Intelligent Location Method and System for Fault Sections in Distribution Network Feeders Based on Feature Recognition

By synchronously acquiring and differentially transforming current and voltage sequences, and combining them with the distribution network topology to construct joint trajectory points, the problem of misjudgment in fault segment location in existing technologies has been solved, achieving accurate fault segment location and high location efficiency.

CN123066483APending Publication Date: 2026-11-10NANJING HAOBO LANNENG POWER TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611270974.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-20
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing fault location technologies for distribution network feeders only utilize a single type of electrical quantity, such as current or voltage, and fail to effectively integrate voltage and current dual electrical quantities to construct phase space trajectory characteristics. This leads to frequent misjudgments and missed judgments in fault section identification, and makes it difficult to quickly locate faults at section boundaries, thus reducing fault location efficiency.

Method used

By synchronously collecting the current and voltage sequences of the feeder section, performing point-by-point differential transformation to generate the deviation sequence, and combining the distribution network topology to construct the joint phase space trajectory points of current and voltage, the outlier significance of the joint trajectory is calculated to determine the fault section.

Benefits of technology

It achieves precise location of faulty sections, improves the basic reliability and location efficiency of fault identification, reduces manual judgment and intervention, and shortens the calculation time.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123066483A_ABST
    Figure CN123066483A_ABST
Patent Text Reader

Abstract

This invention proposes a method and system for intelligent location of faulty feeder sections in distribution networks based on feature recognition, relating to the field of distribution network technology. The method includes: acquiring the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence of a feeder section in the distribution network; performing point-by-point differential transformation on the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence to obtain an intra-section current deviation sequence and an intra-section voltage deviation sequence; performing joint phase space reconstruction on the intra-section current deviation sequence and intra-section voltage deviation sequence to obtain joint current trajectory points and joint voltage trajectory points; determining the outlier significance of the joint trajectory of the feeder section based on the joint current trajectory points and the joint voltage trajectory points; and identifying the faulty feeder section of the distribution network based on the outlier significance of the joint trajectory. This invention can improve the efficiency of intelligent location of faulty feeder sections in distribution networks.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of power distribution network technology, and in particular to a method and system for intelligent location of fault sections in power distribution network feeders based on feature recognition. Background Technology

[0002] As the final power supply carrier of the power system, the distribution network's feeder sections are the core units for power transmission and load distribution. Feeder short circuits, grounding faults, and other faults can directly cause regional power outages and equipment damage. Fault location is a crucial preliminary step in the operation and maintenance of the distribution network. The mainstream approach in the industry relies on the timing measurement data of voltage and current at the beginning and end of the feeder to extract fault characteristics and distinguish faulty sections from normal sections based on differences in electrical quantities. The accuracy of location and the efficiency of calculation directly determine the speed of distribution network fault recovery.

[0003] Existing fault location technologies for distribution network feeders only utilize single electrical quantities such as current or voltage for fault feature analysis, failing to combine them to construct phase-space trajectory features. Furthermore, they lack a joint trajectory outlier quantification evaluation mechanism for adjacent feeder sections. Single electrical quantity features are susceptible to load fluctuations and line noise interference, exhibiting weak differentiation between normal and faulty sections, making it impossible to accurately quantify the degree of deviation of fault features. This leads to frequent misjudgments and missed detections in fault section identification, and makes it difficult to quickly locate faults at section boundaries, significantly reducing the overall computational and identification efficiency of fault location. Therefore, how to integrate voltage and current as dual electrical quantities to construct joint trajectory features, quantify fault outlier levels, and improve the intelligent location efficiency of faulty sections in distribution network feeders has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a method and system for intelligent location of fault sections in distribution network feeders based on feature recognition, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for intelligent location of fault sections in distribution network feeders based on feature recognition, comprising:

[0006] S01. Synchronously collect the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence of the feeder section in the distribution network.

[0007] S02. Perform point-by-point differential transformation on the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence and the last-end aligned voltage sequence to obtain the intra-region current deviation sequence and intra-region voltage deviation sequence of the feeder section.

[0008] S03. Perform joint phase space reconstruction on the current deviation sequence and the voltage deviation sequence in the region respectively to obtain the joint current trajectory point and the joint voltage trajectory point of the feeder section.

[0009] S04. Determine the outlier significance of the joint trajectory of the feeder segment based on the current joint trajectory point and the voltage joint trajectory point;

[0010] S05. Based on the outlier significance of the joint trajectory, determine the faulty feeder section of the distribution network.

[0011] In a preferred embodiment, the synchronous acquisition of the head-end aligned current sequence, the end-end aligned current sequence, the head-end aligned voltage sequence, and the end-end aligned voltage sequence of the feeder section in the distribution network includes:

[0012] Synchronously collect time-series current and voltage data at the beginning and end of feeder sections in the distribution network;

[0013] The time-series current data and the time-series voltage data are aligned with a time reference to obtain the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence of the feeder segment.

[0014] In a preferred embodiment, the step of performing a point-by-point differential transformation on the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence to obtain the intra-region current deviation sequence and intra-region voltage deviation sequence of the feeder section includes:

[0015] The point-by-point amplitude difference operation is performed on the first-end aligned current sequence and the last-end aligned current sequence to obtain the in-region current attenuation of the feeder section.

[0016] The difference between the magnitude of the current attenuation in the region and the mean of the current attenuation in the region is calculated, and the absolute value is taken to obtain the current deviation sequence in the feeder section.

[0017] The voltage drop within the feeder section is obtained by performing point-by-point amplitude difference calculation on the first-end aligned voltage sequence and the last-end aligned voltage sequence.

[0018] The difference between the magnitude of the voltage drop in the region and the mean of the voltage drop in the region is calculated, and the absolute value is taken to obtain the voltage deviation sequence of the feeder section.

[0019] In a preferred embodiment, the step of performing joint phase space reconstruction on the current deviation sequence and the voltage deviation sequence within the region to obtain the joint current trajectory point and the joint voltage trajectory point of the feeder segment includes:

[0020] Based on the topological adjacency relationship of the distribution network, the feeder section is divided into adjacent section pairs, and each adjacent section pair consists of two directly adjacent feeder sections;

[0021] Using the sampling time of the feeder section as the synchronization anchor point, the deviation amplitudes of the current deviation sequences in the middle zone of the adjacent sections are paired and arranged into ordered current binary pairs.

[0022] According to the sampling sequence, the ordered current binary groups are arranged sequentially in the two-dimensional amplitude plane to obtain the current joint trajectory points of the adjacent segment pairs;

[0023] Based on the synchronization anchor point, the deviation amplitudes of the voltage deviation sequences in the adjacent segments within the central region are paired and arranged into ordered voltage binary pairs.

[0024] According to the sampling sequence, the ordered voltage pairs are arranged sequentially in the two-dimensional amplitude plane to obtain the voltage joint trajectory points of the adjacent segment pairs.

[0025] In a preferred embodiment, the current joint trajectory point and the voltage joint trajectory point include:

[0026] The horizontal axis of the current joint trajectory point is the deviation amplitude of the first feeder segment in the adjacent segment alignment at the current sampling time, and the vertical axis is the deviation amplitude of the second feeder segment in the adjacent segment alignment at the same sampling time.

