A multi-branch power distribution network double-end traveling wave fault location method and system

Abstract

The application relates to a kind of multi-branch power distribution network double-end traveling wave fault location method and system, comprising: based on the deployment of multiple measuring points in the head and tail end and branch terminal of radial power distribution network, the line mode wave velocity is calculated by collecting topological information;Using the line mode and zero mode components extracted from the modulus decoupling of three-phase voltage data after fault;Select the strongest line mode component of global mutation as reference, generate wave head timing table through wavelet decomposition and mutation detection;Combined with topology, timing table and line mode wave velocity, corrected zero mode wave velocity, determine the fault interval and output fault location by using line mode and zero mode wave head time difference compensation correction.This application overcomes the positioning error caused by clock asynchronization and parameter disturbance through the wave head timing analysis and time difference compensation mechanism of multiple measuring points cooperation, realizes accurate ranging without the influence of clock error, significantly improves the ranging accuracy;At the same time, the point distribution requirement is simplified, and only branch single measuring point can realize accurate discrimination and positioning of fault interval, with high economy and deployment flexibility.

Description

Technical Field

[0001] This invention relates to the field of distribution network fault location technology, and in particular to a method and system for locating double-ended traveling wave faults in a multi-control power grid. Background Technology

[0002] Traditional two-terminal traveling wave fault location methods face significant challenges in distribution networks. Due to the complexity of distribution network topology and numerous branches, existing technologies typically rely on high-precision clock synchronization between the two ends of the equipment. Clock drift or communication delays can lead to fixed ranging errors, significantly reducing location accuracy. Furthermore, line parameter disturbances further amplify these errors, making it difficult to meet the demands of intelligent distribution networks for rapid and accurate fault location.

[0003] Therefore, there is an urgent need to provide a technical solution to address the above problems. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method and system for locating double-ended traveling wave faults in a multi-control power grid.

[0005] Firstly, the present invention provides a method for locating a two-terminal traveling wave fault in a multi-control power grid, the technical solution of which is as follows:

[0006] By setting traveling wave detection terminals at the beginning, end and branch terminals of the radial multi-branch power grid, the topology information of the main line and branch lines is collected, and the line mode wave velocity is calculated based on the line parameters.

[0007] The three-phase voltage and current data after the fault were collected at each measuring point, and the three-phase voltage data were decoupled by modulus to extract the line mode component and zero mode component of the initial voltage traveling wave at the fault point.

[0008] Based on the voltage change amplitude of the line mode components at each measuring point, the line mode component with the most significant global jump is selected as the unified analysis benchmark.

[0009] Multi-scale wavelet decomposition and abrupt change detection are performed on the selected linear mode components to extract the arrival time of the linear mode wavefront and the arrival time of the zero mode wavefront at each measurement point, and a wavefront arrival time series table is generated.

[0010] Based on topology information, wavefront arrival time table, linear mode wave velocity, and corrected zero-mode wave velocity, the fault range is determined. During the process of determining the fault range and during the final location, the time difference between the arrival time of the linear mode wavefront and the arrival time of the zero-mode wavefront is used for compensation and correction, and the fault location is output.

[0011] The beneficial effects of the multi-control power grid two-end traveling wave fault location method of the present invention are as follows:

[0012] The method of this invention effectively overcomes the positioning error caused by clock asynchrony and parameter disturbance through wavefront timing analysis and time difference compensation mechanism with multi-measurement point collaboration, and realizes accurate ranging unaffected by clock error, significantly improving ranging accuracy; at the same time, it simplifies the deployment requirements, requiring only a single measurement point on the branch to achieve accurate identification and positioning of fault sections, combining high economy and deployment flexibility.

[0013] Based on the above scheme, the method for locating double-ended traveling wave faults in a multi-control power grid according to the present invention can be further improved as follows.

[0014] In one alternative approach, the topology information includes the connection relationships and physical distances between the trunk line, primary branches, and secondary branches.

[0015] Among the above-mentioned optional methods, by explicitly defining the topology information to include the connection relationship and physical distance between the trunk and branch lines, the fault zone judgment is supported by structured data, avoiding misjudgment caused by topological ambiguity during the positioning process, and improving the positioning reliability in complex multi-branch networks.

[0016] In one alternative approach, the step of calculating the line mode wave velocity based on line parameters includes: calculating the line mode wave velocity using the inductance and capacitance parameters of the corresponding line.

[0017] Among the above-mentioned optional methods, the line mode wave velocity is directly calculated through the line inductance and capacitance parameters, avoiding the errors caused by the reliance on fixed wave velocity values ​​in traditional methods, enhancing the algorithm's adaptability to line parameter fluctuations, and providing an accurate wave velocity reference for subsequent time difference compensation.

[0018] In one alternative approach, the modulus decoupling process includes converting the three-phase voltage signal into line-mode and zero-mode components via a Kelenberger transform.

[0019] In the above-mentioned optional methods, the three-phase voltage signals are rapidly decoupled through standardized Kelenberger transformation, electromagnetic coupling interference is eliminated, and the extracted line mode / zero mode components are kept pure and reliable, providing a high-quality signal source for wavefront detection.

[0020] In one alternative approach, the step of selecting the line-mode component with the most significant global jump as a unified analysis benchmark, based on the voltage change amplitude of the line-mode components at each measuring point, includes:

[0021] Calculate the voltage variation amplitude of each line modulus component at each measuring point; the voltage variation amplitude is obtained by differential calculation to obtain the maximum difference value;

[0022] By comparing the maximum difference values ​​of all measurement points globally, the linear mode component with the strongest abrupt change is selected as the unified analysis benchmark.

[0023] Among the above-mentioned optional methods, a global comparison is performed based on the maximum difference value of the line modulus components of each measuring point, and the optimal analysis benchmark is dynamically selected. This avoids misselection caused by weak signals or noise interference at a single measuring point, and significantly improves the robustness of wavefront feature recognition.

[0024] In one alternative approach, the steps of performing multi-scale wavelet decomposition and abrupt change detection on the selected line mode components, and extracting the arrival time of the line mode wavefront and the arrival time of the zero mode wavefront at each measurement point, include:

[0025] Wavelet decomposition is performed on the selected linear mode components to extract high-frequency detail components, and differential processing is performed on the high-frequency detail components to generate abrupt change rate curves.

[0026] Based on the abrupt change rate curve, the wavefront position is detected by combining a preset time window and threshold conditions, and the arrival time of the line mode wavefront and the arrival time of the zero mode wavefront are recorded respectively.

