability of data from different sampling points or line segments. Then, the preprocessed traveling wave signals are divided into several single windows of fixed time length, forming single-window traveling wave feature sets. For each single-window traveling wave feature set, the dynamic characteristics of the traveling wave over time are captured, generating corresponding traveling wave sample time-series features. Next, the traveling wave sample time-series features of each single window are arranged along the time axis to construct a one-dimensional traveling wave feature row vector, forming a traveling wave feature sample set. This method systematically extracts the temporal characteristic changes of the traveling wave signal, providing a reliable data foundation for subsequent traveling wave correlation analysis, fault criterion construction, and removal of switching operation interference from mixed samples.

[0022] S3. Extract the first and second associated features from the traveling wave feature sample set, and determine whether the traveling wave generated by the switch operation affects the traveling wave of the faulty line based on the fusion and discrimination of the two types of associated features.

[0023] In distribution network fault monitoring, to determine whether the traveling wave generated by switch operations affects the traveling wave of the faulty line, a traveling wave feature sample set is obtained from the monitoring data. Based on the data source, this set is divided into a switch operation sample reference set, a fault sample reference set, and a mixed sample reference set. The switch operation sample reference set records the traveling wave signal characteristics generated during switch operations, the fault sample reference set records the traveling wave signal characteristics of the line under fault conditions, and the mixed sample reference set includes traveling wave signals that may be affected by switch operations during the fault occurrence.

[0024] Subsequently, a first traveling wave signal correlation analysis model was constructed based on the switch operation sample reference set and the mixed sample reference set, respectively. By calculating the absolute difference between the two types of samples in a single window time sequence and selecting the traveling wave feature set corresponding to the smallest absolute difference, the first correlation feature was obtained through fusion analysis. This feature was used to quantify the performance of the switch operation traveling wave in the mixed sample. Similarly, a second traveling wave signal correlation analysis model was constructed based on the fault sample reference set and the mixed sample reference set. Using the same calculation steps as the first correlation feature, the second correlation feature was obtained through fusion analysis. This feature was used to reflect the performance of the fault line traveling wave in the mixed sample.

[0025] After obtaining the first and second correlation features, the standard deviation of the mixed sample reference set is further calculated, and a distinction threshold is set based on the standard deviation. An adjustment coefficient is introduced to adapt to actual signal fluctuations. Then, the difference between the first and second correlation features is compared with the distinction threshold: if the difference is greater than or equal to the distinction threshold, it indicates that the traveling wave generated by the switching operation has a significant impact on the traveling wave of the faulty line; if the difference is less than the distinction threshold, it indicates that the traveling wave from the switching operation has a relatively small impact on the traveling wave of the faulty line, and the traveling wave in the mixed sample is mainly generated by the faulty line itself. Through the above implementation method, it is possible to accurately distinguish between the traveling wave from the switching operation and the fault traveling wave in complex distribution network environments, ensuring the accurate extraction of fault traveling wave features, providing a reliable basis for subsequent fault location, isolation, and analysis, while effectively reducing the risk of misjudgment caused by switching operation interference.

[0026] Preferably, the current traveling wave signal and the voltage traveling wave signal at the line endpoint are acquired by the sensor group, and the traveling wave signal features of the current traveling wave signal and the voltage traveling wave signal are extracted respectively; the traveling wave signal features include a first traveling wave signal feature, a second traveling wave signal feature and a third traveling wave signal feature; The traveling wave signal characteristics of the current traveling wave signal and the voltage traveling wave signal when switching operation is performed under fault-free conditions are obtained as the first traveling wave signal characteristics; the traveling wave signal characteristics of the current traveling wave signal and the voltage traveling wave signal when switching operation is performed under fault conditions are obtained as the second traveling wave signal characteristics; the traveling wave signal characteristics of the current traveling wave signal and the voltage traveling wave signal when switching operation is performed under fault conditions are obtained as the third traveling wave signal characteristics.

[0027] It should be noted that the first, second, and third traveling wave signal features need to be time aligned, meaning that the three traveling wave signal features are acquired at the same start time.