[0027] The horizontal axis of the voltage joint trajectory point is the deviation amplitude of the first feeder segment in the adjacent segment alignment at the current sampling time, and the vertical axis is the deviation amplitude of the second feeder segment in the adjacent segment alignment at the same sampling time.

[0028] In a preferred embodiment, determining the outlier significance of the joint trajectory of the feeder segment based on the current joint trajectory point and the voltage joint trajectory point includes:

[0029] Using the abscissa and ordinate values ​​of the current joint trajectory points as the calculation basis, the distance between the current joint trajectory points and the origin is calculated according to the square root rule. The distance values ​​are then arranged in the order of sampling time to obtain the current distance values ​​of the adjacent segment pairs.

[0030] The sequence center value of the current distance values ​​is obtained by dividing the cumulative sum of the current distance values ​​by the number of current distance values.

[0031] Calculate the deviation magnitude between the current distance value and the center value of the sequence one by one;

[0032] The deviation amplitude is squared and summed, and the sum of squares is divided by the number of values ​​to obtain the mean square deviation value. Then, the mean square deviation value is squared to obtain the discrete sequence value of the current distance value.

[0033] Dividing the discrete value of the sequence by the center value of the sequence yields the current joint trajectory convergence / divergence degree of the adjacent segment pairs;

[0034] Based on the method for obtaining the current joint trajectory convergence, the voltage joint trajectory convergence of the adjacent segment pairs is obtained.

[0035] Based on the current joint trajectory convergence and divergence and the voltage joint trajectory convergence and divergence, the convergence and divergence comprehensive index of the adjacent segment pairs is calculated;

[0036] Based on the aggregation-dispersion comprehensive index, the outlier significance of the joint trajectory of the adjacent segment pairs is calculated.

[0037] In a preferred embodiment, calculating the combined convergence / dispersion index of the adjacent segment pairs based on the current joint trajectory convergence and the voltage joint trajectory convergence includes:

[0038] The formula for calculating the aggregation and dispersion comprehensive index is:

[0039] ;

[0040] In the formula, For the first The aggregation and dispersion comprehensive index of adjacent segment pairs For the first Current joint trajectory convergence and divergence of adjacent segments For the first The convergence and divergence of voltage joint trajectories for adjacent segments.

[0041] In a preferred embodiment, calculating the outlier significance of the joint trajectory of the adjacent segment pairs based on the aggregation-dispersion composite index includes:

[0042] The aggregation and dispersion comprehensive index is averaged and merged to obtain the set center value of the aggregation and dispersion comprehensive index;

[0043] The second-order central moments of the aggregation-dispersion composite index are obtained by calculating the set discrete values ​​of the aggregation-dispersion composite index.

[0044] A nonlinear outlier mapping is performed on the aggregation-dispersion composite index to obtain the joint trajectory outlier significance of the adjacent segment pairs. The formula for calculating the joint trajectory outlier significance is as follows:

[0045] ;

[0046] In the formula, For the first Outlier significance of joint trajectory of adjacent segment pairs For the discrete values ​​of the set, For the first The aggregation and dispersion comprehensive index of adjacent segment pairs The set center value, This is the exponentiation operator.

[0047] In a preferred embodiment, determining the faulty feeder section of the distribution network based on the joint trajectory outlier significance includes:

[0048] The adjacent segments are arranged from largest to smallest according to the numerical value of the outlier significance of the joint trajectory.

[0049] Based on the sorting results, the direction of current flow is traced from the power supply side to the load side along the feeder transmission path, and the feeder section located downstream of the fault current in the first adjacent section pair is identified as the faulty feeder section.

[0050] To address the aforementioned problems, this invention also provides a feature-based intelligent location system for fault sections in distribution network feeders, the system comprising:

[0051] The data synchronization acquisition module is used to synchronously acquire the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence of the feeder section in the distribution network.

[0052] The differential transformation module is used to perform point-by-point differential transformation on the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence and the last-end aligned voltage sequence to obtain the intra-region current deviation sequence and intra-region voltage deviation sequence of the feeder section.

[0053] The joint phase space reconstruction module is used to perform joint phase space reconstruction on the current deviation sequence and the voltage deviation sequence in the region respectively, so as to obtain the joint current trajectory point and the joint voltage trajectory point of the feeder section.

[0054] The outlier significance determination module is used to determine the outlier significance of the joint trajectory of the feeder segment based on the current joint trajectory point and the voltage joint trajectory point.

[0055] The fault section determination module is used to determine the faulty feeder section of the distribution network based on the outlier significance of the joint trajectory.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] 1. This invention synchronously collects voltage and current time series aligned at both ends of a feeder section, generates intra-section deviation sequences through point-by-point differential transformation, and completes joint phase space reconstruction of adjacent section pairs based on the distribution network topology. It synchronously constructs two-dimensional joint trajectory points of current and voltage, fully preserves the voltage and current coupling change characteristics under fault conditions, fully explores the fault information contained in the dual electrical quantities, strengthens the data feature differentiation between fault sections and normal sections, and improves the basic reliability of fault identification from the underlying data feature level.

[0058] 2. This invention constructs a convergence and dispersion comprehensive index that integrates the convergence and dispersion characteristics of current and voltage trajectories. Then, it calculates the outlier significance of the joint trajectory through nonlinear mapping and uses quantitative indicators to automatically identify fault sections. This enables precise location of fault sections downstream of the fault current. The entire process achieves full-process quantitative calculation, reduces manual judgment intervention, significantly shortens the calculation time for fault section location in distribution network feeders, and effectively improves the overall efficiency of intelligent location of fault sections in distribution network feeders. Attached Figure Description

[0059] Figure 1 This is a flowchart illustrating a method for intelligent location of fault sections in a distribution network feeder based on feature recognition, according to an embodiment of the present invention.

[0060] Figure 2 A functional block diagram of a distribution network feeder fault section intelligent positioning system based on feature recognition provided in an embodiment of the present invention;

[0061] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0062] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0063] This application provides a method for intelligent location of faulty sections in distribution network feeders based on feature recognition. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for intelligent location of faulty sections in distribution network feeders based on feature recognition can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0064] Reference Figure 1 The diagram shown is a flowchart illustrating a method for intelligent location of fault sections in distribution network feeders based on feature recognition, according to an embodiment of the present invention. In this embodiment, the method for intelligent location of fault sections in distribution network feeders based on feature recognition includes:

[0065] S01. Synchronously collect the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence of the feeder section in the distribution network.

[0066] In this embodiment of the invention, the synchronous acquisition of the head-end aligned current sequence, the end-end aligned current sequence, the head-end aligned voltage sequence, and the end-end aligned voltage sequence of the feeder section in the distribution network includes:

[0067] Synchronously collect time-series current and voltage data at the beginning and end of feeder sections in the distribution network;

[0068] The time-series current data and the time-series voltage data are aligned with a time reference to obtain the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence of the feeder segment.

[0069] The data acquisition devices deployed at the beginning and end of the feeder section of the distribution network receive the same standard time signal. The two sets of data acquisition devices continuously capture the real-time electrical quantities of the line at the same sampling interval. The data acquisition device at the beginning captures the real-time current and voltage values ​​of the line. Each capture action is bound to a corresponding capture time mark. The set of all the current values ​​at the beginning with time marks forms the time-series current data, and the set of all the voltage values ​​at the beginning with time marks forms the time-series voltage data.