[0027] Among the above-mentioned optional methods, the mutation rate curve is generated by differential processing of high-frequency detail components. Combined with the dual criteria of time window and threshold, the first fault wavefront is accurately captured and power frequency jitter is shielded, which solves the pain point of traditional methods being susceptible to noise interference.

[0028] In one alternative approach, the steps for compensation and correction include:

[0029] A time difference elimination term is constructed based on the arrival time of the line mode wavefront and the arrival time of the zero mode wavefront to eliminate clock synchronization errors;

[0030] The least squares method is used to fit the functional relationship between the fault distance difference and the zero-mode time difference. The corrected zero-mode wave velocity is calculated based on the time difference elimination term. The linear mode wave velocity and the corrected zero-mode wave velocity are used to perform double-ended traveling wave ranging and output the fault point distance.

[0031] In one alternative approach, the steps for calculating the corrected zero-mode wave velocity include:

[0032] The least squares method was used to fit a linear function relationship between the fault distance difference and the zero-mode time difference, and the coefficients of the fitted function relationship were directly used as the corrected zero-mode wave velocity.

[0033] Among the aforementioned optional methods, a time difference elimination term is constructed by using the arrival time difference between the linear mode and the zero-mode wavefront, fundamentally offsetting the impact of clock synchronization error on ranging. Simultaneously, the least squares method is used to directly fit the linear functional relationship between the fault distance difference and the zero-mode time difference, using the fitting coefficients as the corrected zero-mode wave velocity. This avoids the accuracy loss caused by zero-mode wave velocity attenuation and significantly simplifies the calculation process of the correction parameters. This scheme can obtain high-precision corrected wave velocity through a single linear fitting, avoiding the complex multi-terminal weighted calculations in traditional methods, and significantly improving ranging stability and anti-interference capability under complex branch power grids.

[0034] In one alternative approach, the step of determining the fault region based on topology information, wavefront arrival time table, linear mode wave velocity, and corrected zero-mode wave velocity includes:

[0035] The arrival times of the line mode wavefront and the zero mode wavefront at the first and last measuring points are obtained from the wavefront arrival time table. Combined with the line mode wave velocity and the corrected zero mode wave velocity, double-end traveling wave ranging is performed on the first and last measuring points of the distribution network to obtain the fault distance.

[0036] Based on the fault distance, combined with the physical distance between nodes and the length of the trunk line in the topology information, it is determined whether the fault is located on the trunk line.

[0037] If the fault is located on the main line, then the faulty section is determined to be on the main line.

[0038] If the fault is not on the main line, the fault location can be determined by combining the wavefront arrival time table and the timing of the measurement points at the end of each branch. This will help pinpoint the primary or secondary branch where the fault is located and determine the fault range.

[0039] Among the above-mentioned optional methods, by combining the ranging results at the beginning and end with the wavefront timing table, the fault interval of the main road / branch can be determined. Only a single measuring point is needed to locate the branch fault location, which greatly reduces the complexity of the point layout and balances economy and positioning efficiency.

[0040] Secondly, this invention provides a dual-end traveling wave fault location system for a multi-control power grid, the technical solution of which is as follows:

[0041] The multi-control power grid dual-end traveling wave fault location system includes: a topology acquisition and wave velocity calculation module, a signal processing and feature extraction module, a time difference compensation and wave velocity correction module, and a fault interval location and output module.

[0042] The topology acquisition and wave velocity calculation module is used to: deploy traveling wave detection terminals at the beginning, end and branch terminals of the radial multi-branch power grid, collect topology information including the connection relationship and physical distance between the main line and the branch line, and calculate the line mode wave velocity based on the inductance and capacitance parameters of the line.

[0043] The signal processing and feature extraction module is used to: collect three-phase voltage and current data after a fault at each measuring point; decouple the three-phase voltage signal by mode through Kelvin transform; extract the line-mode component and zero-mode component of the initial voltage traveling wave at the fault point; calculate the maximum difference value of the line-mode component at each measuring point; select the line-mode component with the most significant jump as a unified analysis benchmark through global comparison; perform wavelet decomposition on the line-mode component; extract high-frequency detail components and generate a sudden change rate curve; detect the wavefront position by combining a preset time window and threshold conditions; record the arrival time of the line-mode wavefront and the arrival time of the zero-mode wavefront at each measuring point; and generate a wavefront arrival time sequence table.

[0044] The time difference compensation and wave velocity correction module is used to: construct a time difference elimination term based on the arrival time of the line mode and the zero mode wavefront to eliminate the clock synchronization error at both ends; use the least squares method to fit a linear function relationship between the fault distance difference and the zero mode time difference, and directly use the coefficients of the fitted function relationship as the corrected zero mode wave velocity; and use the line mode wave velocity and the corrected zero mode wave velocity to perform dual-end traveling wave ranging and output the fault point distance.

[0045] The fault range location and output module is used to: obtain the arrival times of the line mode wavefront and the zero mode wavefront at the first and last measuring points according to the wavefront arrival time table, and combine the line mode wave velocity and the corrected zero mode wave velocity to perform double-end traveling wave ranging on the first and last measuring points of the distribution network to obtain the fault distance; based on the fault distance, combined with the node physical distance and the length of the main line in the topology information, determine whether the fault is located on the main line; if the fault is located on the main line, determine that the fault range is on the main line; if the fault is not on the main line, combine the wavefront arrival time table and the time sequence of the measuring points at the end of each branch to locate the first-level or second-level branch line where the fault is located and determine the fault range.

[0046] The beneficial effects of the multi-control power grid dual-end traveling wave fault location system of the present invention are as follows:

[0047] The system of this invention effectively overcomes the positioning error caused by clock asynchrony and parameter disturbance through multi-measurement point collaborative wavefront timing analysis and time difference compensation mechanism, realizing accurate ranging unaffected by clock error and significantly improving ranging accuracy; at the same time, it simplifies the deployment requirements, requiring only a single measurement point on the branch to achieve accurate identification and positioning of fault sections, combining high economy and deployment flexibility.

[0048] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0049] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0050] Figure 1 This is a flowchart illustrating a method for locating a two-terminal traveling wave fault in a multi-control power grid according to the present invention.

[0051] Figure 2 A detailed flowchart illustrating the overall fault location process;

[0052] Figure 3This is a simulation topology diagram of a 35kV distribution network fault.