[0028] In actual operation of overhead power distribution lines, the embodiments of the present invention can capture transient traveling wave signals of the line in real time by using high-speed current and voltage sensor groups installed at the beginning and end of the feeder. Taking a 10kV overhead line as an example, when maintenance personnel close the load switch at the end of the line under fault-free conditions, the current and voltage traveling wave signals recorded by the sensor only exhibit transient fluctuations with limited amplitude and frequency concentrated in a specific narrow band range. The extracted traveling wave signal characteristics can be used as the first traveling wave signal characteristics. However, when a single-phase ground fault occurs about 3km from the beginning of the line, but without any accompanying switching operation, the current and voltage traveling wave signals recorded by the sensor show transient characteristics with significantly increased amplitude, steep wavefront, and wideband energy. The extracted traveling wave signal characteristics can be used as the second traveling wave signal characteristics. If the circuit breaker at the beginning of the line trips due to protection action at the same time as the single-phase ground fault occurs, the current and voltage traveling wave signals captured by the sensor contain both strong components of the fault transient and high-frequency oscillation components brought about by the switching operation. The extracted signal characteristics can be used as the third traveling wave signal characteristics.

[0029] It is important to emphasize that, to ensure effective comparison of the three types of features under the same time base, the start time must be uniformly set for the switching contact action time, fault trigger time, or circuit breaker tripping time, and strict time alignment processing must be performed on the three types of traveling wave signal features. Through this process, it is possible to clearly distinguish between normal switching operation, pure fault, and fault-superimposed switching operation in real-world applications, providing accurate data support for subsequent traveling wave correlation analysis and fault detection model construction. Preferably, a traveling wave sample set is constructed by preprocessing and time-series segmentation based on the traveling wave signal features. The time-series features of the traveling wave samples are captured based on the traveling wave sample set, and a one-dimensional traveling wave feature row vector is generated to construct the traveling wave feature sample set. The specific steps are as follows: The traveling wave signal features are obtained separately, and preprocessing and time-series segmentation are performed on the traveling wave signal features to obtain a single-window traveling wave feature set. ,in, The characteristics of the traveling wave signal at time t are... To construct a traveling wave sample set based on the single-window traveling wave characteristics, the sample set is defined with a single window length. , For the i-th single-window traveling wave feature set, The number of features in a single-window traveling wave feature set; Based on traveling wave sample set A time series feature analysis model for traveling wave samples is constructed to capture the time series features of traveling wave samples corresponding to the single-window traveling wave feature set. One-dimensional traveling wave feature row vectors are generated based on the temporal characteristics of the traveling wave samples to form a traveling wave feature sample set.

[0030] In the practical application of fault detection on overhead power lines in a distribution network, this invention uses a single-phase grounding fault on a 10kV overhead line during thunderstorms as an example. First, transient traveling wave signal characteristics of current and voltage are acquired using high-speed sensor arrays installed at both ends of the feeder. For ease of analysis, these traveling wave signal characteristics undergo preprocessing operations, including noise filtering, amplitude normalization, and feature alignment, to ensure data stability and comparability. Subsequently, the continuous traveling wave signal characteristics are divided into fixed time windows, such as 20μs, to obtain a single-window traveling wave feature set. For example, a window may contain T=200 sampling points, and the corresponding set is represented as... ,in This represents the eigenvalue of the traveling wave at time t. As time progresses, multiple such windows are segmented and sequentially form a complete set of traveling wave samples. Each of these features corresponds to a local time-series traveling wave characteristic. Next, a time-series feature analysis model is constructed for the traveling wave sample set to capture the dynamic changes within each single window. Finally, the time-series features of each window are compressed and represented as a one-dimensional traveling wave feature row vector, thus forming a traveling wave feature sample set. This process enables rapid differentiation of different types of traveling wave patterns in real-world operating environments, thereby supporting subsequent fault location, switching operation interference elimination, and intelligent criterion design.

[0031] Preferably, based on the traveling wave sample set A time-series feature analysis model for traveling wave samples is constructed to capture the time-series features of traveling wave samples corresponding to the single-window traveling wave feature set. The calculation formula for the time-series feature analysis model for traveling wave samples is as follows: In the formula: For the time series features of the traveling wave samples corresponding to the i-th single-window traveling wave feature set, For single window length, Let be the maximum value of the traveling wave signal feature within the i-th single-window traveling wave feature set. Let be the minimum value of the traveling wave signal feature within the i-th single-window traveling wave feature set. The traveling wave signal features at time t+1 within the i-th single-window traveling wave feature set. Let be the traveling wave signal feature at time t within the i-th single-window traveling wave feature set.