[0070] The data acquisition device at the end position synchronously performs the capture work in complete sync with the action rhythm of the acquisition device at the head position. The data acquisition device at the end position captures the real-time current value and real-time voltage value of the line. Each capture action is bound to a corresponding capture time mark. The set of all end current values ​​with time marks forms the time-series current data, and the set of all end voltage values ​​with time marks forms the time-series voltage data.

[0071] The time-series current data, time-series voltage data, and time-series current data generated at the beginning and end of the feeder section are uniformly collected and stored in the same data storage medium. The time-series current data is a collection formed by summarizing the current values ​​captured by the beginning and end acquisition devices in this section with time stamps, and the time-series voltage data is a collection formed by summarizing the voltage values ​​captured by the beginning and end acquisition devices in this section with time stamps.

[0072] Extract all time markers carried by all time-series current data and all time-series voltage data stored in the storage medium. All time markers are directly generated by the acquisition device after receiving the standard time signal when a single electrical quantity acquisition is completed. Each set of time markers corresponds to the current value or voltage value obtained in a single acquisition.

[0073] The time markers that appear with completely overlapping frequencies among all time markers are selected as the unified time reference. The unified time reference is determined by the time markers that appear in multiple sets of time series data. Only the current and voltage values ​​that carry the unified time reference mark are retained in each type of time series data, and the values ​​that do not carry the unified time reference mark are removed.

[0074] Collect and retain the current values ​​of each time at the beginning of the time reference mark, and arrange them in the order in which the time reference mark appears. After the arrangement is completed, the current sequence of the beginning is obtained. The current sequence of the beginning is a dedicated data set generated after this step is completed.

[0075] The end current values ​​of each time with a unified time reference mark are collected and arranged in the order in which the unified time reference mark appears. After the arrangement is completed, the end-aligned current sequence is obtained. The end-aligned current sequence is a dedicated data set generated after this step is completed.

[0076] Collect and retain the voltage values ​​of each time at the beginning of the time reference mark, and arrange them in the order in which the time reference mark appears. After the arrangement is completed, the voltage sequence with the beginning aligned is obtained. The voltage sequence with the beginning aligned is a dedicated data set generated after this step is completed.

[0077] Collect and retain the end voltage values ​​of the unified time reference mark, and arrange them in the order of appearance of the unified time reference mark. After the arrangement is completed, the end-aligned voltage sequence is obtained. The end-aligned voltage sequence is a dedicated data set generated after this step is completed.

[0078] The first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence together constitute a complete electrical sequence set for the feeder section after time base alignment processing. The time markers corresponding to each position in the four types of sequences are completely matched, meeting the usage standards for synchronous acquisition and output data.

[0079] The beneficial effects are that, relying on a unified standard time signal, the acquisition devices at the beginning and end of the line execute the same capture rhythm. All electrical values ​​are bound to corresponding time markers to form time-series current and voltage data. By extracting all time markers and filtering overlapping markers as a unified time reference, mismatched values ​​are eliminated and the data are arranged according to the order of the markers to generate four types of alignment sequences. Throughout the process, the time markers at corresponding positions within the four types of sequences are completely matched. This ensures that the voltage and current values ​​at both ends of the feeder section at the same instant are completely preserved, eliminating data correspondence deviations caused by timing misalignment. The electrical change correlation of the line at the same time is completely preserved, providing a complete set of matching basic data for subsequent section status analysis based on the electrical values ​​at both ends. No additional data correction operations are required throughout the process, and a complete electrical sequence set that meets the synchronous use standard is stably output.

[0080] S02. Perform point-by-point differential transformation on the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence and the last-end aligned voltage sequence to obtain the intra-region current deviation sequence and intra-region voltage deviation sequence of the feeder section.

[0081] In this embodiment of the invention, the step of performing point-by-point differential transformation on the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence to obtain the intra-region current deviation sequence and intra-region voltage deviation sequence of the feeder segment includes:

[0082] The point-by-point amplitude difference operation is performed on the first-end aligned current sequence and the last-end aligned current sequence to obtain the in-region current attenuation of the feeder section.

[0083] The difference between the magnitude of the current attenuation in the region and the mean of the current attenuation in the region is calculated, and the absolute value is taken to obtain the current deviation sequence in the feeder section.

[0084] The voltage drop within the feeder section is obtained by performing point-by-point amplitude difference calculation on the first-end aligned voltage sequence and the last-end aligned voltage sequence.

[0085] The difference between the magnitude of the voltage drop in the region and the mean of the voltage drop in the region is calculated, and the absolute value is taken to obtain the voltage deviation sequence of the feeder section.

[0086] Extract all amplitude values ​​arranged sequentially according to time markers within the first-end aligned current sequence, and extract all amplitude values ​​arranged at the same time marker position within the last-end aligned current sequence. The two sets of amplitude values ​​are generated by the preceding time base alignment operation, and the time markers at the same sequential position within the two types of sequences are completely consistent.

[0087] According to the order of the time markers within the sequence, the amplitude values ​​of the first-aligned current sequence and the last-aligned current sequence at the same position are extracted sequentially. The amplitude value of the first-aligned current sequence is subtracted from the amplitude value of the last-aligned current sequence. After each subtraction operation, a set of exclusive values ​​is obtained. The values ​​produced by all the operations are integrated and summarized according to the original order to form the current attenuation within the region. The current attenuation within the region is the exclusive data set produced by the amplitude difference operation in this step.

[0088] Extract all sequentially arranged numerical values ​​of the current attenuation within the region. All numerical values ​​are generated by point-by-point amplitude difference operation of the preceding head-aligned current sequence and the end-aligned current sequence. Count the total number of contents contained in all numerical values. Superimpose and summarize all numerical values ​​to obtain the superimposed sum. Divide the superimposed sum by the total number of contents to obtain the mean of the current attenuation within the region. The mean of the current attenuation within the region is determined by the superimposed sum of all numerical values ​​within the current attenuation within the region and the total number of contents. The mean of the current attenuation within the region is taken as the average attenuation of 100 sampling points in the two power frequency cycles before the fault. This steady-state mean remains fixed during the fault process and is not updated with the fault transient data.

[0089] According to the original order of the current attenuation within the zone, the numerical content corresponding to each order is extracted sequentially. The mean value of the current attenuation within the zone is subtracted from the numerical content of the order to obtain a set of difference content. The positive and negative attributes of the difference content are identified, and the positive and negative labels of the difference content are discarded, leaving only the numerical part corresponding to the difference content. Each order operation produces numerical content without positive and negative labels. All numerical content without positive and negative labels produced by the order are integrated and summarized according to the original order to form the current deviation sequence within the zone. The current deviation sequence within the zone is a unique data set produced by the absolute value operation of the difference in this step.

[0090] Extract all amplitude values ​​arranged sequentially according to time markers within the first-end aligned voltage sequence, and extract all amplitude values ​​arranged at the same time marker position within the last-end aligned voltage sequence. The two sets of amplitude values ​​are generated by the preceding time base alignment operation, and the time markers corresponding to the same sequential position within the two types of sequences are completely identical.