[0053] Figure 4 A schematic diagram of the wave head positioning process;

[0054] Figure 5 A comparison chart of linear errors under different clock errors;

[0055] Figure 6 This is a schematic diagram illustrating the relationship between fault distance difference and zero-mode time difference;

[0056] Figure 7 To correct the schematic diagram of the D-type traveling wave ranging process that is not affected by time difference;

[0057] Figure 8 This is a simulation model diagram for PSCAD / EMTDC;

[0058] Figure 9 This is a schematic diagram of the structure of a multi-control power grid dual-end traveling wave fault location system according to the present invention. Detailed Implementation

[0059] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0060] Figure 1 This diagram illustrates a flowchart of an embodiment of a two-terminal traveling wave fault location method for a multi-control power grid provided by the present invention. Figure 1 As shown, it includes the following steps:

[0061] S1. Traveling wave detection terminals are installed at the beginning, end, and branch terminals of the radial multi-branch power grid to collect topology information of the main line and branch lines, and the line mode wave velocity is calculated based on the line parameters. In S1:

[0062] 1) Traveling wave detection terminal refers to a signal acquisition device deployed at the beginning, end and branch terminals of the power distribution network to monitor line voltage and current data in real time.

[0063] 2) The topology information includes the physical connection relationship of the main line, primary branch line, and secondary branch line, as well as the physical distance of each line segment, providing a structural basis for the division of fault sections.

[0064] Specifically, in a radial multi-branch power grid, traveling wave detection terminals are installed at the head end (power supply side), tail end (load side), and all branch terminals; the line topology (such as node connection relationship and branch length) is collected synchronously, and the line mode wave velocity reflecting the propagation speed of high-frequency traveling waves is calculated based on the inductance and capacitance parameters of the line.

[0065] S2. Collect three-phase voltage and current data after the fault using various measuring points, and perform modulus decoupling processing on the three-phase voltage data to extract the line-mode and zero-mode components of the initial voltage traveling wave at the fault point. In S2:

[0066] 1) Modulus decoupling refers to converting electromagnetically coupled three-phase voltage signals into independent modal components to eliminate inter-phase interference.

[0067] 2) Among the line-mode component and the zero-mode component, the line-mode component represents the phase-to-phase traveling wave characteristics, while the zero-mode component represents the ground fault traveling wave characteristics, together forming the basis of fault analysis.

[0068] Specifically, after a fault occurs, three-phase voltage and current data are collected synchronously at each measuring point; the three-phase voltage signal is decoupled into uncoupled line-mode and zero-mode components through Kelenberger transformation, and the initial voltage traveling wave mode signal generated at the fault point is extracted.

[0069] S3. Based on the voltage change amplitude of the line-mode components at each measuring point, select the line-mode component with the most significant global jump as the unified analysis benchmark. In S3:

[0070] 1) The voltage change amplitude is obtained by differential calculation of the instantaneous jump intensity of the line mode component, which reflects the sudden change characteristics of the fault traveling wave.

[0071] 2) The most significant global jump refers to selecting the linear mode component with the largest overall jump amplitude as the unified analysis object by comprehensively considering the abrupt change intensity of all measurement points.

[0072] Specifically, the maximum difference value (instantaneous jump peak value) of the three linear mode components at each measuring point is calculated; the calculation results of all measuring points are summarized, and the linear mode component with the strongest global cumulative jump amplitude is selected as the unified signal reference for subsequent wavefront detection.

[0073] S4. Perform multi-scale wavelet decomposition and abrupt change detection on the selected linear mode components, extract the arrival time of the linear mode wavefront and the arrival time of the zero mode wavefront at each measurement point, and generate a wavefront arrival time series table. In S4:

[0074] 1) Multiscale wavelet decomposition refers to decomposing a signal into components of different frequencies through wavelet transform, focusing on high-frequency abrupt changes.

[0075] 2) The rate of change curve is generated by differential processing of high-frequency detail components to produce a curve that reflects the rate of signal change, which is used to accurately locate the wavefront.

[0076] Specifically, wavelet decomposition is performed on the selected line mode components to extract the first-level high-frequency detail components (d1); differential processing is performed on the d1 components to generate a sudden change rate curve, and the wavefront is located by combining a preset time window and amplitude threshold; the precise arrival time of the line mode wavefront and the zero mode wavefront at each measurement point is recorded respectively, and the wavefront arrival time sequence table is generated by arranging them in chronological order.

[0077] S5. Based on topology information, wavefront arrival time table, linear mode wave velocity, and corrected zero-mode wave velocity, determine the fault region. During the fault region determination process and final location, compensate and correct using the time difference between the linear mode wavefront arrival time and the zero-mode wavefront arrival time, and output the fault location. In S5:

[0078] 1) The time difference elimination term refers to the use of the difference in arrival time between the linear mode and the zero mode wavefront to offset the fixed error introduced by the asynchronous clocks of the two devices.

[0079] 2) The fault section refers to the section of the line where the fault point is located, such as the main line, primary branch line or secondary branch line.

[0080] Specifically, when performing double-end traveling wave ranging at the beginning and end measuring points, a time difference elimination term is constructed using the time difference between the line mode and the zero mode wavefronts, and the corrected zero mode wave velocity is directly used for compensation correction to output the fault point distance. Then, by comparing this distance with the node distances in the topology information, it is determined whether the fault is located on the main line. If the fault is on the main line, the corrected ranging result is directly used as the final location; if the fault is on a branch line, the specific branch is located according to the wavefront timing table, and the corrected ranging method is reused within that branch interval to output the precise location of the fault point.

[0081] Figure 2 The overall fault location process is shown, which includes six core steps: First, deploy detection terminals at key nodes of the distribution network and collect topology information; after the fault occurs, perform signal decoupling and modulus analysis; generate a timing table through wavefront detection; and finally achieve accurate location by combining topology and corrected ranging methods.

[0082] The technical solution of this embodiment effectively overcomes the positioning error caused by clock asynchrony and parameter disturbance through wavefront timing analysis and time difference compensation mechanism with multi-measurement point collaboration, realizes accurate ranging unaffected by clock error, and significantly improves ranging accuracy; at the same time, it simplifies the deployment requirements, requiring only a single measurement point on the branch to achieve accurate identification and positioning of the fault section, combining high economy and deployment flexibility.

[0083] In one alternative approach, the topology information includes the connection relationships and physical distances between the trunk line, primary branches, and secondary branches.

[0084] It should be noted that, as Figure 3As shown, the straight line from the first measuring point ① to the last measuring point ⑥ is the main line. Branches such as ② and ③ that branch directly from the main line are first-level branches. Branches such as ④, ⑤, ⑦, ⑧, and ⑨ that branch further from the first-level branches are second-level branches.

[0085] In one alternative approach, the step of calculating the line mode wave velocity based on line parameters includes: calculating the line mode wave velocity using the inductance and capacitance parameters of the corresponding line.