[0032] This invention is illustrated in a real-world application of single-phase grounding fault detection on a 10kV overhead line, using a lightning strike as an example. After the sensor arrays at both ends of the line collect transient traveling wave signals of current and voltage, multiple single-window traveling wave feature sets are first obtained through time-series segmentation, such as each window having a length of T=100 sampling points. Taking a certain window i as an example, the characteristic values ​​of the sampled traveling wave signal range from −2.3kV to 3.1kV, with the corresponding maximum value... minimum value The amplitude normalization factor of this window can be obtained through calculation. Meanwhile, for two adjacent sampling points within the window, such as the traveling wave signal at time t... and the traveling wave signal at time t+1 It can calculate its relative rate of change. Multiply the relative rates of change of all adjacent points by the magnitude normalization factor and average them over the entire window to obtain the time-series characteristic value of that window. In this way, the traveling wave feature set for each single window is obtained. Each corresponds to a unique This results in a one-dimensional time-series characteristic row vector sequence. In actual line operation, this method can be used to quickly capture the dynamic changes of traveling wave signals within different time windows, effectively distinguishing between fault traveling waves and switching operation interference traveling waves, thus improving the accuracy and robustness of fault location.

[0033] Preferably, the traveling wave feature sample set includes a traveling wave feature sample set composed of the time-series features of the traveling wave samples obtained based on the features of the first traveling wave signal, and is labeled as the switching operation sample reference set. ,in, Let i be the set of single-window traveling wave features in the switch operation sample reference set. This is the nth single-window traveling wave feature set in the switch operation sample reference set. The number of traveling wave feature sets in a single window; the traveling wave feature sample set composed of the traveling wave sample time-series features obtained based on the second traveling wave signal features is labeled as the fault sample reference set. ,in, Let i be the set of single-window traveling wave features in the fault sample reference set. The nth single-window traveling wave feature set in the fault sample reference set; the traveling wave feature sample set composed of the traveling wave sample time series features obtained based on the third traveling wave signal features is labeled as the mixed sample reference set. ,in, Let i be the set of traveling wave features in the i-th single-window sample reference set. Let n be the nth single-window traveling wave feature set in the mixed sample reference set.

[0034] To improve the accuracy of overhead line fault detection, this invention employs high-speed sampling sensor arrays at both ends of the line to collect transient traveling wave signals of current and voltage. The collected signals are then classified and feature extracted according to different operating conditions. First, under fault-free conditions, maintenance personnel perform multiple switching operations on the line and record the current and voltage traveling wave signals. By constructing a time window, the timing features of the traveling wave samples are extracted, ultimately forming a reference set of switching operation samples. Each of them The feature vector corresponds to a single-phase ground fault. Secondly, in actual single-phase ground faults, researchers also extracted traveling wave timing features using time windows to generate a fault sample reference set. ,in, The characteristics of the fault traveling wave within this window are described. Finally, under the complex operating condition of a ground fault occurring during thunderstorms and accompanied by switch tripping, both the fault traveling wave and the interference traveling wave caused by switch operation are simultaneously collected. After feature extraction, a mixed sample reference set is obtained. By comparison and The correlation between the features of each window can effectively remove the traveling wave components that are highly correlated with the switching operation in the mixed samples, thereby retaining the features that are truly related to the fault event and improving the accuracy and robustness of fault location and classification.

[0035] Preferably, by constructing a switch operation sample reference set, a fault sample reference set, and a mixed sample reference set, first correlation features and second correlation features are extracted respectively. Based on the fusion and discrimination of the two types of correlation features, it is determined whether the traveling wave generated by the switch operation affects the traveling wave of the faulty line. The specific steps are as follows: The switching operation sample reference set, the fault sample reference set, and the mixed sample reference set are extracted from the traveling wave feature sample set respectively; A first traveling wave signal correlation analysis model is constructed by using a switch operation sample reference set and a mixed sample reference set, and the first correlation feature is obtained by fusion analysis. A second traveling wave signal correlation analysis model is constructed by using a fault sample reference set and a mixed sample reference set, and the second correlation feature is obtained by fusion analysis. Based on the first and second correlation features, it is determined whether the traveling wave generated by the switch operation affects the traveling wave of the faulty line.