[0091] According to the order of the time markers within the sequence, the amplitude values ​​of the first-aligned voltage sequence and the last-aligned voltage sequence at the same position are extracted in sequence. The amplitude value of the last-aligned voltage sequence is subtracted from the amplitude value of the first-aligned voltage sequence. After each subtraction operation, a set of unique values ​​is obtained. The values ​​produced by all the operations are integrated and summarized according to the original order to form the voltage drop in the region.

[0092] Extract all sequentially arranged numerical values ​​within the voltage drop range. All numerical values ​​are generated by point-by-point amplitude difference operations on the preceding head-aligned voltage sequence and the tail-aligned voltage sequence. Count the total number of content segments contained in all numerical values. Superimpose and summarize all numerical values ​​to obtain the superimposed sum. Divide the superimposed sum by the total number of content segments to obtain the mean of the voltage drop range within the range. The mean of the voltage drop range within the range is determined by the sum of all numerical values ​​within the range and the total number of content segments. The voltage drop range within the range is a dedicated dataset generated by the amplitude difference operation in this step. The mean of the voltage drop range within the range is the average of the voltage drop ranges of 100 sampling points over the two power frequency cycles prior to the fault.

[0093] According to the original order of voltage drop within the zone, the numerical content corresponding to each order is extracted sequentially. The average value of voltage drop within the zone is subtracted from the numerical content of the order to obtain a set of difference content. The positive and negative attributes of the difference content are identified, and the positive and negative labels of the difference content are discarded, leaving only the numerical part corresponding to the difference content. Each order operation produces numerical content without positive and negative labels. All numerical content without positive and negative labels produced by the order are integrated and summarized according to the original order to form the voltage deviation sequence within the zone. The voltage deviation sequence within the zone is a unique data set produced by the absolute value operation of the difference in this step.

[0094] The current deviation sequence and the voltage deviation sequence within the region together constitute two exclusive data sets generated after the point-by-point differential transformation operation. Each position within the two sets corresponds to the same original acquisition time marker, fully carrying the corresponding data content of the current and voltage at both ends of the feeder section after differential processing.

[0095] The beneficial effect is that, based on the four types of electrical sequences completed by alignment, the magnitude subtraction operation is performed to generate the current attenuation and voltage drop within the region. Then, based on the average value of all values ​​within the corresponding attenuation and drop, the difference is calculated and the positive and negative labels are removed to obtain two types of deviation sequences. The corresponding data matching processing is completed according to the unified time mark throughout the process. It can remove the basic data interference caused by the average change of the overall electrical quantity of the line, and retain the unique change information of the electrical quantity within the section that deviates from the average level. The deviation data of the current dimension and the voltage dimension are retained separately at the same time. The two types of deviation sequences maintain the correspondence of the original acquisition time mark, and completely retain the corresponding relationship of the current and voltage deviations at the same instant. This provides sufficient underlying data material with high discrimination for subsequent section status identification based on the two types of electrical deviation information.

[0096] S03. Perform joint phase space reconstruction on the current deviation sequence and the voltage deviation sequence in the region respectively to obtain the joint current trajectory point and the joint voltage trajectory point of the feeder section.

[0097] In this embodiment of the invention, the step of performing joint phase space reconstruction on the current deviation sequence and the voltage deviation sequence within the region to obtain the joint current trajectory point and the joint voltage trajectory point of the feeder segment includes:

[0098] Based on the topological adjacency relationship of the distribution network, the feeder section is divided into adjacent section pairs, and each adjacent section pair consists of two directly adjacent feeder sections;

[0099] Using the sampling time of the feeder section as the synchronization anchor point, the deviation amplitudes of the current deviation sequences in the middle zone of the adjacent sections are paired and arranged into ordered current binary pairs.

[0100] According to the sampling sequence, the ordered current binary groups are arranged sequentially in the two-dimensional amplitude plane to obtain the current joint trajectory points of the adjacent segment pairs;

[0101] Based on the synchronization anchor point, the deviation amplitudes of the voltage deviation sequences in the adjacent segments within the central region are paired and arranged into ordered voltage binary pairs.

[0102] According to the sampling sequence, the ordered voltage pairs are arranged sequentially in the two-dimensional amplitude plane to obtain the voltage joint trajectory points of the adjacent segment pairs.

[0103] The current joint trajectory point and the voltage joint trajectory point include:

[0104] The horizontal axis of the current joint trajectory point is the deviation amplitude of the first feeder segment in the adjacent segment alignment at the current sampling time, and the vertical axis is the deviation amplitude of the second feeder segment in the adjacent segment alignment at the same sampling time.

[0105] The horizontal axis of the voltage joint trajectory point is the deviation amplitude of the first feeder segment in the adjacent segment alignment at the current sampling time, and the vertical axis is the deviation amplitude of the second feeder segment in the adjacent segment alignment at the same sampling time.

[0106] In this embodiment, the joint phase space reconstruction is a two-dimensional feature mapping method oriented towards adjacent sections. It does not require the embedding dimension and delay time parameters of traditional phase space reconstruction. It only uses the synchronous deviation amplitude of two adjacent sections as two-dimensional coordinates to construct trajectory points, which intuitively reflects the linkage characteristics of electrical deviation between the two sections.

[0107] This step involves identifying the interconnections between all feeder sections within the distribution network. These interconnections form the basis for determining the topological adjacency relationships in this step. Topological adjacency relationships refer to the arrangement relationships formed by the direct interconnections of all feeder sections.

[0108] Two groups of feeder segments with direct connection are selected from all feeder segments and combined. The combined content of each group is defined as an adjacent segment pair. The adjacent segment pair is determined by two groups of directly connected feeder segments selected from the topological adjacency relationship.

[0109] Extract the sampling time associated with each group of arrangement content within the intra-regional current deviation sequence produced by the preceding operation. The sampling time is a time marker generated when the acquisition device captures electrical values. The sampling time serves as a synchronization anchor point, which is determined by the sampling time content carried by the intra-regional current deviation sequence.

[0110] Extract the internal deviation amplitude of the internal current deviation sequence corresponding to the first feeder segment within the adjacent segment, and extract the deviation amplitude corresponding to the same synchronous anchor point within the internal current deviation sequence corresponding to the second feeder segment within the adjacent segment. Pair the deviation amplitudes of the two sets of corresponding synchronous anchor points in a fixed order of amplitude of the first feeder segment first and amplitude of the second feeder segment last. The two sets of amplitude contents completed by a single synchronous anchor point pairing arrangement form an ordered current binary group. The ordered current binary group is the exclusive content produced by the amplitude pairing arrangement in this step.

[0111] According to the order in which the synchronization anchors are generated by the acquisition action, the ordered current binary pairs corresponding to each set of synchronization anchors are extracted in sequence. The amplitude content of the first set of ordered current binary pairs is placed in the horizontal extension direction of the two-dimensional amplitude plane, and the amplitude content of the last set of ordered current binary pairs is placed in the vertical extension direction of the two-dimensional amplitude plane. After each set of ordered current binary pairs is placed, a unique point content is formed. The point content generated corresponding to all synchronization anchors is summarized to form the current joint trajectory point. The current joint trajectory point is the unique point content produced by the ordered current binary pair placement operation in this step.

[0112] The synchronization anchor point is used as the unified reference standard for amplitude pairing and arrangement. The synchronization anchor point is determined by the sampling time content carried by the voltage deviation sequence within the region.