[0086] In this embodiment, by deploying corresponding traveling wave detection terminals at the beginning and end of the 35kV distribution system with a typical radial multi-branch transmission line structure and at the end of each branch, the topology information of the main line and each primary and secondary branch is stored. At the same time, the line mode wave velocity of the line is obtained through the line parameters. The calculation formula is as follows:

[0087] (1)

[0088] in, The wave velocity of the traveling wave linear mode component; , These are the zero-mode inductance and capacitance of the circuit, respectively.

[0089] In one alternative approach, the modulus decoupling process includes converting the three-phase voltage signal into line-mode and zero-mode components via a Kelenberger transform.

[0090] In this embodiment, after a fault occurs, traveling wave detection terminals deployed at various measuring points in the distribution network synchronously acquire three-phase voltage signals. The three-phase voltage signals are denoted as... , , , representing the instantaneous voltage values ​​of phase A, phase B, and phase C, respectively.

[0091] Because of electromagnetic coupling between the three-phase conductors, traveling waves are generated on the non-faulty phases, and the equations describing these waves are not independent of each phase. To eliminate the coupling effect and effectively extract fault characteristics, the acquired three-phase voltage signals need to be decoupled and converted into independent modulus components.

[0092] The Karenbauer transform is used to convert the phase-domain system (i.e., the ABC three-phase system) into an uncoupled mode-domain system. The mathematical expression of the transform is as follows:

[0093] (2)

[0094] in, , , These are the instantaneous values ​​of the three-phase voltages before decoupling (unit: volts V); It is the zero-mode voltage component (unit: volts V) obtained after transformation, which represents the common-mode part of the three-phase voltage; It is one of the line-mode voltage components obtained after transformation (unit: volts V), which represents the differential mode part flowing between phase A conductor and phase B conductor; It is one of the line-mode voltage components obtained after transformation (unit: volts V), which represents the differential mode portion flowing between the A-phase conductor and the C-phase conductor.

[0095] To facilitate modulus analysis and form a complete three-phase circuit description, a third linear modulus component is introduced. This component represents the differential mode portion flowing between the B-phase and C-phase conductors. and , Together they constitute a complete set describing the differential-mode voltage between three phases.

[0096] After the above Kelenberg transformation, the original three-phase coupled voltage signal was successfully transformed. , , Decoupling to zero-mode voltage component and line mode voltage components , and introduce These decoupled mode components, especially the linear and zero-mode components of the initial voltage traveling wave at the fault point, will serve as the basis signals for subsequent wavefront detection and fault location analysis.

[0097] In one alternative approach, the step of selecting the line-mode component with the most significant global jump as a unified analysis benchmark, based on the voltage change amplitude of the line-mode components at each measuring point, includes:

[0098] Calculate the voltage variation amplitude of each line modulus component at each measuring point; the voltage variation amplitude is obtained by differential calculation to obtain the maximum difference value;

[0099] By comparing the maximum difference values ​​of all measurement points globally, the linear mode component with the strongest abrupt change is selected as the unified analysis benchmark.

[0100] In this embodiment, the linear mode component includes those obtained through mode decoupling processing (such as the aforementioned Kalenberger transform). , The components, and the flow between phase B and phase C conductors introduced based on the three-phase circuit integrity derivation. Components. Therefore, the line-mode voltage signal available for analysis at each measuring point is: , , Three types.

[0101] In calculating the voltage change amplitude at a single measurement point, for each traveling wave detection terminal measurement point, the three line mode voltage components are calculated separately. , , The voltage variation amplitudes of each signal are calculated. Specifically, the maximum difference value of the line-mode voltage signal sequence is calculated to identify the most significant transitions in the signal.

[0102] All measurement points are subjected to the same linear mode component (e.g., all measurement points). The maximum difference values ​​calculated are then compared globally. Specifically, the strategy is to sum the maximum difference values ​​calculated for a particular linear modulus component from all measuring points (i.e., to obtain the total global variation of that linear modulus component). The three linear modulus components are calculated separately. , , The corresponding global total change magnitude.

[0103] The total global variation amplitudes of the three line mode components are compared, and the line mode component with the largest total variation amplitude is selected. This selected line mode component will serve as the unified signal input and unified analysis benchmark for wavefront detection and analysis at all subsequent measurement points. In addition, based on the global comparison strategy, fault type information can be used as an auxiliary criterion, but the core selection criterion is based on the cumulative sum of the global maximum difference values.

[0104] In one alternative approach, the steps of performing multi-scale wavelet decomposition and abrupt change detection on the selected line mode components, and extracting the arrival time of the line mode wavefront and the arrival time of the zero mode wavefront at each measurement point, include:

[0105] Wavelet decomposition is performed on the selected linear mode components to extract high-frequency detail components, and differential processing is performed on the high-frequency detail components to generate abrupt change rate curves.

[0106] Based on the abrupt change rate curve, the wavefront position is detected by combining a preset time window and threshold conditions, and the arrival time of the line mode wavefront and the arrival time of the zero mode wavefront are recorded respectively.

[0107] In this embodiment, the selected optimal line-mode voltage (i.e., the line-mode component with the most significant global jump) is used as the unified signal input. A multi-scale wavelet transform method is employed to decompose the selected line-mode voltage signal at each measurement point. Specifically, the Daubechies wavelet db4 is used as the wavelet basis function to perform a 5-level decomposition of the signal, extracting the first-level detail component, denoted as d1 (high-frequency detail component). This d1 component effectively reflects the abrupt changes in the high-frequency part of the signal and is the main carrier of the fault traveling wavefront.

[0108] Differential processing is performed on the d1 high-frequency detail component extracted at each measurement point. Differential processing calculates the amplitude change between adjacent sampling points, obtaining the abrupt change rate curve of the d1 component. This curve characterizes the rate of change of the signal amplitude over time, and the wavefront position typically corresponds to a significant abrupt change point on the curve.

[0109] Based on the generated rate of change curve (i.e., the differentially processed d1 signal, denoted as...) , where k represents the k-th sampling point), and the wavefront position is detected by combining a preset time window and threshold conditions. The specific detection process is as follows: Figure 4 As shown:

[0110] The preset sampling frequency is 10MHz. In the differential results... In the middle, skip a preset initial time window. (Value: 0.01ms). This window is used to shield the steady-state power frequency component in the transition signal before and after the fault, as well as the slight jitter generated at the beginning of sampling, to avoid the slight disturbance in the steady-state region or at the moment of sampling being misjudged as a wavefront, thereby improving the reliability of detection.