[0036] In actual power distribution network operation, when a line fault occurs, switching operations may also trigger transient traveling waves, causing the fault traveling wave and the switching operation traveling wave to overlap, increasing the difficulty of fault identification. To accurately determine the impact of switching operations on the fault traveling wave, a traveling wave feature sample set can be constructed. On a 10kV overhead line, when a single-phase ground fault occurs and maintenance personnel perform switching operations, the system collects three types of traveling wave data: a switching operation sample reference set... Fault Sample Reference Set And a mixed sample reference set containing switching operations and fault superposition. First, features of the traveling wave signal, such as wavefront amplitude, spectral energy distribution, duration, and attenuation characteristics, are extracted from each type of sample. Then, a first traveling wave signal correlation analysis model is constructed using a switching operation sample reference set and a mixed sample reference set. By comparing the differences in the corresponding window features of the two sets and performing weighted fusion, the first correlation feature is obtained. This reflects the contribution of the switching operation traveling wave to the mixed traveling wave. Similarly, a second traveling wave signal correlation analysis model is constructed using the fault sample reference set and the mixed sample reference set to obtain the second correlation features. This is used to quantify the performance of the fault traveling wave in the mixed signal. Finally, combining the first and second correlation features, and by setting a threshold and fusion judgment rules, it is determined whether the traveling wave generated by the switching operation has a significant impact on the traveling wave of the faulty line. If the first correlation feature is significantly higher than the threshold and the second correlation feature shows that the fault traveling wave is masked or shifted, it can be determined that the switching operation traveling wave interferes with the fault analysis. Therefore, this interference can be corrected or eliminated in subsequent fault location or protection actions to ensure the accuracy and reliability of fault identification.

[0037] Preferably, a first traveling wave signal correlation analysis model is constructed by using a switch operation sample reference set and a mixed sample reference set, and the first correlation feature is obtained through fusion analysis. The specific steps are as follows: Obtain a reference set of switch operation samples Extract the reference set of switch operation samples. With mixed sample reference set Based on the time-series characteristics of all traveling wave samples in the two reference sets, the minimum absolute difference is calculated to determine the single-window traveling wave feature set corresponding to the minimum absolute difference. A first traveling wave signal correlation analysis model is then constructed, and the first correlation feature is obtained through fusion analysis. The formula for the first traveling wave signal correlation analysis model is as follows: In the formula: As the first associated feature, The time window length, Let be the traveling wave signal characteristics at time t in the j-th single-window traveling wave feature set of the switch operation sample reference set. Let be the traveling wave signal feature at time t in the j-th single-window traveling wave feature set of the mixed sample reference set.

[0038] Preferably, the minimum absolute difference is calculated based on the time-series characteristics of the traveling wave samples in the two sample reference sets, and the single-window traveling wave feature set corresponding to the minimum absolute difference is determined. The formula for calculating the minimum absolute difference is: ; In the formula: For the reference set of switch operation samples With mixed sample reference set The minimum absolute difference of the time-series features of the traveling wave samples in the dataset is denoted as the j-th single-window traveling wave feature set. For mixed sample reference set The time series features of the traveling wave sample corresponding to the i-th single-window traveling wave feature set in the data. For the reference set of switch operation samples The time series features of the traveling wave sample corresponding to the i-th single-window traveling wave feature set in the data. For the reference set of switch operation samples With mixed sample reference set The minimum absolute difference of the time series characteristics of the traveling wave samples in the data.

[0039] In a power distribution network fault monitoring exercise, this invention records a short-circuit fault on a certain line, and simultaneously, nearby switching equipment operates within the same time period. To determine whether the traveling wave generated by the switching operation affects the traveling wave characteristics of the faulty line, three sample reference sets need to be constructed first: a switching operation sample reference set, a fault sample reference set, and a mixed sample reference set.

[0040] In practice, traveling wave characteristics are extracted from the historical monitoring data of the switch operation sample reference set, including the timing characteristics of the current and voltage traveling wave signals generated at the moment the switch is closed or opened; the fault sample reference set includes the traveling wave signal characteristics of historical fault lines under different fault types; and the mixed sample reference set contains the traveling wave signal characteristics when faults and switch operations occur simultaneously.

[0041] A first traveling wave signal correlation analysis model is constructed by using a switching operation sample reference set and a mixed sample reference set. The traveling wave time series characteristics of the two sample sets are compared, and the absolute difference between each single-window traveling wave feature set in the time series is calculated. The single-window traveling wave feature set corresponding to the smallest absolute difference is selected to extract the first correlation feature. This method reflects the consistency between the performance of the switching operation traveling wave in the mixed sample and the traveling wave characteristics of the switching operation itself.

[0042] Meanwhile, a second traveling wave signal correlation analysis model is constructed by using the fault sample reference set and the mixed sample reference set. The method is the same as described above. The second correlation feature is extracted to reflect the consistency between the performance of the traveling wave of the faulted line in the mixed sample and the characteristics of the fault itself.