[0113] Extract the internal deviation amplitude of the voltage deviation sequence corresponding to the first feeder segment within the adjacent segment, and extract the deviation amplitude corresponding to the same synchronization anchor point within the voltage deviation sequence corresponding to the second feeder segment within the adjacent segment. Pair the deviation amplitudes of the two sets of corresponding synchronization anchor points in a fixed order, with the amplitude of the first feeder segment preceding the amplitude of the second feeder segment. The two sets of amplitude contents completed in a single synchronization anchor point pairing arrangement form an ordered voltage binary. The ordered voltage binary is the exclusive content produced by the amplitude pairing arrangement in this step.

[0114] According to the order in which the synchronization anchors are generated by the acquisition action, the ordered voltage binary pairs corresponding to each set of synchronization anchors are extracted in sequence. The amplitude content of the first set of ordered voltage binary pairs is placed in the horizontal extension direction of the two-dimensional amplitude plane, and the amplitude content of the last set of ordered voltage binary pairs is placed in the vertical extension direction of the two-dimensional amplitude plane. After each set of ordered voltage binary pairs is placed, a unique point content is formed. The point content generated by all synchronization anchors is summarized to form a voltage joint trajectory point. The voltage joint trajectory point is the unique point content produced by the ordered voltage binary pair placement operation in this step.

[0115] The value corresponding to the lateral extension direction inside the current joint trajectory point is taken from the deviation amplitude of the intra-region current deviation sequence under the synchronous anchor point of the first feeder section in the adjacent section. The value corresponding to the longitudinal extension direction inside the current joint trajectory point is taken from the deviation amplitude of the intra-region current deviation sequence under the synchronous anchor point of the second feeder section in the adjacent section. The two sets of amplitude contents are determined by the internal arrangement of the preceding ordered current binary group.

[0116] The values ​​corresponding to the horizontal extension direction inside the voltage joint trajectory point are taken from the voltage deviation sequence deviation amplitude under the synchronous anchor point of the first feeder section in the adjacent section. The values ​​corresponding to the vertical extension direction inside the voltage joint trajectory point are taken from the voltage deviation sequence deviation amplitude under the synchronous anchor point of the second feeder section in the adjacent section. The two sets of amplitude contents are determined by the internal arrangement of the preceding ordered voltage binary.

[0117] The current joint trajectory points and voltage joint trajectory points completely preserve the corresponding arrangement relationship of the deviation amplitude of the current dimension and voltage dimension under the same synchronous anchor point of the two sets of feeder sections in the adjacent section pair. The content of both types of points is generated by arranging them based on the unified synchronous anchor point. The source of the corresponding amplitude in the horizontal and vertical directions within the point is completely matched with the two feeder sections of the adjacent section pair.

[0118] The beneficial effects are as follows: by dividing adjacent segments based on the connection and correlation of feeder segments, using a unified sampling time as a synchronous anchor point to pair the deviation amplitudes of the two segments to generate ordered binary pairs, and then generating two types of trajectory points based on the two-dimensional amplitude plane arrangement, the segment sources corresponding to the horizontal and vertical amplitudes are unified throughout the process, and the point content of the current dimension and voltage dimension are constructed synchronously and separately. This can integrate the electrical deviation information of two adjacent segments at the same instant into the same plane carrier, fully retain the synchronous correspondence between the two types of electrical quantities, current and voltage, between adjacent segments, and transform the single deviation information carried by the one-dimensional sequence into point content with bidirectional comparison relationship, enriching the comparison dimension of electrical changes between segments, and providing complete comparison reference material for subsequent quantification of the aggregation state of electrical deviations of segments.

[0119] S04. Determine the outlier significance of the joint trajectory of the feeder segment based on the current joint trajectory point and the voltage joint trajectory point;

[0120] In this embodiment of the invention, determining the outlier significance of the joint trajectory of the feeder segment based on the current joint trajectory point and the voltage joint trajectory point includes:

[0121] Using the abscissa and ordinate values ​​of the current joint trajectory points as the calculation basis, the distance between the current joint trajectory points and the origin is calculated according to the square root rule. The distance values ​​are then arranged in the order of sampling time to obtain the current distance values ​​of the adjacent segment pairs.

[0122] The sequence center value of the current distance values ​​is obtained by dividing the cumulative sum of the current distance values ​​by the number of current distance values.

[0123] Calculate the deviation magnitude between the current distance value and the center value of the sequence one by one;

[0124] The deviation amplitude is squared and summed, and the sum of squares is divided by the number of values ​​to obtain the mean square deviation value. Then, the mean square deviation value is squared to obtain the discrete sequence value of the current distance value.

[0125] Dividing the discrete value of the sequence by the center value of the sequence yields the current joint trajectory convergence / divergence degree of the adjacent segment pairs;

[0126] Based on the method for obtaining the current joint trajectory convergence, the voltage joint trajectory convergence of the adjacent segment pairs is obtained.

[0127] Based on the current joint trajectory convergence and divergence and the voltage joint trajectory convergence and divergence, the convergence and divergence comprehensive index of the adjacent segment pairs is calculated;

[0128] Based on the aggregation-dispersion comprehensive index, the outlier significance of the joint trajectory of the adjacent segment pairs is calculated.

[0129] The calculation of the combined convergence / dispersion index of the adjacent segment pairs based on the current joint trajectory convergence / dispersion and the voltage joint trajectory convergence / dispersion includes:

[0130] The formula for calculating the aggregation and dispersion comprehensive index is:

[0131] ;

[0132] In the formula, For the first The aggregation and dispersion comprehensive index of adjacent segment pairs For the first Current joint trajectory convergence and divergence of adjacent segments For the first The convergence and divergence of voltage joint trajectories for adjacent segments.

[0133] The corresponding current joint trajectory convergence and divergence is a unique value derived from the distribution operation performed after calculating the distance between points on the current joint trajectory, the sequence center value, the deviation amplitude, and the sequence discrete value in adjacent segments.

[0134] The corresponding voltage joint trajectory convergence / divergence is derived from the entire generation process of the current joint trajectory convergence / divergence for adjacent segments, and is the exclusive content generated after processing the voltage joint trajectory points. This is a comprehensive index of aggregation and dispersion, and this content is the output obtained after fusion processing here.

[0135] This processing procedure incorporates the current-coupled trajectory convergence and voltage-coupled trajectory convergence of adjacent segments into the processing flow. First, the two types of convergence are multiplied and then square-rooted. Then, the distribution result between the result of multiplying the two types of convergence by two times and the result of multiplying each type of convergence by itself and then adding them together is obtained. The content obtained from the square root process and the content obtained from the distribution process are multiplied again to obtain the convergence comprehensive index. This process fully absorbs the trajectory convergence and divergence state information of both the current dimension and the voltage dimension, and obtains a unified output content that can simultaneously reflect the state of both types of trajectory.

[0136] When the content obtained by multiplying and taking the square root of the current joint trajectory convergence and voltage joint trajectory convergence changes, or when the allocation content of the result of doubling the convergence and the sum of the results of multiplying and taking the sum of the two types of convergence changes, the output content of the convergence comprehensive index will change accordingly. When the two types of input content change simultaneously and equally, the allocated content remains unchanged, and the output content of the convergence comprehensive index will only change accordingly with the content obtained by multiplying and taking the square root.