[0111] Skip In the data behind the window, search for points that meet the following conditions as mutation starting points:

[0112] (3)

[0113] in, This represents the signal amplitude at the k-th sampling point (from the d1 component after differential processing). This is a threshold coefficient, ranging from 5 to 10 times. This value is an empirical value determined through statistical analysis of multiple simulations and noise conditions. express The global average amplitude of the signal (usually excluding the shielded area in the calculation). (Data within).

[0114] For determining the location of candidate wavefronts, the time window following the found mutation initiation point is... Within a range of 0.02ms, the point with the largest amplitude is searched and selected as the candidate wavefront position (denoted as the kth point).

[0115] For wavefront position verification, the following two criteria are applied to verify the found candidate wavefront position k:

[0116] First criterion: The wavefront is a local maximum point within a certain range, that is, the candidate point k must be a local peak, satisfying:

[0117] (4)

[0118] in, These represent the signal amplitudes of the left and right adjacent sampling points of candidate point k, respectively. The time interval between two adjacent sampling points (at a sampling rate of 10MHz). ).

[0119] The second criterion is that the abrupt change rate of the wavefront is much greater than the average abrupt change rate. That is, the time derivative of the equivalent discrete signal is used to measure the rate of change of the signal at point n. The larger the abrupt change rate, the more obvious the wavefront characteristics. As shown in equation (5), when the following conditions are met... When the time is right, the k-th point is determined as the wave head position, where Let $k$ be the global average mutation rate. Therefore, the mutation rate of candidate site $k$ must be significantly higher than the global average, satisfying the following condition:

[0120] (5)

[0121] in, This represents the mutation rate of candidate point k (equivalent to the time derivative of a discrete signal). for The global average mutation rate of the signal is calculated as follows: = ; This represents the signal amplitude at the nth sampling point;

[0122] If candidate point k satisfies both the first and second criteria, then point k is determined to be the position of the line-mode voltage wavefront. If neither criterion is satisfied, the time window is expanded. Then, the candidate wavefront position determination and wavefront position criterion verification are re-executed, that is, the candidate point with the largest amplitude is searched again in a larger window and the criterion verification is performed until a wavefront position that meets the conditions is found.

[0123] Regarding wavefront time recording, for each measurement point: record the arrival time corresponding to the finally determined line-mode voltage wavefront. Similarly, for the zero-mode voltage component signal, repeat wavelet decomposition and high-frequency detail component extraction to wavefront position detection, that is, perform wavelet decomposition to extract d1, differential, and wavefront detection, and record the arrival time corresponding to the zero-mode voltage wavefront at that measurement point.

[0124] During the generation of the wavefront timing table, the arrival times of the line mode wavefront and the zero mode wavefront recorded at all n measurement points are arranged in an ordered manner (usually sorted from earliest to latest time) to generate an accurate wavefront arrival timing table. This timing table provides a reliable and accurate time-stamped basis for subsequent fault zone determination and D-type ranging unaffected by time difference.

[0125] In one alternative approach, the steps for compensation and correction include:

[0126] A time difference elimination term is constructed based on the arrival time of the line mode wavefront and the arrival time of the zero mode wavefront to eliminate clock synchronization errors;

[0127] The least squares method is used to fit the functional relationship between the fault distance difference and the zero-mode time difference. The corrected zero-mode wave velocity is calculated based on the time difference elimination term. The linear mode wave velocity and the corrected zero-mode wave velocity are used to perform double-ended traveling wave ranging and output the fault point distance.

[0128] In one alternative approach, the steps for calculating the corrected zero-mode wave velocity include:

[0129] The least squares method was used to fit a linear function relationship between the fault distance difference and the zero-mode time difference, and the coefficients of the fitted function relationship were directly used as the corrected zero-mode wave velocity.

[0130] It should be noted that after extracting and sorting the wavefront times at each measurement point, the fixed error caused by clock asynchrony at both ends is reduced by utilizing the arrival time difference between the linear and zero-mode wavefronts. Furthermore, the traditional D-type ranging formula is improved by introducing the linear-zero-mode time difference, eliminating its influence from the calculation formula. This ensures high accuracy even with insufficient time synchronization or communication delays. Considering the frequency-dependent attenuation of the zero-mode wave velocity during long-distance propagation, a polynomial fitting based on least squares is introduced to compensate for and correct the zero-mode wave velocity, further improving ranging stability. Through simulation analysis of the relationship between the fault distance difference and the zero-mode time difference, the zero-mode time difference from the fault point to both measurement points is calculated for different fault distance differences. The equivalent average zero-mode wave velocity is then calculated as the corrected zero-mode wave velocity. This corrected zero-mode wave velocity is substituted into the D-type traveling wave ranging, which is unaffected by time differences, to achieve precise fault location.

[0131] Distance from the fault point to the left end As a calculation benchmark, the improved D-type traveling wave ranging method, which is unaffected by time difference, is as follows:

[0132] Traditional D-type traveling wave ranging is as follows:

[0133] (6)

[0134] However, due to issues such as clock drift or communication delays, introducing a fixed clock error will significantly reduce the accuracy of ranging. The result after adding clock error is as follows:

[0135] (7)

[0136] in, Indicates the length of the lines at the left and right ends; For linear mode wave velocity; , These are the arrival times for the left and right ends, respectively; Clock error (generally not exceeding) ); This is to assume that the distance between the fault point and the left and right ends of the line is not considered under time difference conditions.

[0137] When the synchronization error of the time synchronization devices at both ends is too large, it will cause a significant error in the fault location accuracy. However, the clock synchronization error can be eliminated by using the arrival time difference between the linear mode and the zero mode wavefront. Therefore, it is possible to improve the traditional D-type traveling wave ranging by utilizing the zero-mode time difference, making it unaffected by time difference.

[0138] In this embodiment, a time difference elimination term is constructed to eliminate clock synchronization errors: the arrival time of the linear mode wavefront and the arrival time of the zero mode wavefront are used to construct a time difference elimination term to eliminate the fixed error (clock error) caused by the clock asynchrony of the two end devices. ). Specifically:

[0139] The core principle of constructing the time difference elimination term lies in the clock error term in the linear mode time difference and the zero mode time difference. They can cancel each other out, as shown in equation (8). Therefore, by utilizing the linear zero-mode time difference, the traditional D-type traveling wave ranging can be improved into a D-type traveling wave ranging that is not affected by the time difference, as shown in equation (9):

[0140] (8)

[0141] (9)

[0142] in, , It is the time when the linear mode traveling wave arrives at the left and right measurement endpoints of the fault section, respectively (obtained from the wavefront arrival timing table). , These are the times when the zero-mode traveling wave arrives at the left and right measurement endpoints of the fault section, respectively (obtained from the wavefront arrival timing table).