[0043] The method involves fusing the first and second correlation features for discrimination. If the first correlation feature is high while the second correlation feature is low, it indicates that the traveling wave in the mixed sample is mainly affected by switch operation, and the traveling wave of the faulty line is less affected by interference. Conversely, if the second correlation feature is high while the first correlation feature is low, it indicates that the traveling wave in the mixed sample is mainly generated by the faulty line itself and is not significantly related to switch operation. If both are high, it indicates that there is a certain superposition effect between the traveling wave caused by switch operation and the traveling wave of the faulty line. This method can effectively distinguish the traveling wave interference caused by switch operation from the true characteristics of the faulty traveling wave, providing an accurate basis for subsequent fault location and analysis.

[0044] It should be noted that a second traveling wave signal correlation analysis model is constructed by using a fault sample reference set and a mixed sample reference set. The second correlation feature is obtained through fusion analysis. The calculation steps for the second correlation feature are the same as those for the first correlation feature, and the specific steps are as follows: Obtain a reference set of fault samples Extract the fault sample reference set With mixed sample reference set Based on the time-series characteristics of all traveling wave samples in the two reference sets, the minimum absolute difference is calculated to determine the single-window traveling wave feature set corresponding to the minimum absolute difference. A second traveling wave signal correlation analysis model is then constructed, and the second correlation features are obtained through fusion analysis. The formula for the second traveling wave signal correlation analysis model is as follows: In the formula: This is the second association feature. The time window length, The traveling wave signal features at time t are located in the j-th single-window traveling wave feature set of the fault sample reference set. Let be the traveling wave signal feature at time t in the j-th single-window traveling wave feature set of the mixed sample reference set.

[0045] Based on the fault sample reference set With mixed sample reference set The minimum absolute difference is calculated from the time-series features of the traveling wave samples to determine the single-window traveling wave feature set corresponding to the minimum absolute difference. The formula for calculating the minimum absolute difference is as follows: In the formula: For fault sample reference set With mixed sample reference set The minimum absolute difference of the time-series features of the traveling wave samples in the dataset is denoted as the j-th single-window traveling wave feature set. For the time series features of the traveling wave sample corresponding to the i-th single-window traveling wave feature set in the mixed sample reference set, For the time series features of the traveling wave sample corresponding to the i-th single-window traveling wave feature set in the fault sample reference set, For fault sample reference set With mixed sample reference set The minimum absolute difference of the time series characteristics of the traveling wave samples in the data.

[0046] In this embodiment of the invention, traveling wave signals generated by switching operations exist. To determine the characteristics of the traveling wave in a faulty line within a mixed signal, a second traveling wave signal correlation analysis model needs to be constructed using a fault sample reference set and a mixed sample reference set. Specifically, a fault sample reference set is first extracted from historical fault data, including the time-series characteristics of current and voltage traveling wave signals under different fault types. Simultaneously, the mixed sample reference set records the traveling wave signal characteristics that may be affected by external factors such as switching operations during the fault occurrence. Each single-window traveling wave feature set of these two sample sets is compared, their absolute differences over time are calculated, and the single-window traveling wave feature set corresponding to the smallest absolute difference is selected. This process ensures that the traveling wave signal segment closest to the fault's own characteristics is selected from the mixed samples, thus reflecting the performance of the faulty line's traveling wave in the mixed signal.

[0047] Subsequently, using the single-window traveling wave feature set corresponding to these minimum absolute differences, a second traveling wave signal correlation analysis model is constructed, and the second correlation feature is obtained by fusion calculation. This correlation feature quantifies the strength and consistency of the fault traveling wave features in the mixed signal, providing a basis for further judging whether the fault traveling wave in the mixed signal is affected by other external traveling wave interference.

[0048] Preferably, the determination of whether the traveling wave generated by the switch operation affects the traveling wave of the faulty line based on the first correlation feature and the second correlation feature involves the following steps: calculating the mixed sample reference set. Standard deviation Based on standard deviation Set the distinction threshold , For adjustment coefficients; The first associated feature With the second associated feature The difference Compared with the discrimination threshold, if the first associated feature With the second associated feature The difference If the value is greater than or equal to the distinguishing threshold, the traveling wave generated by the switching operation will affect the traveling wave of the faulty line; if the first associated feature... With the second associated feature The difference If the value is less than the distinguishing threshold, the traveling wave generated by the switching operation will not affect the traveling wave of the faulty line.

[0049] In this embodiment of the invention, a short-circuit fault is recorded on a certain line, and a traveling wave signal generated by a switch operation is also present. In order to determine the characteristics of the traveling wave of the faulty line in the mixed signal, a second traveling wave signal correlation analysis model needs to be constructed by using the fault sample reference set and the mixed sample reference set.

