Power distribution network fault traveling wave positioning method and system

Abstract

The invention relates to the technical field of power distribution network fault positioning, in particular to a power distribution network fault traveling wave positioning method and system, and the method comprises the steps: synchronously collecting a transient signal through a distributed traveling wave sensing unit, and extracting an initial traveling wave arrival time; constructing an adaptive wave velocity model based on line physical parameters to calculate an equivalent propagation velocity; dynamically generating a traveling wave propagation tree according to the real-time switch state to determine an effective monitoring path; based on the propagation path and the actual arrival time, an optimization algorithm is adopted to determine a preliminary fault point; and calculating a consistency index of the polarity of the traveling waves of the adjacent nodes, starting a multi-hypothesis tracking mechanism in a preliminary point neighborhood when the consistency index is lower than a threshold value, performing confidence coefficient weighted evaluation by combining a traveling wave attenuation model, and outputting an accurate positioning result. According to the method, through cooperation of dynamic topology mapping, adaptive wave velocity correction and multi-source closed-loop verification, the precision, reliability and adaptive capability of fault positioning in a complex power distribution network environment are remarkably improved, and the operation and maintenance requirements of rapid self-healing are met.

Description

Technical Field

[0001] This invention relates to the field of power system fault diagnosis and location technology, specifically to a method and system for locating traveling waves in a distribution network fault. Background Technology

[0002] With the ongoing development of smart distribution networks, rapid fault location and isolation capabilities have become a key requirement for improving power supply reliability. The traveling wave method, due to its advantages of fast response and high location accuracy, has been widely applied in the field of power system fault location.

[0003] However, directly applying the traveling wave method to the distribution network faces many challenges, and existing technical solutions mainly have the following shortcomings:

[0004] High dependence on time synchronization: Mainstream solutions rely on global satellite navigation systems (such as GPS and BeiDou) to provide microsecond-level time synchronization. However, in urban underground utility tunnels, densely built-up areas, or in severe weather, satellite signals are easily blocked or interfered with, leading to synchronization loss and positioning failure.

[0005] The wave velocity model is overly simplistic: most methods use a fixed wave velocity or only roughly set the wave velocity value based on the line type (cable / overhead line). In reality, the propagation speed of traveling waves in cables is affected by many factors such as insulation material, ambient temperature, service life, and laying method, and varies significantly dynamically. Using a fixed wave velocity introduces systematic errors that are difficult to correct, especially in cable lines, where positioning accuracy is severely reduced.

[0006] Insufficient topology adaptability: Distribution networks have numerous branches and flexible operating modes. Existing methods often perform location calculations based on preset static topologies, which cannot effectively handle real-time topology changes caused by switch relocations. This can easily lead to misjudgments near branch points and loop closure points, or failure to identify effective monitoring nodes.

[0007] Weak anti-interference and fault tolerance capabilities: Traveling wave signals in distribution networks attenuate rapidly and are susceptible to interference from load switching, distributed power source access, line reflection and refraction, making it difficult to accurately identify the traveling wave front. Existing solutions mostly rely on single signal processing or threshold judgment, which lacks reliability in complex electromagnetic environments and lacks an effective verification mechanism for the credibility of positioning results.

[0008] Therefore, existing technologies are insufficient to simultaneously meet the requirements for high-precision, high-reliability, and highly adaptive fault location in complex power distribution network environments. Summary of the Invention

[0009] This invention aims to overcome the shortcomings of existing technologies and provide a method and system for fault traveling wave location in distribution networks. Through the synergy of three technologies—dynamic topology mapping, adaptive wave velocity correction, and multi-source closed-loop verification—it constructs a fault traveling wave location system that does not rely on satellite time synchronization, can adapt to the changing operating modes of distribution networks, and has strong anti-interference capabilities.