[0143] in, This represents the distance from the fault point to the left measurement endpoint (a quantity to be determined). This indicates the total length of the line between the left and right measurement endpoints; The line mode wave velocity is known and calculated from the line parameters. This is the zero-mode wave velocity (a basic value calculated from line parameters, but requiring subsequent correction). This is the calculation result of the time difference elimination term, which no longer includes clock error. .

[0144] Since the zero-mode wave velocity gradually decreases with increasing fault distance, the error caused by this decrease becomes more pronounced with increasing distance. Moreover, when the distances at the left and right ends of the line differ too much, it will also have an adverse effect on the ranging results. Therefore, equation (9) is rewritten as follows:

[0145] (10)

[0146] It should be noted that the basis for the rewriting is: due to the zero-mode wave velocity In long-distance propagation, a frequency-dependent attenuation effect exists, and directly using a time difference elimination term to eliminate it will still lead to ranging errors. Specifically, when the fault point is close to the left measurement endpoint, due to the attenuation effect of the zero-mode wave velocity, the calculated actual value of the zero-mode wave velocity will be greater than its true value. Furthermore, this error is more significant because the attenuation levels at the left and right ends are different, and vice versa. This is specifically reflected in the formula as follows: because Using an excessively large value relative to the actual value causes... In practice, this will lead to increased errors. Furthermore, it needs to be clarified that the main function of formula (10) is theoretical analysis, revealing how the attenuation effect leads to errors even when clock errors are eliminated when using uncorrected zero-mode wave velocity. This results in errors. Formula (10) itself was not designed for the final calculation.

[0147] Given the strong monotonicity between the fault distance difference and the zero-mode time difference at the left and right ends, the least squares method can be used to fit their correspondence. Specifically, the relationship between the fault distance difference and the zero-mode time difference is fitted through simulation analysis. The zero-mode time difference from the fault point to the measurement points at both ends is calculated for different fault distance differences. The calculated fitting coefficients, used as a correction for the zero-mode wave velocity, have better accuracy than ignoring the attenuation of the zero-mode wave velocity, replacing the original actual zero-mode wave velocity. This yields a corrected zero-mode wave velocity with a moderate attenuation relative to the left and right ends, without disrupting the principle of distance measurement using linear zero-mode time difference. By using a moderately attenuated zero-mode wave velocity to replace the original calculated wave velocity, the clock error is completely eliminated from the formula, thereby improving the accuracy of distance measurement and obtaining a more precise fault distance. Simultaneously, the linear error is defined as the relative error in the linear part of the distance measurement formula after introducing the clock error. Compared with traditional methods, [the following is a comparison]. Figure 5 As shown.

[0148] In this embodiment, after completing the construction of the D-type traveling wave ranging formula (Equation 9) which is unaffected by time difference, considering the zero-mode wave velocity... During long-distance propagation, it exhibits frequency-dependent attenuation characteristics, and there is a good monotonic relationship between its propagation time difference and that of the left and right ends. Therefore, zero-mode wave velocity needs to be compensated and corrected to improve ranging stability and accuracy. The compensation and correction process specifically includes the following three steps:

[0149] Linear fitting of the relationship between fault distance difference and zero-mode time difference: The least squares method was used to fit the fault distance difference. With zero-mode time difference Performing a linear fit yields the following relationship:

[0150] (11)

[0151] in, The difference in fault distance between the left and right ends. , and These are the distances from the fault point to the left and right ends, respectively. The time difference between the arrival of the zero-mode wavefronts on the left and right ends. = ; These are the fitting coefficients, i.e., the functional relationship coefficients (which will later be used as the corrected zero-mode wave velocity). ).

[0152] It should be noted that a model was built using actual line parameters, sampling points were set at both ends of the line, and fault points were set every 1 km to collect zero-mode wavefront time data at different fault distances. For example... Figure 6 As shown, the experiment verifies that the coefficient of determination for the first-order fit is greater than 0.99, proving the rationality of directly using the fitting coefficient as the correction for the zero-mode wave velocity, and when hour; Therefore, the fitted equation does not contain a constant term.

[0153] Determine the corrected zero-mode wave velocity : The fitting coefficients As a correction zero-mode wave velocity Satisfying the relation:

[0154] (12)

[0155] The corrected wave velocity has the following characteristics: It is a fixed value and does not change with clock error. It changes with the change; it completely eliminates clock errors from the ranging formula; it represents the equivalent wave velocity with the relative attenuation at the left and right ends being moderate.

[0156] The corrected zero-mode wave velocity Substitute the D-type traveling wave ranging formula (9), which is unaffected by time difference, into the original formula. The distance to the fault point was calculated. The final distance measurement formula is as follows:

[0157] (13)

[0158] Overall, the distance measurement correction process is as follows: Figure 7 As shown: First, the linear relationship between the fault distance difference and the zero-mode time difference is fitted offline; second, the fitting coefficient is directly used as the correction wave velocity; finally, the D-type ranging formula that resists the influence of time difference is substituted into the output positioning result.

[0159] In one alternative approach, the step of determining the fault region based on topology information, wavefront arrival time table, linear mode wave velocity, and corrected zero-mode wave velocity includes:

[0160] The arrival times of the line mode wavefront and the zero mode wavefront at the first and last measuring points are obtained from the wavefront arrival time table. Combined with the line mode wave velocity and the corrected zero mode wave velocity, double-end traveling wave ranging is performed on the first and last measuring points of the distribution network to obtain the fault distance.

[0161] Based on the fault distance, combined with the physical distance between nodes and the length of the trunk line in the topology information, it is determined whether the fault is located on the trunk line.

[0162] If the fault is located on the main line, then the faulty section is determined to be on the main line.

[0163] If the fault is not on the main line, the fault location can be determined by combining the wavefront arrival time table and the timing of the measurement points at the end of each branch. This will help pinpoint the primary or secondary branch where the fault is located and determine the fault range.

[0164] It should be noted that when determining fault zones based on the acquired line topology, physical distances to measurement points, wave velocity information, and the generated wavefront arrival time table, an error margin needs to be introduced during the fault selection process due to factors such as line parameter deviations and data acquisition unit measurement errors in the actual distribution network environment. (Value is 0.01). This margin setting is based on the fact that the current traveling wave ranging algorithm generally has a positioning relative error of less than 1% under typical operating conditions. Through multi-condition simulation and field test verification, this margin can avoid the decrease in accuracy caused by the expansion of the positioning interval (margin too large) and prevent missed detection caused by slight jitter in wavefront extraction (margin too small), thus taking into account both engineering feasibility and the accuracy range of mainstream algorithms. Therefore, this scheme uses D-type double-ended traveling wave ranging, which is not affected by time difference, to locate the fault in two stages: first, ranging is performed on the measuring points at the beginning and end of the distribution network to determine whether the fault is located in the main line; if it is not a main line fault, the first-level or second-level branch interval is located according to the wavefront timing sequence of the measuring points at the end of each branch.