[0050] In practice, a fault sample reference set is first extracted from historical fault data, including the time-series characteristics of current and voltage traveling wave signals under different fault types. Simultaneously, a mixed sample reference set records the traveling wave signal characteristics that may have been affected by external factors such as switching operations during the fault occurrence. Each single-window traveling wave feature set from these two sample sets is compared, their absolute differences over time are calculated, and the single-window traveling wave feature set corresponding to the smallest absolute difference is selected. This process ensures that the traveling wave signal segment closest to the fault's own characteristics is selected from the mixed sample, thus reflecting the performance of the faulty line's traveling wave in the mixed signal.

[0051] Subsequently, using the single-window traveling wave feature set corresponding to these minimum absolute differences, a second traveling wave signal correlation analysis model was constructed, and the second correlation feature was obtained by fusion calculation. This correlation feature quantifies the strength and consistency of the fault traveling wave features in the mixed signal, providing a basis for further judging whether the fault traveling wave in the mixed signal is affected by other external traveling wave interference.

[0052] By comparing the first and second correlation features, the source of the traveling wave signal in the mixed sample can be distinguished. For example, when the second correlation feature is significantly higher than the first correlation feature, it indicates that the mixed signal is mainly generated by the faulty line itself and has little relation to the interference of switch operation. If both the second and first correlation features are high, it indicates that there is a certain superposition effect between the fault traveling wave and the switch operation traveling wave in the mixed signal. This method can accurately extract the true traveling wave characteristics of the faulty line, providing a reliable basis for subsequent fault location and analysis.

[0053] A fault detection system for overhead power distribution lines based on double-ended traveling waves includes a traveling wave signal feature extraction module, a traveling wave feature sample aggregation module, and a traveling wave feature influence discrimination module; the traveling wave signal feature extraction module is connected to the traveling wave feature sample aggregation module, and the traveling wave feature sample aggregation module is connected to the traveling wave feature influence discrimination module.

[0054] The traveling wave signal feature extraction module is used to acquire the current traveling wave signal and voltage traveling wave signal at the line endpoints through the sensor group, and extract the traveling wave signal features of the current traveling wave signal and voltage traveling wave signal respectively. The traveling wave feature sample aggregation module performs preprocessing and time-series segmentation based on the traveling wave signal features to construct a traveling wave sample set. It captures the time-series features of the traveling wave samples based on the traveling wave sample set and generates a one-dimensional traveling wave feature row vector to construct the traveling wave feature sample set. The traveling wave feature impact discrimination module extracts the first and second associated features from the traveling wave feature sample set, and determines whether the traveling wave generated by the switch operation affects the traveling wave of the fault line based on the fusion discrimination of the two types of associated features.

[0055] like Figure 2 The diagram shown is a system block diagram of a fault detection system for overhead lines in a distribution network based on a double-ended traveling wave, according to an embodiment of the present invention. Correspondingly, it can be used to execute... Figure 1 The steps in the method embodiments shown are implemented in a similar manner and have similar technical effects, and will not be repeated here.

[0056] Through the above embodiments, the sensor group of the present invention acquires the current traveling wave signal and the voltage traveling wave signal at the line endpoint, and extracts the traveling wave signal features of the current traveling wave signal and the voltage traveling wave signal respectively; based on the traveling wave signal features, preprocessing and time-series segmentation are performed to construct a traveling wave sample set; the traveling wave sample time-series features are captured according to the traveling wave sample set, and a one-dimensional traveling wave feature row vector is generated to construct a traveling wave feature sample set; based on the traveling wave feature sample set, the first correlation feature and the second correlation feature are extracted respectively, and based on the fusion and discrimination of the two types of correlation features, it is determined whether the traveling wave generated by the switching operation affects the traveling wave of the faulty line.

[0057] This invention acquires current and voltage traveling wave signals by deploying high-precision sensor groups at both ends of the line and extracts their traveling wave features. This allows the method to capture the high-frequency transient characteristics generated by the line at the moment of a fault, enabling rapid response to fault events. Compared to traditional methods based on steady-state quantities or single-end measurements, this approach can acquire fault information within milliseconds, improving the timeliness of fault detection. Preprocessing and time-series segmentation based on traveling wave signal features construct a traveling wave sample set and capture its temporal features. A one-dimensional traveling wave feature row vector is then generated to construct the traveling wave feature sample set. This allows the method to systematically characterize the dynamic changes of the traveling wave over time, effectively reducing background noise interference and enhancing the accuracy and robustness of fault feature extraction. By extracting the first and second correlation features of switch operation samples and fault samples respectively, and fusing these two types of correlation features for discrimination, the method can identify the difference between the interference traveling wave generated by the switch operation and the actual fault traveling wave, thereby eliminating interference components. This not only improves the accuracy of fault initial wave identification but also avoids false positives and false negatives, improving the reliability of fault location.