[0010] To achieve the above objectives, the present invention adopts the following technical solution:

[0011] In a first aspect, the present invention provides a method for locating traveling wave faults in a power distribution network, comprising the following steps:

[0012] Step S1: Within a preset high-frequency sampling period, transient voltage and current signals are synchronously acquired by multiple distributed traveling wave sensing units deployed at key nodes of the distribution network; Step S2: When any distributed traveling wave sensing unit detects a fault triggering condition, data backtracking is initiated to acquire waveform data of each node, time-frequency analysis is performed on the waveform data, and the initial traveling wave arrival time of each node is extracted. For overhead line sections, the traveling wave propagation speed is taken as a fixed value of 0.97 times the speed of light; Step S3: Based on the line physical parameter database containing information on line type, conductor material, insulation medium, ambient temperature, and service life, an adaptive wave velocity model is constructed to calculate the equivalent propagation speed of the fault path. Step S4: Based on the real-time switch status information obtained from the distribution automation system through the standard communication interface, construct the connectivity topology diagram under the current operating mode. Using the node that first triggers fault recording as the root node, generate a traveling wave propagation tree using a depth-first search algorithm. This step can adapt in real-time to network topology changes caused by switch displacement, accurately identify effective monitoring nodes and traveling wave propagation paths, fundamentally avoiding misjudgments in location near branch points and loop points caused by using a static topology model. Step S5: Based on the electrical path defined by the traveling wave propagation tree, calculate the actual arrival time of each node. Based on the equivalent propagation speed The calculated theoretical arrival time is compared, and the initial fault location is determined on the propagation tree path by minimizing the sum of squared residuals; Step S6: To cope with the complex multipath reflection and noise interference in the distribution network and improve the confidence of the location result, the consistency index CI of the polarity of the first wave of the traveling wave of adjacent monitoring nodes in the traveling wave propagation tree is calculated. When CI is lower than a preset threshold, it is determined that there is multipath interference, and a multi-hypothesis tracking mechanism is activated: multiple candidate intervals are generated in the neighborhood of the initial fault point, and the optimization search in step S5 is re-executed in each candidate interval. The confidence weight evaluation of the search results is combined with the traveling wave attenuation model, and the optimal solution is selected as the accurate location result; Step S7: The accurate location result is output.

[0013] Furthermore, in step S3, for the cable line segment, its wave velocity... Calculated using the following formula: in, At the speed of light, Where is the relative permittivity of the insulation material, T is the real-time ambient temperature, A is the number of years the line has been in operation, and k and m are preset aging coefficients, which are set according to the type of insulation material and operating experience. For example, for XLPE insulated cables, k can be 0.002 / ℃ and m can be 0.015 / year. The aging coefficients k and m can be preset to initial values ​​according to the type of insulation material and can be adaptively adjusted online based on historical positioning deviation data accumulated during system operation, so that the wave velocity model continuously approximates the actual physical state of the line.

[0014] Further, in step S6, the consistency index CI is defined as the proportion of node pairs in the traveling wave propagation tree where the polarity of the first wave of adjacent nodes on the same line segment is the same to the total number of node pairs. Preferably, the preset threshold is 0.7. When CI ≥ 0.7, the traveling wave propagation path is considered clear with minimal interference, and the preliminary positioning result can be directly accepted; when CI < 0.7, the preliminary positioning result is considered to be clear. At a value of 0.7, a warning is issued indicating the possibility of strong reflections or distortions, requiring the activation of a multi-hypothesis tracking mechanism for result verification and optimization.

[0015] Furthermore, in step S6, the attenuation factor of the traveling wave attenuation model... This can be expressed by the following formula: in, The dominant frequency of the traveling wave (unit: MHz); As a reference frequency, we take 1 MHz; In the reference frequency The attenuation value below; The decay growth coefficient is dimensionless. and Based on the line type, a table is consulted. In one specific implementation, the system's pre-set line parameter feature library stores attenuation characteristic parameters corresponding to lines of different models, specifications, and insulation materials. These parameters can be directly derived from the attenuation-frequency characteristic curves specified in the official technical specifications provided by the line manufacturer. The coefficients required for the model can be obtained by fitting these curves.

[0016] Secondly, the present invention provides a power distribution network fault traveling wave location system for implementing the above-mentioned method, comprising: multiple distributed traveling wave sensing units deployed at key nodes of the power distribution network for synchronous signal acquisition; a local edge computing node connected to all the distributed traveling wave sensing units via a communication network, and integrating the following modules working collaboratively: a dynamic wave velocity modeling unit and a real-time topology analysis engine, which together constitute an adaptive coupled model for accurately characterizing the propagation path and velocity of the traveling wave; a preliminary location calculation unit for performing location calculations based on the output of the coupled model; and a multi-source traveling wave consistency verification module for performing closed-loop verification and optimization of the results of the preliminary location calculation unit based on polarity consistency and multiple hypothesis tracking.