[0137] The calculation of the outlier significance of the joint trajectory of the adjacent segment pairs based on the aggregation-dispersion comprehensive index includes:

[0138] The aggregation and dispersion comprehensive index is averaged and merged to obtain the set center value of the aggregation and dispersion comprehensive index;

[0139] The second-order central moments of the aggregation-dispersion composite index are obtained by calculating the set discrete values ​​of the aggregation-dispersion composite index.

[0140] A nonlinear outlier mapping is performed on the aggregation-dispersion composite index to obtain the joint trajectory outlier significance of the adjacent segment pairs. The formula for calculating the joint trajectory outlier significance is as follows:

[0141] ;

[0142] In the formula, For the first Outlier significance of joint trajectory of adjacent segment pairs For the discrete values ​​of the set, For the first The aggregation and dispersion comprehensive index of adjacent segment pairs The set center value, This is the exponentiation operator.

[0143] The corresponding convergence / dispersion integrated index is derived from the specific data generated after fusing the convergence / dispersion of the corresponding current joint trajectory and the convergence / dispersion of the voltage joint trajectory for a single group of adjacent segments. The corresponding set center value is determined by the sum of the aggregate and dispersion indices generated from all adjacent segments and the distribution of the total number of indices. The corresponding discrete value is determined by the superposition and distribution of the squared second-order differences of all aggregation and dispersion comprehensive indices. This is the joint trajectory outlier significance, which is the output obtained after nonlinear outlier degree mapping processing.

[0144] The process first takes a single pair of adjacent segments and performs a difference operation between the corresponding aggregation-dispersion comprehensive index and the set center value. The difference is then distributed with the set discrete value to obtain the first part of the processed content. Next, the same difference is distributed with twice the set discrete value. The output of the distribution is sent to the exponential calculation process to complete the transformation and obtain the second part of the processed content. The first part of the processed content and the second part of the processed content are multiplied to obtain the joint trajectory outlier significance. This process relies on the overall distribution benchmark of all segment pairs to complete the deviation transformation of a single pair of segment pairs and complete the nonlinear mapping from the aggregation-dispersion comprehensive index to the outlier representation content.

[0145] When the difference between the aggregation and dispersion comprehensive index and the set center value changes, the two parts of the processing content allocated are changed synchronously. The transformation content produced by the index operation process changes synchronously with the allocated input content. The outlier significance of the joint trajectory obtained after multiplication changes synchronously with the corresponding change. The set center value and the set discrete value are taken from the unified processing results of all adjacent segment pairs. After this part of the content is fixed, the final output content is determined only by the specific content of a single set of aggregation and dispersion comprehensive indexes.

[0146] Extract the amplitude content corresponding to the horizontal extension direction and the amplitude content corresponding to the vertical extension direction within the current joint trajectory point. The two types of amplitude content are determined by the arrangement of the previous ordered current binary tuples. After multiplying each of the two sets of amplitude content with itself, merge and summarize the output of the two sets of multiplication. Then, perform square root processing on the summarized content. After the processing of a single current joint trajectory point is completed, a unique distance content is generated. The distance content generated by all current joint trajectory points is arranged in the order of the synchronous anchor point generation to form the current distance value. The current distance value is the unique content set produced by the point distance processing in this step.

[0147] Extract all the complete distance content within the current distance value, count the total number of parts corresponding to all distance content, and sum up all the distance content to obtain the sum of the sum. Then, use the sum of the sum to allocate each distance content to the corresponding part. After the allocation is completed, the sequence center value is obtained. The sequence center value is determined by the sum of all the current distance values ​​and the total number of parts.

[0148] Extract each set of distance values ​​arranged within the current distance values ​​in sequence. Perform a difference operation between each set of distance values ​​and the center value of the sequence. Remove the positive and negative labels of the difference and retain the pure numerical content. Each operation produces the deviation amplitude of the corresponding set of distance values. The deviation amplitude is determined by removing the positive and negative labels from the difference between the set of current distance values ​​and the center value of the sequence.

[0149] Each deviation amplitude is multiplied by itself, and the results of all multiplications are summed to obtain the squared cumulative value. The squared cumulative value is then distributed to the total number of deviation amplitudes. After distribution, the mean square deviation value is obtained. The mean square deviation value is then squared to obtain the discrete value of the sequence. The discrete value of the sequence is determined by square rooting the result of the summation and distribution of the squared deviation amplitudes.

[0150] The discrete values ​​of the sequence are assigned to the corresponding center values ​​of the sequence. After the assignment is completed, the current joint trajectory convergence degree is obtained. The current joint trajectory convergence degree is the exclusive content produced by the assignment operation in this step.

[0151] The voltage distance value is obtained by processing the horizontal and vertical extension amplitude values ​​within the voltage joint trajectory points. The corresponding sequence center value is then obtained based on the voltage distance value. The deviation amplitude corresponding to the voltage distance value is then calculated one by one. The deviation amplitude is then processed by square superposition and square root to obtain the discrete value of the voltage dimension sequence. The discrete value of the voltage dimension sequence is then distributed with the center value of the voltage dimension sequence to obtain the voltage joint trajectory convergence degree. The voltage joint trajectory convergence degree completely replicates the complete generation process of the current joint trajectory convergence degree.

[0152] The current joint trajectory convergence and voltage joint trajectory convergence of adjacent segments in a single group are integrated, and the electrical dimension characteristics carried by each type of convergence are taken into account simultaneously to complete the fusion process. After the fusion process is completed, a comprehensive convergence index is generated. The comprehensive convergence index is the exclusive content generated by the fusion operation of the two types of convergence in this step.

[0153] Collect all the clustering and dispersion comprehensive indicators generated by each of the adjacent segments, count the total number of clustering and dispersion comprehensive indicators, sum all the clustering and dispersion comprehensive indicators to obtain the sum, allocate the sum to each clustering and dispersion comprehensive indicator, and obtain the set center value after allocation. The set center value is determined by the sum of all clustering and dispersion comprehensive indicators and the total number of indicators.

[0154] Each aggregate-dispersion composite index is taken out sequentially and its difference is calculated with the set center value. The difference value of each individual index is then multiplied by itself. All the multiplied values ​​are summed and distributed to the total number of aggregate-dispersion composite indices. After the distribution is completed, the set discrete value is obtained. The set discrete value is determined by the distribution result of the sum of the squares of the second-order differences of all aggregate-dispersion composite indices.

[0155] The difference between the single aggregation and dispersion comprehensive index and the set central value, along with the set discrete value, are allocated. The allocated output is then substituted into the index mapping process to complete the nonlinear transformation. After the transformation, the joint trajectory outlier significance is generated. The joint trajectory outlier significance is the exclusive judgment content generated by the nonlinear outlier degree mapping process in this step.