[0058] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

[0059] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for fault detection of overhead lines in a distribution network based on double-ended traveling waves, characterized in that, Includes the following steps: The current traveling wave signal and voltage traveling wave signal at the line endpoints are acquired by a sensor array, and the traveling wave signal characteristics of the current traveling wave signal and voltage traveling wave signal are extracted respectively. Based on the characteristics of the traveling wave signal, a traveling wave sample set is constructed by preprocessing and time-series segmentation. The time-series characteristics of the traveling wave samples are captured based on the traveling wave sample set, and a one-dimensional traveling wave feature row vector is generated to construct the traveling wave feature sample set. The first and second associated features are extracted from the traveling wave feature sample set, and the traveling wave generated by the switch operation is determined to affect the traveling wave of the faulty line based on the fusion and discrimination of the two types of associated features.

2. The method for detecting faults in overhead lines of a distribution network based on a double-ended traveling wave, as described in claim 1, is characterized in that... The characteristics of traveling wave signals include the first traveling wave signal characteristics, the second traveling wave signal characteristics, and the third traveling wave signal characteristics.

3. The method for fault detection of overhead lines in a distribution network based on a double-ended traveling wave as described in claim 2, characterized in that, The traveling wave signal characteristics of the current traveling wave signal and the voltage traveling wave signal during switching operation under fault-free line conditions are obtained and used as the first traveling wave signal characteristics. The characteristics of the traveling wave signals of the current and voltage traveling waves under line fault conditions without switching operations are obtained and used as the second traveling wave signal characteristics. The traveling wave signal characteristics of the current traveling wave signal and the voltage traveling wave signal during switching operations under line fault conditions are acquired and used as the third traveling wave signal characteristics.

4. The method for detecting faults in overhead lines of a distribution network based on double-ended traveling waves according to claim 1, characterized in that, Based on the characteristics of the traveling wave signal, a traveling wave sample set is constructed through preprocessing and time-series segmentation. The time-series features of the traveling wave samples are then captured from the sample set, and a one-dimensional traveling wave feature row vector is generated to construct the traveling wave feature sample set. The specific steps are as follows: The traveling wave signal features are obtained separately, and preprocessing and time-series segmentation are performed on the traveling wave signal features to obtain a single-window traveling wave feature set. ,in, The characteristics of the traveling wave signal at time t are... To construct a traveling wave sample set based on the single-window traveling wave characteristics, the sample set is defined with a single window length. , For the i-th single-window traveling wave feature set, The number of features in a single-window traveling wave feature set; Based on traveling wave sample set A time series feature analysis model for traveling wave samples is constructed to capture the time series features of traveling wave samples corresponding to the single-window traveling wave feature set. One-dimensional traveling wave feature row vectors are generated based on the temporal characteristics of the traveling wave samples to form a traveling wave feature sample set.

5. The method for detecting faults in overhead lines of a distribution network based on double-ended traveling waves according to claim 4, characterized in that, The traveling wave feature sample set includes the traveling wave feature sample time series features obtained based on the features of the first traveling wave signal, and is labeled as the switching operation sample reference set. ,in, Let i be the set of single-window traveling wave features in the switch operation sample reference set. This is the nth single-window traveling wave feature set in the switch operation sample reference set. The number of traveling wave feature sets in a single window; the traveling wave feature sample set composed of the traveling wave sample time-series features obtained based on the second traveling wave signal features is labeled as the fault sample reference set. ,in, Let i be the set of single-window traveling wave features in the fault sample reference set. The nth single-window traveling wave feature set in the fault sample reference set; the traveling wave feature sample set composed of the traveling wave sample time series features obtained based on the third traveling wave signal features is labeled as the mixed sample reference set. ,in, Let i be the set of traveling wave features in the i-th single-window sample reference set. Let n be the nth single-window traveling wave feature set in the mixed sample reference set.