[0017] Compared with the prior art, the beneficial effects of the present invention are:

[0018] Based on the real-time switch status, a "traveling wave propagation tree" is dynamically constructed, which can automatically adapt to changes in the operation mode of the distribution network, accurately identify effective monitoring nodes and propagation paths, and fundamentally avoid positioning errors caused by topology changes.

[0019] By creating a polarity consistency index (CI) verification and a multiple hypothesis tracking mechanism, a closed-loop verification of the positioning results is established. This mechanism can effectively identify and suppress interference such as multipath reflection and noise, significantly improving the reliability and confidence of the positioning results under complex working conditions.

[0020] The system architecture is clear, and sensors only need to be deployed at key nodes, eliminating the need for dense deployment across the entire network. It supports standard communication protocols, making it easy to integrate with existing distribution automation master stations, thus reducing deployment and maintenance costs.

[0021] This invention systematically solves the three major sources of uncertainty (propagation speed, propagation path, and signal interference) in distribution network fault location by combining dynamic wave velocity correction, real-time topology mapping, and closed-loop multi-source verification technologies, thereby achieving a significant simultaneous improvement in location accuracy, reliability, and adaptability under complex operating conditions. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the overall architecture of the power distribution network fault traveling wave location system provided in an embodiment of the present invention; Figure 2 This is a flowchart illustrating the principle of dynamic wave velocity modeling and topology mapping in an embodiment of the present invention. Figure 3 This is a flowchart illustrating the logic of multi-source traveling wave consistency verification and precise positioning in an embodiment of the present invention. Detailed Implementation

[0023] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0024] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” and “described” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the described features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0025] Example 1:

[0026] This embodiment provides a traveling wave fault location system for a distribution network, comprising: distributed traveling wave sensing units deployed at key locations such as substation outgoing lines, ring main units, and important branch points; each distributed traveling wave sensing unit adopts a standardized and modular design, which can be flexibly deployed at key nodes such as ring main units and pole-mounted switches, and can be networked using existing or newly built fiber optic communication channels in the distribution network, without requiring large-scale modifications to primary equipment, making engineering implementation convenient and easy to promote; a fiber optic Ethernet network connecting each sensing unit; a local edge computing node as the core processing unit; and a distribution automation master station located in the dispatch center. The sensing units incorporate wideband Rogowski coils and capacitive voltage dividers, with a sampling accuracy of 14 bits. Each unit synchronizes with the master clock deployed in the substation communication room via fiber optic Ethernet using a precise time protocol (IEEE 1588), with a measured synchronization error ≤ ±200 nanoseconds. The edge computing node is an industrial-grade embedded server running a Linux real-time operating system.

[0027] refer to Figure 2 and Figure 3 The positioning method flow in this embodiment is as follows:

[0028] Step S1 (Signal Acquisition and Triggering): Each sensing unit synchronously acquires three-phase voltage and current transient signals at a period of 10 microseconds (sampling rate 100 kHz). The raw data is processed by the digital filter inside the field-programmable gate array (50 Hz power frequency suppression + wavelet threshold denoising), then timestamped with PTP and stored cyclically in the local buffer. When any unit detects that the amplitude of three consecutive sampling points exceeds three times the standard deviation of the fundamental effective value, a fault is determined to have occurred, and a trigger command is immediately sent to the entire network; the triggering condition here is that the sudden change amplitude, rate of change, or energy of voltage or current exceeds a set threshold.

[0029] Step S2 (Traveling Wave Arrival Time Extraction): Each sensing unit uploads waveform data from 2 milliseconds before the fault to 10 milliseconds after the fault to the edge computing node. The node performs a complex Morlet wavelet transform (center frequency) on each channel's data (e.g., phase A voltage). =6, scale range 1-128, corresponding frequency band approximately 7.8 kHz-1 MHz), to obtain the time-frequency energy spectrum.