[0165] In this embodiment, the measurement points at the beginning of the distribution network (e.g., [missing information]) are obtained according to the wavefront arrival time table. Figure 3 Measurement point 1) and tail end measurement point (such as Figure 3 The arrival time of the linear mode wavefront and the arrival time of the zero mode wavefront at measurement point 6) are measured. The fault distance is calculated using the compensated and corrected D-type traveling wave ranging formula that is not affected by time difference, i.e., formula (13). This is the calculated fault distance. .

[0166] Combining the physical distance between nodes and the length of the backbone line in the topology information Calculate the trunk coefficient :

[0167] (14)

[0168] in, This refers to the fault distance measured on the main line. To and The nearest node physical distance; The length of the main trunk line.

[0169] like ≥ If the fault is located on the main line, then the fault section is determined to be on the main line; at this point, the fault location is the specific fault location on the main line. .

[0170] like < If the fault is not located on the main line, the branch line location procedure is executed.

[0171] When the fault is not located on the main line, the measuring point with the earliest wavefront arrival time is located according to the wavefront arrival time table (denoted as the minimum measuring point). Using the minimum measuring point as a reference, double-ended traveling wave ranging is performed sequentially with the remaining n measuring points on the branch to obtain the fault distance. ( ), and calculate the branch coefficient. ( ):

[0172] (15)

[0173] Where x is the length of the branch route (obtained from topology information).

[0174] At this point, the distance measurement result falls into two categories:

[0175] like < If all branch coefficients are less than the error margin, then the fault is determined to be located on that first-level branch. In this case, the smallest measuring point and the nearest measuring point other than the current first-level branch are selected for traveling wave ranging; the resulting distance is denoted as... Based on the line topology, the length of this first-level branch line is utilized. Subtract the measured fault distance, then add the line length of the secondary branch where the smallest measuring point is located. Obtain the specific fault location in the first-level branch. The announcement is as follows:

[0176] (16)

[0177] Otherwise, not satisfied hour:

[0178] like If the fault is located in the secondary branch where the minimum measuring point is located, then traveling wave ranging is performed using the minimum measuring point and the measuring point with the second shortest outgoing time. At this point, the distance from the minimum measuring point to the fault location is... (km):

[0179] (17)

[0180] If only one set of distance measurement results shows an anomaly, it is recorded as an abnormal measurement point. ,Right now All other results are satisfied. If the branch containing the abnormal measuring point is determined to be the faulty branch, then traveling wave ranging is performed using the minimum measuring point and the abnormal measuring point to obtain the result. At this time, the distance from the fault location to the abnormal measuring point is... (km):

[0181] (18)

[0182] in, The length of the line at the faulty branch measuring point.

[0183] To further illustrate the effect, experiments were conducted for verification. Furthermore, methods such as... Figure 8 The PSCAD-MATLAB software was used to co-simulate a 35kV radial multi-branch distribution line. Detection devices were installed at the beginning and end points of the distribution network and at the end nodes of each secondary branch line. For ease of verification, all lines were set to the same parameters, and each detection device simultaneously acquired fault traveling wave signals. A phase-A ground fault was set at a location 10km from the branch start point on the primary branch connected to node 4, 10km from the end of the line. The sampling frequency was 10MHz, the fault occurrence time was 0.02s, the fault duration was 0.04s, the transition resistance was 50Ω, and the initial phase angle of the fault was... Table 1 shows the wavefront time sequence at each measuring point:

[0184] Table 1 (Wavefront Time Sequence at Each Measurement Point):

[0185]

[0186] Table 1 provides the wavefront times at the beginning and end measuring points of the distribution network. Distance measurement determines the location of the fault on the main line. The main line criterion is then applied between the fault and the nearest node 4, satisfying the condition. Therefore, it can be determined that the fault is located on the first-level or second-level branch of node 4. Branch criteria are then used for further analysis. Three measuring points (numbered 7, 8, and 9) were set up on this branch road. The measuring point with the earliest wavefront (point 8) was used as the reference, and two sets of distance measurements were performed sequentially to meet the requirements. Therefore, it can be determined that the fault is located on the first-level branch of node 4.

[0187] According to Table 1, the distance is measured using the smallest measuring point (8) and the nearest measuring point (5) excluding this first-level branch. The line distance between measuring points 8 and 5 is obtained from the line topology. km. The relationship between zero-mode travel time and fault distance difference is analyzed through simulation in PSCAD. In MATLAB, the scatter plot obtained from the simulation is fitted using a first-order polynomial with the least squares method, thereby obtaining the relationship between zero-mode travel time and fault distance difference, and thus the corrected zero-mode wave velocity. Substitution (10):

[0188] ;

[0189] The distance measurement results show that the distance to the fault is 0.13 km different from the simulated fault distance, with a relative error of only 0.65%. The simulation results indicate that the measured fault location is very close to the set actual fault location, and the relative error is less than 1%.

[0190] Figure 9 A schematic diagram of an embodiment of a dual-end traveling wave fault location system 200 for a multi-control power grid provided by the present invention is shown. Figure 9 As shown, the system 200 includes: a topology acquisition and wave velocity calculation module 210, a signal processing and feature extraction module 220, a time difference compensation and wave velocity correction module 230, and a fault zone location and output module 240.

[0191] The topology acquisition and wave velocity calculation module 210 is used to: deploy traveling wave detection terminals at the beginning, end and branch terminals of the radial multi-dominated power grid, collect topology information including the connection relationship and physical distance between the main line and the branch line, and calculate the line mode wave velocity based on the inductance and capacitance parameters of the line.

[0192] The signal processing and feature extraction module 220 is used to: collect three-phase voltage and current data after the fault at each measuring point; perform modal decoupling on the three-phase voltage signal through Kelvin transform; extract the line-mode component and zero-mode component of the initial voltage traveling wave at the fault point; calculate the maximum difference value of the line-mode component at each measuring point; select the line-mode component with the most significant jump as a unified analysis benchmark through global comparison; perform wavelet decomposition on the line-mode component; extract high-frequency detail components and generate a sudden change rate curve; detect the wavefront position by combining a preset time window and threshold conditions; record the arrival time of the line-mode wavefront and the arrival time of the zero-mode wavefront at each measuring point; and generate a wavefront arrival time sequence table.