[0156] The beneficial effects are as follows: corresponding distance content is generated based on current joint trajectory points and voltage joint trajectory points respectively, and two types of convergence and divergence are derived layer by layer. After fusing the two types of convergence and divergence to obtain a comprehensive convergence and divergence index, the mean and discrete state are extracted based on all indicators. Then, the outlier significance of the joint trajectory is generated through nonlinear mapping. The distribution state of the two types of electrical trajectories, current and voltage, is covered synchronously throughout the process. The density of point arrangement is transformed into a unified quantitative judgment content. The trajectory distribution characteristics of the two electrical dimensions are fully taken into account, eliminating the information loss caused by the evaluation of a single electrical quantity. The deviation degree is calibrated based on the overall index distribution of the section. The state measurement of all adjacent section pairs is completed with a unified standard, providing a unified and complete quantitative reference for section fault identification.

[0157] S05. Based on the outlier significance of the joint trajectory, determine the faulty feeder section of the distribution network.

[0158] In this embodiment of the invention, determining the faulty feeder section of the distribution network based on the joint trajectory outlier significance includes:

[0159] The adjacent segments are arranged from largest to smallest according to the numerical value of the outlier significance of the joint trajectory.

[0160] Based on the sorting results, the direction of current flow is traced from the power supply side to the load side along the feeder transmission path, and the feeder section located downstream of the fault current in the first adjacent section pair is identified as the faulty feeder section.

[0161] Extract the outlier saliency of the joint trajectory corresponding to each pair of adjacent segments. The joint trajectory outlier saliency is produced by the prior nonlinear outlier degree mapping. All adjacent segments are rearranged according to the arrangement order of their respective joint trajectory outlier saliencies. After rearrangement, the arrangement order of adjacent segments follows the arrangement rule of joint trajectory outlier saliency from the first position to the last position. After the rearrangement is completed, the sorting result is obtained. The sorting result is the exclusive content produced by the rearrangement of all adjacent segments in this step.

[0162] Read the feeder transmission path information stored in the distribution network in advance. The feeder transmission path information is determined by the distribution network topology adjacency relationship. The feeder transmission path records the current flow direction from the power supply side to the load side. The current flow direction is determined by the electrical connection relationship between feeder sections. Take out the adjacent section pair with the first position in the sorting result. The adjacent section pair contains two directly adjacent feeder sections. According to the current flow direction, distinguish the feeder section corresponding to the downstream direction of the fault current in the adjacent section pair. The downstream direction of the fault current is determined by the current flow direction from the power supply side to the load side.

[0163] The feeder segment corresponding to the downstream direction of the fault current identified within the first adjacent segment pair is marked as the fault feeder segment. The fault feeder segment is the exclusive object content finally determined in this step. This determination result is jointly determined based on the first adjacent segment pair of the sorting result and the current flow direction recorded in the feeder transmission path.

[0164] The physical mechanism of the judgment rule is as follows: During normal operation, the current decay and voltage drop of each section are in a stable state, the deviation sequence fluctuates little, and the joint trajectory point clustering is high; after the fault occurs, the electrical quantity of the upstream section of the fault point is less affected by the fault than the fault section itself, and the deviation amplitude of the fault section will change abruptly, resulting in a significant increase in the dispersion of the adjacent pair trajectory containing the fault section. Therefore, the adjacent pair with the highest outlier significance is the fault boundary pair; the fault current flows from the power source to the fault point, and the fault point is located in the downstream section. Therefore, the downstream section is judged as the fault section.

[0165] Multi-point fault handling: If there are two or more pairs of adjacent segments with outlier significance exceeding [a certain threshold]... If the threshold is set, then downstream judgment is performed on each group of segments that exceed the threshold, and all faulty segments are output.

[0166] Active distribution network adaptation: For feeders containing distributed generation, the power supply side and load side of each section are first identified according to the power flow direction before the fault; if the current direction of the section is reversed after the fault, the upstream and downstream directions of the section are updated synchronously to ensure that the downstream direction of the fault current always points to the fault point.

[0167] Scope of application: This method supports various faults such as metallic short circuit, grounding through transition resistor, and high resistance grounding. It is suitable for neutral point ungrounded, arc suppression coil grounded, and low resistance grounding systems.

[0168] The beneficial effect is that, based on the already generated joint trajectory outlier significance, the order of all adjacent segment pairs is rearranged to obtain the sorting result. Then, combined with the current flow direction from the power supply side to the load side recorded in the feeder transmission path, the feeder segment downstream of the fault current in the first adjacent segment pair is located as the faulty feeder segment. The segment discrimination is carried out directly using the trajectory outlier characterization content obtained in the previous steps, without the need to introduce other external reference materials. The segment location is completed along the inherent flow direction of the feeder, realizing the convergence location from the set of adjacent segment pairs to a single faulty feeder segment, simplifying the fault segment discrimination process, making full use of the deviation information obtained from the current and voltage joint trajectory processing, and ensuring that the fault segment location result is consistent with the feeder topology connection relationship.

[0169] like Figure 2 The diagram shown is a functional block diagram of a distribution network feeder fault section intelligent positioning system based on feature recognition provided in an embodiment of the present invention.

[0170] The feature-based intelligent fault location system 10 for distribution network feeder sections described in this invention can be installed in an electronic device. Depending on the functions implemented, the feature-based intelligent fault location system 10 may include a data synchronization acquisition module 11, a differential differential transformation module 12, a joint phase space reconstruction module 13, an outlier saliency determination module 14, and a fault section determination module 15. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0171] In this embodiment, the functions of each module / unit are as follows:

[0172] The data synchronization acquisition module 11 is used to synchronously acquire the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence and the last-end aligned voltage sequence of the feeder section in the distribution network.

[0173] The differential transformation module 12 is used to perform point-by-point differential transformation on the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence and the last-end aligned voltage sequence to obtain the intra-region current deviation sequence and intra-region voltage deviation sequence of the feeder section.

[0174] The joint phase space reconstruction module 13 is used to perform joint phase space reconstruction on the current deviation sequence and the voltage deviation sequence in the region respectively, so as to obtain the joint current trajectory point and the joint voltage trajectory point of the feeder section.

[0175] The outlier significance determination module 14 is used to determine the outlier significance of the joint trajectory of the feeder segment based on the current joint trajectory point and the voltage joint trajectory point.

[0176] The fault section determination module 15 is used to determine the fault feeder section of the distribution network based on the joint trajectory outlier significance.

[0177] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0178] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0179] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0180] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0181] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0182] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for intelligent location of fault sections in distribution network feeders based on feature recognition, characterized in that, The method includes: S01. Synchronously collect the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence of the feeder section in the distribution network. S02. Perform point-by-point differential transformation on the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence and the last-end aligned voltage sequence to obtain the intra-region current deviation sequence and intra-region voltage deviation sequence of the feeder section. S03. Perform joint phase space reconstruction on the current deviation sequence and the voltage deviation sequence in the region respectively to obtain the joint current trajectory point and the joint voltage trajectory point of the feeder section. S04. Determine the outlier significance of the joint trajectory of the feeder segment based on the current joint trajectory point and the voltage joint trajectory point; S05. Based on the outlier significance of the joint trajectory, determine the faulty feeder section of the distribution network.

2. The intelligent location method for fault sections of distribution network feeders based on feature recognition as described in claim 1, characterized in that, The synchronous acquisition of the head-end aligned current sequence, the end-end aligned current sequence, the head-end aligned voltage sequence, and the end-end aligned voltage sequence of the feeder section in the distribution network includes: Synchronously collect time-series current and voltage data at the beginning and end of feeder sections in the distribution network; The time-series current data and the time-series voltage data are aligned with a time reference to obtain the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence of the feeder segment.