[0030] In the energy spectrum, search along the time axis for the first local maximum point that satisfies the following condition:

[0031] Along the time axis, for each time point t, a sliding window (41 points in total) is formed by taking N=20 sampling points before and after it. The mean μ(t) and standard deviation σ(t) of the energy values ​​within this window are calculated. Then, the first local maximum point that satisfies the following conditions is searched:

[0032] Energy value > μ(t) + 2σ(t);

[0033] Furthermore, the duration of this high-energy state is ≥ 5 consecutive sampling points (i.e., 50 microseconds).

[0034] The time corresponding to the maximum value is taken as the suspected arrival time of the channel. Clarke transform is performed on all three-phase channels to obtain the zero-mode component; within a window of ±10 microseconds before and after the suspected arrival time t0, the first-order difference of the zero-mode component is calculated; if there are three consecutive sampling points with the absolute value of the first-order difference greater than the preset threshold Δ and the difference signs are consistent, then it is determined that there is a valid polarity change, and t0 is determined as the initial traveling wave arrival time of the sensing node.

[0035] Step S3 (Dynamic Wavespeed Modeling): The dynamic wavespeed modeling unit in the edge computing node is invoked. This unit accesses the locally stored SQLite format line parameter database.

[0036] The fault current flows through the cable segment identified based on the real-time topology. For the cable segment, the relative permittivity of its insulation medium, XLPE, is determined. =2.3, read the real-time temperature T=30℃ reported by the associated temperature sensor via Modbus RTU protocol. Calculate the number of years in operation A = current year - 2015. Substitute into the formula: Calculated .

[0037] For the aging coefficients k and m, their initial values ​​can be set according to the aging characteristic reference values ​​provided by the international or national industry standards corresponding to the insulation material, or by using the time-domain reflectometry method to conduct actual measurements on typical line samples of the same type, and calibrating the coefficients by fitting the wave velocity data of the samples under different temperatures and simulated aging conditions. During system operation, the above coefficients can be iteratively optimized based on feedback from historical fault location data and actual inspection results, thereby achieving adaptive correction of the model.

[0038] For overhead line segments, directly fix the wave speed. Finally, the equivalent wave velocity of the entire fault path is calculated by weighted averaging of the lengths of each segment. .

[0039] Step S4 (Topology Mapping and Propagation Tree Generation): The real-time topology analysis engine obtains the remote signaling status of all circuit breakers and disconnectors from the distribution automation master station in real time via the IEC61850 GOOSE service. The engine abstracts the distribution network as an undirected graph G=(V, E), where vertices V are electrical nodes, edges E are line segments, and edge weights are line lengths. Taking the node where the sensor unit that first triggered the waveform recording is located as the root node, a depth-first search algorithm is used to traverse graph G, automatically ignoring all branches isolated by switches with a connection status of "separate," thereby generating an accurate "traveling wave propagation tree." This tree clearly defines all monitoring nodes that the traveling wave can propagate to under the current operating mode and their propagation path order.

[0040] Step S5 (Preliminary Location): On the main path of the generated propagation tree, assume the electrical distance from the fault point to the root node is x (unknown), and the fault initiation time is... (Unknown). For any monitoring node i on the propagation tree, its theoretical arrival time is: ,in Let x be the path length from the hypothetical failure point to node i (calculated based on the propagation tree and x). Construct the objective function: , where n is the number of effective monitoring nodes on the propagation tree. The particle swarm optimization algorithm is used to minimize... Set the particle swarm size to 50, the maximum number of iterations to 100, and the particle position vector to [x, After optimization and convergence, the initial fault location x and starting time that minimize the sum of squared residuals are obtained. .

[0041] Step S6 (Multi-source verification and precise positioning): The multi-source traveling wave consistency verification module begins operation. First, it extracts the polarity (positive or negative transition) of the first wave of the traveling wave from adjacent nodes on the propagation tree. It then calculates the polarity consistency index CI = (number of adjacent node pairs with the same polarity) / (total number of adjacent node pairs).

[0042] when If the interference is small and the preliminary positioning result is reliable, proceed directly to step S7.