6. The method for detecting faults in overhead lines of a distribution network based on a double-ended traveling wave, as described in claim 1, is characterized in that... By constructing a reference set of switch operation samples, a reference set of fault samples, and a mixed reference set, the first correlation feature and the second correlation feature are extracted respectively. Based on the fusion and discrimination of the two types of correlation features, it is determined whether the traveling wave generated by the switch operation affects the traveling wave of the faulty line. The specific steps are as follows: The switching operation sample reference set, the fault sample reference set, and the mixed sample reference set are extracted from the traveling wave feature sample set respectively; A first traveling wave signal correlation analysis model is constructed by using a switch operation sample reference set and a mixed sample reference set, and the first correlation feature is obtained by fusion analysis. A second traveling wave signal correlation analysis model is constructed by using a fault sample reference set and a mixed sample reference set, and the second correlation feature is obtained by fusion analysis. Based on the first and second correlation features, it is determined whether the traveling wave generated by the switch operation affects the traveling wave of the faulty line.

7. The method for fault detection of overhead lines in a distribution network based on a double-ended traveling wave as described in claim 6, characterized in that, The first traveling wave signal correlation analysis model is constructed by using a switch operation sample reference set and a mixed sample reference set. The first correlation feature is obtained through fusion analysis. The specific steps are as follows: Obtain a reference set of switch operation samples ,in, Let i be the set of single-window traveling wave features in the switch operation sample reference set. For the nth single-window traveling wave feature set in the switch operation sample reference set, extract the switch operation sample reference set. With mixed sample reference set The time series characteristics of all traveling wave samples in the data, among which, Let i be the set of traveling wave features in the i-th single-window sample reference set. For the nth single-window traveling wave feature set in the mixed sample reference set, calculate the minimum absolute difference based on the time series characteristics of the traveling wave samples in the two sample reference sets, determine the single-window traveling wave feature set corresponding to the minimum absolute difference, construct the first traveling wave signal correlation analysis model, and obtain the first correlation feature through fusion analysis.

8. The method for fault detection of overhead lines in a distribution network based on a double-ended traveling wave as described in claim 7, characterized in that, The formula for the correlation analysis model of the first traveling wave signal is: ; In the formula: As the first associated feature, The time window length, Let be the traveling wave signal characteristics at time t in the j-th single-window traveling wave feature set of the switch operation sample reference set. Let be the traveling wave signal feature at time t in the j-th single-window traveling wave feature set of the mixed sample reference set.

9. The method for fault detection of overhead lines in a distribution network based on a double-ended traveling wave as described in claim 6, characterized in that, The specific steps for determining whether the traveling wave generated by the switch operation affects the traveling wave of the faulty line based on the first and second correlation features are as follows: Calculate the mixed sample reference set Standard deviation ,in, Let i be the set of traveling wave features in the i-th single-window sample reference set. For the nth single-window traveling wave feature set in the mixed sample reference set, based on standard deviation Set the distinction threshold , For adjustment coefficients; A second traveling wave signal correlation analysis model is constructed using a fault sample reference set and a mixed sample reference set, and the second correlation features are obtained through fusion analysis. ; The first associated feature With the second associated feature The difference Compared with the discrimination threshold, if the first associated feature With the second associated feature The difference If the value is greater than or equal to the distinguishing threshold, the traveling wave generated by the switching operation will affect the traveling wave of the faulty line; if the first associated feature... With the second associated feature The difference If the value is less than the distinguishing threshold, the traveling wave generated by the switching operation will not affect the traveling wave of the faulty line.

10. A fault detection system for overhead lines in a distribution network based on a double-ended traveling wave, applied to the fault detection method for overhead lines in a distribution network based on a double-ended traveling wave as described in any one of claims 1-9, characterized in that, The system includes: a traveling wave signal feature extraction module, a traveling wave feature sample aggregation module, and a traveling wave feature influence discrimination module; The traveling wave signal feature extraction module is used to acquire the current traveling wave signal and voltage traveling wave signal at the line endpoints through the sensor group, and extract the traveling wave signal features of the current traveling wave signal and voltage traveling wave signal respectively. The traveling wave feature sample aggregation module performs preprocessing and time-series segmentation based on the traveling wave signal features to construct a traveling wave sample set. It captures the time-series features of the traveling wave samples based on the traveling wave sample set and generates a one-dimensional traveling wave feature row vector to construct the traveling wave feature sample set. The traveling wave feature impact discrimination module extracts the first and second associated features from the traveling wave feature sample set, and determines whether the traveling wave generated by the switch operation affects the traveling wave of the fault line based on the fusion discrimination of the two types of associated features.

Images