3. The intelligent location method for fault sections of distribution network feeders based on feature recognition as described in claim 1, characterized in that, The step of performing point-by-point differential transformation on the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence to obtain the intra-region current deviation sequence and intra-region voltage deviation sequence of the feeder section includes: The point-by-point amplitude difference operation is performed on the first-end aligned current sequence and the last-end aligned current sequence to obtain the in-region current attenuation of the feeder section. The difference between the magnitude of the current attenuation in the region and the mean of the current attenuation in the region is calculated, and the absolute value is taken to obtain the current deviation sequence in the feeder section. The voltage drop within the feeder section is obtained by performing point-by-point amplitude difference calculation on the first-end aligned voltage sequence and the last-end aligned voltage sequence. The difference between the magnitude of the voltage drop in the region and the mean of the voltage drop in the region is calculated, and the absolute value is taken to obtain the voltage deviation sequence of the feeder section.

4. The intelligent location method for fault sections of distribution network feeders based on feature recognition as described in claim 2, characterized in that, The step of performing joint phase space reconstruction on the current deviation sequence and the voltage deviation sequence within the region to obtain the joint current trajectory point and the joint voltage trajectory point of the feeder segment includes: Based on the topological adjacency relationship of the distribution network, the feeder section is divided into adjacent section pairs, and each adjacent section pair consists of two directly adjacent feeder sections; Using the sampling time of the feeder section as the synchronization anchor point, the deviation amplitudes of the current deviation sequences in the middle zone of the adjacent sections are paired and arranged into ordered current binary pairs. According to the sampling sequence, the ordered current binary groups are arranged sequentially in the two-dimensional amplitude plane to obtain the current joint trajectory points of the adjacent segment pairs; Based on the synchronization anchor point, the deviation amplitudes of the voltage deviation sequences in the adjacent segments within the central region are paired and arranged into ordered voltage binary pairs. According to the sampling sequence, the ordered voltage pairs are arranged sequentially in the two-dimensional amplitude plane to obtain the voltage joint trajectory points of the adjacent segment pairs.

5. The intelligent location method for fault sections of distribution network feeders based on feature recognition as described in claim 4, characterized in that, The current joint trajectory point and the voltage joint trajectory point include: The horizontal axis of the current joint trajectory point is the deviation amplitude of the first feeder segment in the adjacent segment alignment at the current sampling time, and the vertical axis is the deviation amplitude of the second feeder segment in the adjacent segment alignment at the same sampling time. The horizontal axis of the voltage joint trajectory point is the deviation amplitude of the first feeder segment in the adjacent segment alignment at the current sampling time, and the vertical axis is the deviation amplitude of the second feeder segment in the adjacent segment alignment at the same sampling time.

6. The intelligent location method for fault sections of distribution network feeders based on feature recognition as described in claim 4, characterized in that, The step of determining the outlier significance of the joint trajectory of the feeder segment based on the joint current trajectory point and the joint voltage trajectory point includes: Using the abscissa and ordinate values ​​of the current joint trajectory points as the calculation basis, the distance between the current joint trajectory points and the origin is calculated according to the square root rule. The distance values ​​are then arranged in the order of sampling time to obtain the current distance values ​​of the adjacent segment pairs. The sequence center value of the current distance values ​​is obtained by dividing the cumulative sum of the current distance values ​​by the number of current distance values. Calculate the deviation magnitude between the current distance value and the center value of the sequence one by one; The deviation amplitude is squared and summed, and the sum of squares is divided by the number of values ​​to obtain the mean square deviation value. Then, the mean square deviation value is squared to obtain the discrete sequence value of the current distance value. Dividing the discrete value of the sequence by the center value of the sequence yields the current joint trajectory convergence / divergence degree of the adjacent segment pairs; Based on the method for obtaining the current joint trajectory convergence, the voltage joint trajectory convergence of the adjacent segment pairs is obtained. Based on the current joint trajectory convergence and divergence and the voltage joint trajectory convergence and divergence, the convergence and divergence comprehensive index of the adjacent segment pairs is calculated; Based on the aggregation-dispersion comprehensive index, the outlier significance of the joint trajectory of the adjacent segment pairs is calculated.

7. The intelligent location method for fault sections of distribution network feeders based on feature recognition as described in claim 6, characterized in that, The calculation of the combined convergence / dispersion index of the adjacent segment pairs based on the current joint trajectory convergence / dispersion and the voltage joint trajectory convergence / dispersion includes: The formula for calculating the aggregation and dispersion comprehensive index is: ; In the formula, For the first The aggregation and dispersion comprehensive index of adjacent segment pairs For the first Current joint trajectory convergence / divergence of adjacent segments For the first The convergence and divergence of voltage joint trajectories for adjacent segments.

8. The intelligent location method for fault sections of distribution network feeders based on feature recognition as described in claim 6, characterized in that, The calculation of the outlier significance of the joint trajectory of the adjacent segment pairs based on the aggregation-dispersion comprehensive index includes: The aggregation and dispersion comprehensive index is averaged and merged to obtain the set center value of the aggregation and dispersion comprehensive index; The second-order central moments of the aggregation-dispersion composite index are obtained by calculating the set discrete values ​​of the aggregation-dispersion composite index. A nonlinear outlier mapping is performed on the aggregation-dispersion composite index to obtain the joint trajectory outlier significance of the adjacent segment pairs. The formula for calculating the joint trajectory outlier significance is as follows: ; In the formula, For the first Outlier significance of joint trajectory of adjacent segment pairs For the discrete values ​​of the set, For the first The aggregation and dispersion comprehensive index of adjacent segment pairs The set center value, This is the exponentiation operator.

9. The intelligent location method for fault sections of distribution network feeders based on feature recognition as described in claim 4, characterized in that, The determination of the faulty feeder section of the distribution network based on the joint trajectory outlier significance includes: The adjacent segments are arranged from largest to smallest according to the numerical value of the outlier significance of the joint trajectory. Based on the sorting results, the direction of current flow is traced from the power supply side to the load side along the feeder transmission path, and the feeder section located downstream of the fault current in the first adjacent section pair is identified as the faulty feeder section.

10. A feature-based intelligent location system for fault sections in distribution network feeders, characterized in that, The system for implementing the feature recognition-based intelligent fault location method for distribution network feeder sections as described in claim 1 includes: The data synchronization acquisition module is used to synchronously acquire the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence, and the last-end aligned voltage sequence of the feeder section in the distribution network. The differential transformation module is used to perform point-by-point differential transformation on the first-end aligned current sequence, the last-end aligned current sequence, the first-end aligned voltage sequence and the last-end aligned voltage sequence to obtain the intra-region current deviation sequence and intra-region voltage deviation sequence of the feeder section. The joint phase space reconstruction module is used to perform joint phase space reconstruction on the current deviation sequence and the voltage deviation sequence in the region respectively, so as to obtain the joint current trajectory point and the joint voltage trajectory point of the feeder section. The outlier significance determination module is used to determine the outlier significance of the joint trajectory of the feeder segment based on the current joint trajectory point and the voltage joint trajectory point. The fault section determination module is used to determine the faulty feeder section of the distribution network based on the outlier significance of the joint trajectory.