[0043] when If significant multipath interference is detected, a multi-hypothesis tracking mechanism is initiated:

[0044] Centered on the initial location point, extend 500 meters forward and backward to form a candidate section with a total length of 1 kilometer;

[0045] Divide the candidate interval into 10 sub-intervals with a step size of 100 meters. Within each sub-interval, repeat the optimization search in step S5 to obtain 10 candidate solutions. , and the corresponding residuals ;

[0046] In the complex Morlet wavelet transform time-frequency energy spectrum obtained in step S2, locate the time series corresponding to the arrival time t_i of the initial traveling wave; in this time series, select the scale with the largest energy value. And according to the center frequency =6 will scale Convert to frequency ,in For the sampling frequency, this This is the dominant frequency of the traveling wave.

[0047] For each candidate solution According to the location of the fault The line type corresponding to the propagation path to each monitoring node i, and the calculated traveling wave dominant frequency. Determine the attenuation factor of this path. Define the confidence weight of the time information of the i-th monitoring node under this candidate solution. Inversely proportional to the degree of path decay, a feasible definition is: in, To effectively monitor the number of nodes, the weighted sum of squared residuals corresponding to this candidate solution is: From all candidate solutions, select the one that results in the weighted sum of squared residuals. The smallest solution is taken as the precise positioning result.

[0048] Select the weighted residual from all candidate solutions. The smallest candidate solution; if its corresponding maximum absolute time residual exceeds a preset dynamic threshold. If the location fails, an alarm will be issued; Dynamically calculated based on the current total fault path length L and equivalent wave velocity v: , where α is an empirical coefficient, ranging from 0.5% to 1%.

[0049] If, after verification using a multi-hypothesis tracking mechanism, the weighted residuals of all candidate solutions far exceed expectations, or the maximum absolute time residual consistently exceeds the limit, the system can determine that the location confidence level is too low. This may stem from severe anomalies in topology information (such as the fault point actually being located on a branch mistakenly identified as disconnected) or extreme interference with the traveling wave signal. In this case, the system will output a "location failed" flag, along with a preliminary assessment of the fault region and the original polarity sequence of all nodes, and simultaneously issue an alarm, prompting maintenance personnel to conduct manual verification in conjunction with the status of the line sectionalizing switches.

[0050] Step S7 (Result Output and Execution): The precise location result is encapsulated into a standard IEC 61850-7-420 FaultLocation report and uploaded to the distribution automation master station via a TLS encrypted channel. Simultaneously, the edge computing node drives a solid-state relay through its GPIO interface, outputting a hard-contact trip signal to the circuit breaker control circuit of the feeder where the fault point is located, triggering a rapid isolation operation. The measured average time for the entire process is approximately 65 milliseconds, meeting the requirements for rapid self-healing.

[0051] Those skilled in the art will understand that although the above embodiments are mainly illustrated using a single-phase grounding fault in a cable-overhead hybrid line as an example, the technical solution provided by this invention is also applicable to other scenarios. For example, for pure cable networks or pure overhead networks, the dynamic wave velocity modeling method only requires corresponding adjustment of the line type parameters; for fault types such as phase-to-phase short circuits, the traveling wave initial wavefront extraction, polarity determination, and multi-source verification mechanism are still applicable. The core of this invention lies in systematically improving the positioning accuracy and reliability under various distribution network topologies and fault types through a triple collaborative mechanism of dynamic wave velocity correction, real-time topology mapping, and closed-loop multi-source verification.

[0052] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for locating traveling wave faults in a distribution network, characterized in that, The process includes the following steps: S1: Within a preset high-frequency sampling period, transient voltage and current signals are synchronously acquired by multiple distributed traveling wave sensing units deployed at key nodes of the distribution network; S2: When any distributed traveling wave sensing unit detects a fault triggering condition, data backtracking is initiated to acquire waveform data from each node, time-frequency analysis is performed on the waveform data, and the initial traveling wave arrival time of each node is extracted. S3: Based on a database of line physical parameters containing information on line type, conductor material, insulation medium, ambient temperature, and service life, an adaptive wave velocity model is constructed to calculate the equivalent propagation velocity of the fault path. S4: Based on the real-time switch status information of the distribution network, construct the connectivity topology under the current operating mode. Using the node that first triggers fault recording as the root node, generate a traveling wave propagation tree using a depth-first search algorithm. This propagation tree excludes all disconnected branches. S5: Based on the electrical paths defined by the traveling wave propagation tree, calculate the actual arrival time of each node. Based on the equivalent propagation speed The calculated theoretical arrival time is compared, and the initial fault location is determined on the propagation tree path by minimizing the sum of squared residuals; S6: The consistency index CI of the first wave polarity of the traveling wave of adjacent monitoring nodes in the traveling wave propagation tree is calculated. When CI is lower than a preset threshold, it is determined that there is multipath interference, and a multi-hypothesis tracking mechanism is initiated: multiple candidate intervals are generated in the neighborhood of the initial fault point, and the optimization search in step S5 is re-executed in each candidate interval. The confidence weight evaluation of the search results is combined with the traveling wave attenuation model, and the optimal solution is selected as the precise location result; S7: The precise location result is output.