[0193] The time difference compensation and wave speed correction module 230 is used to: construct a time difference elimination term based on the arrival time of the line mode and the zero mode wavefront to eliminate the clock synchronization error at both ends; use the least squares method to fit a linear function relationship between the fault distance difference and the zero mode time difference, and directly use the fitted function relationship coefficients as the corrected zero mode wave speed; and use the line mode wave speed and the corrected zero mode wave speed to perform dual-end traveling wave ranging and output the fault point distance.

[0194] The fault range location and output module 240 is used to: obtain the arrival time of the line mode wavefront and the arrival time of the zero mode wavefront at the first and last measuring points according to the wavefront arrival time table, and combine the line mode wave velocity and the corrected zero mode wave velocity to perform double-end traveling wave ranging on the first and last measuring points of the distribution network to obtain the fault distance; based on the fault distance, combined with the node physical distance and the length of the main line in the topology information, determine whether the fault is located on the main line; if the fault is located on the main line, determine that the fault range is on the main line; if the fault is not on the main line, combine the wavefront arrival time table and the time sequence of the measuring points at the end of each branch to locate the first-level or second-level branch line where the fault is located and determine the fault range.

[0195] The technical solution of this embodiment effectively overcomes the positioning error caused by clock asynchrony and parameter disturbance through wavefront timing analysis and time difference compensation mechanism with multi-measurement point collaboration, realizes accurate ranging unaffected by clock error, and significantly improves ranging accuracy; at the same time, it simplifies the deployment requirements, requiring only a single measurement point on the branch to achieve accurate identification and positioning of the fault section, combining high economy and deployment flexibility.

[0196] The parameters and steps for implementing the corresponding functions of each module in the multi-control power grid double-end traveling wave fault location system 200 of this embodiment can be referred to the parameters and steps in the embodiments of the multi-control power grid double-end traveling wave fault location method above, and will not be repeated here.

[0197] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for locating a two-terminal traveling wave fault in a multi-control power grid, characterized in that, include: By setting traveling wave detection terminals at the beginning, end and branch terminals of the radial multi-branch power grid, the topology information of the main line and branch lines is collected, and the line mode wave velocity is calculated based on the line parameters. The three-phase voltage and current data after the fault were collected at each measuring point, and the three-phase voltage data were decoupled by modulus to extract the line mode component and zero mode component of the initial voltage traveling wave at the fault point. Based on the voltage change amplitude of the line mode components at each measuring point, the line mode component with the most significant global jump is selected as the unified analysis benchmark. Multi-scale wavelet decomposition and abrupt change detection are performed on the selected linear mode components to extract the arrival time of the linear mode wavefront and the arrival time of the zero mode wavefront at each measurement point, and a wavefront arrival time series table is generated. Based on topology information, wavefront arrival time table, line mode wave velocity and corrected zero mode wave velocity, the fault range is determined. During the process of determining the fault range and during the final location, the time difference between the line mode wavefront arrival time and the zero mode wavefront arrival time is used for compensation and correction, and the fault location is output. The steps of performing multi-scale wavelet decomposition and abrupt change detection on the selected line mode components, and extracting the arrival time of the line mode wavefront and the arrival time of the zero mode wavefront at each measurement point, include: Wavelet decomposition is performed on the selected linear mode components to extract high-frequency detail components, and differential processing is performed on the high-frequency detail components to generate abrupt change rate curves. Based on the abrupt change rate curve, the wavefront position is detected by combining the preset time window and threshold conditions, and the arrival time of the line mode wavefront and the arrival time of the zero mode wavefront are recorded respectively. The steps for compensation and correction include: A time difference elimination term is constructed based on the arrival time of the line mode wavefront and the arrival time of the zero mode wavefront to eliminate clock synchronization errors; The least squares method is used to fit the functional relationship between the fault distance difference and the zero-mode time difference. The corrected zero-mode wave velocity is calculated based on the time difference elimination term. The line-mode wave velocity and the corrected zero-mode wave velocity are used to perform double-ended traveling wave ranging and output the fault point distance. The steps for calculating the corrected zero-mode wave velocity include: The least squares method was used to fit a linear function relationship between the fault distance difference and the zero-mode time difference, and the coefficients of the fitted function relationship were directly used as the corrected zero-mode wave velocity.

2. The method for locating a two-terminal traveling wave fault in a multi-control power grid according to claim 1, characterized in that, The topology information includes the connection relationships and physical distances between the main lines, primary branches, and secondary branches.

3. The method for locating a two-terminal traveling wave fault in a multi-control power grid according to claim 1, characterized in that, The step of calculating the line mode wave velocity based on line parameters includes: calculating the line mode wave velocity using the inductance and capacitance parameters of the corresponding line.

4. The method for locating a two-terminal traveling wave fault in a multi-control power grid according to claim 1, characterized in that, The modulus decoupling process includes converting the three-phase voltage signal into line-mode and zero-mode components using a Kelenberger transform.

5. The method for locating a two-terminal traveling wave fault in a multi-control power grid according to claim 1, characterized in that, The steps for selecting the line-mode component with the most significant global jump as a unified analysis benchmark, based on the voltage change amplitude of the line-mode components at each measuring point, include: Calculate the voltage variation amplitude of each line modulus component at each measuring point; the voltage variation amplitude is obtained by differential calculation to obtain the maximum difference value; By comparing the maximum difference values ​​of all measurement points globally, the linear mode component with the strongest abrupt change is selected as the unified analysis benchmark.

6. The method for locating a two-terminal traveling wave fault in a multi-control power grid according to claim 1, characterized in that, The steps for determining the fault region based on topology information, wavefront arrival time table, linear mode wave velocity, and corrected zero-mode wave velocity include: The arrival times of the line mode wavefront and the zero mode wavefront at the first and last measuring points are obtained from the wavefront arrival time table. Combined with the line mode wave velocity and the corrected zero mode wave velocity, double-end traveling wave ranging is performed on the first and last measuring points of the distribution network to obtain the fault distance. Based on the fault distance, combined with the physical distance between nodes and the length of the trunk line in the topology information, it is determined whether the fault is located on the trunk line. If the fault is located on the main line, then the faulty section is determined to be on the main line. If the fault is not on the main line, the fault location can be determined by combining the wavefront arrival time table and the timing of the measurement points at the end of each branch. This will help pinpoint the primary or secondary branch where the fault is located and determine the fault range.

7. A dual-end traveling wave fault location system for a multi-control power grid, characterized in that, The method for locating double-ended traveling wave faults in a multi-control power grid as described in any one of claims 1-6 is adopted.

Images