2. The method for locating traveling waves in a distribution network fault according to claim 1, characterized in that, In step S3, for the cable line segment, its wave velocity Through formula Calculate, where, At the speed of light, is the relative permittivity of the insulating material, T is the real-time ambient temperature, A is the number of years the line has been in operation, and k and m are preset aging coefficients.

3. The method for locating traveling waves in a distribution network fault according to claim 1, characterized in that, In step S6, the consistency index CI is defined as the proportion of the number of nodes with the same polarity of the first wave of the traveling wave on the same line segment in the traveling wave propagation tree to the total number of node pairs. The preset threshold is 0.7, and the decision condition for activating the multi-hypothesis tracking mechanism is based on the comparison result between the consistency index CI and the preset threshold.

4. The method for locating traveling waves of faults in a distribution network according to claim 1, characterized in that, In step S1, the distributed traveling wave sensing unit includes a wideband Rogowski coil current transformer and a capacitive voltage divider; the configurable range of the high-frequency sampling period is 1 to 50 microseconds, with a default value of 10 microseconds; each unit is connected via fiber optic Ethernet and time synchronization is achieved using a precision time protocol, with the synchronization error controlled within ±200 nanoseconds.

5. The method for locating traveling waves of faults in a distribution network according to claim 1, characterized in that, In step S5, the optimization search algorithm is a particle swarm optimization algorithm; the particle position vector is... ,in The electrical distance from the fault point to the root node. This is the time when the fault begins.

6. The method for locating traveling waves of faults in a distribution network according to claim 1, characterized in that, In step S6, the attenuation factor of the traveling wave attenuation model The expression is ,in The dominant frequency of the traveling wave. For reference frequency, and The coefficients are obtained by looking up a table based on the line type.

7. The method for locating traveling waves of a distribution network fault according to claim 1, characterized in that, In step S6, in the multi-hypothesis tracking mechanism, the candidate interval is centered on the initial fault point and extended 500 meters before and after it, and divided into several sub-intervals with a step size of 100 meters. The search of the optimized search algorithm is performed independently in each sub-interval.

8. The method for locating traveling waves of a distribution network fault according to claim 1 or 2, characterized in that, In step S3, the physical parameter database of the line supports online updates. The system automatically adjusts the aging coefficients k and m in the adaptive wave velocity model based on the deviation between historical fault location data and actual inspection results.

9. The method for locating traveling waves of a distribution network fault according to claim 1, characterized in that: In step S4, when real-time topology information is abnormal or lost, the system automatically switches to the most recent valid topology snapshot and issues an alarm message.

10. A power distribution network fault traveling wave location system, used to implement the method according to any one of claims 1 to 9, characterized in that, include: Multiple distributed traveling wave sensing units are deployed at key nodes of the distribution network to achieve synchronous signal acquisition. Local edge computing nodes are connected to all the distributed traveling wave sensing units through a communication network and integrate the following modules that work together: a dynamic wave velocity modeling unit and a real-time topology analysis engine, which together constitute an adaptive coupled model for accurately characterizing the propagation path and velocity of traveling waves; a preliminary positioning calculation unit, which performs positioning calculations based on the output of the coupled model; and a multi-source traveling wave consistency verification module, which performs closed-loop verification and optimization of the results of the preliminary positioning calculation unit based on polarity consistency and multiple hypothesis tracking.

Images