Single-phase Grounding Fault Section Location Method for Distribution Network Based on Multi-source Data Fusion

Abstract

The present invention discloses a method for locating single-phase grounding fault sections in a distribution network based on multi-source data fusion, including: obtaining distribution network cloud master station and dispatching tele-signaling data, and respectively obtaining distribution terminal grounding, bus grounding, small current grounding line selection, arc suppression coil, and bus voltage information through data cleaning, and using any one of the distribution terminal grounding alarm signal, bus grounding, and small current grounding line selection information as the grounding judgment trigger condition; after any one of the distribution terminal grounding alarm signal, bus grounding, and small current grounding line selection information triggers the judgment of the grounding fault section, the distribution network cloud master station uses three kinds of information, namely grounding fault alarm information, zero-sequence current telemetry value, and grounding fault recording wave, to respectively judge the grounding fault section. The present invention simultaneously considers the in-station line selection result, the relationship between the neutral point grounding method and the zero-sequence current, the terminal recording wave file, and the terminal alarm information, and comprehensively obtains a grounding fault section location method with high reliability, strong practicability, and high accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution networks, and particularly to a method for locating single-phase grounding fault sections in a distribution network based on multi-source data fusion. Background Art

[0002] With the development of energy globalization, the construction and safe operation of distribution networks have attracted more and more attention. In a 10kV power system, single-phase grounding fault is one of the common fault types, and its occurrence rate accounts for more than 70% of the total system fault rate. A single-phase grounding fault not only affects the stable operation of the power system but may also trigger more serious electrical accidents, such as phase-to-phase short circuit, fire, etc. The electrical quantity information of a single-phase grounding fault mainly depends on zero-sequence voltage and zero-sequence current. The signals are much weaker than those of short circuit faults and are affected by equipment performance, equipment operation status, etc. It is difficult to accurately discriminate faults. After a single-phase grounding fault occurs, trial switching to handle the fault is still common. Problems such as large-scale power outages, fires, electric shock casualties, and equipment damage caused by inaccurate fault discrimination and untimely isolation exist to varying degrees. Therefore, in-depth research on single-phase grounding fault protection technology, improving the accuracy of locating single-phase grounding fault sections in a distribution network, and quickly isolating the fault sections are of great significance for improving the safety and reliability of the power system.

[0003] The current judgment of single-phase grounding faults in distribution networks mainly relies on in-station grounding line selection devices and feeder terminals on 10kV feeders. The current common method for locating grounding fault sections is based on the grounding alarm information of in-station line selection devices and the grounding alarm information sent from feeder terminals to the master station. By combining the alarm information of the two and comparing the line topology diagram, the fault location section can be obtained. However, due to the limited accuracy of in-station grounding line selection devices, the line selection accuracy of early devices was not high, and limited by the transformation progress, in-station line selection devices are not installed in some substations. In addition, since there are many manufacturers of feeder terminal equipment put into operation in the grid, the grounding fault judgment algorithms of each unit are different, and the grounding protection setting values of the terminals are not configured reasonably, and the upper and lower switches are mismatched, resulting in different credibility of the grounding alarm signals sent from the terminals to the master station. At present, the main grid and the distribution automation system operate separately, and the information is not integrated, reducing the efficiency of grounding fault handling. The dispatching mainly selects and switches based on the results of the main grid line selection device or tries to switch according to the sequence list, and the fault finding time is long; grass-roots personnel need to understand the grounding information by phone notification or checking the system interface, and the information acquisition is fragmented, which is not conducive to the rapid handling of grounding faults. Therefore, there is still no method with high reliability and strong practicability for locating single-phase grounding fault sections in distribution networks at present.

[0004] To improve the accuracy of single-phase grounding fault section location in the distribution network and overcome the problems of insufficient reliability in the discrimination accuracy of single-phase grounding faults by in-station line selection devices and feeder terminals, the distribution network fault section location method needs to comprehensively consider the line selection results of in-station line selection devices, the telemetry values of switch zero-sequence currents, the grounding alarm information of feeder terminals, the grounding fault recording waves of feeder terminals, etc. The invention patent publication document CN 118275820 A mentions that by comprehensively comparing four types of information, namely the terminal grounding alarm information, the switch zero-sequence current, the feeder zero-sequence voltage, and the terminal recording wave file, the single-phase grounding fault section information is obtained. However, when determining the fault section by comparing the magnitudes of zero-sequence currents in the invention patent publication document CN 118275820 A, the grounding method of the main transformer neutral point is not considered, and only based on the change amount of the real-time zero-sequence current value and the historical steady-state current value of the switch for sorting, it is determined that the grounding fault point is behind the switch with the largest change amount of zero-sequence current. In an ungrounded system, when a single-phase grounding fault occurs, if the zero-sequence current is correctly collected, it can be determined that the grounding fault point is behind the switch with the largest change amount of zero-sequence current. However, in a system with arc suppression coil grounded at the neutral point, due to the over-compensation effect of the arc suppression coil, the zero-sequence current of the switch decreases successively to the subsequent switches after compensation. Therefore, the fault point is not necessarily located behind the switch with the largest change amount of zero-sequence current.

[0005] In addition, in the invention patent publication document CN 118275820 A, the section where the grounding fault occurs is determined by comparing the similarity of the switch recording wave files. However, when arc grounding occurs, a terminal will record multiple groups of waveforms in a short period of time. Comparing the similarity of the recording waveforms of multiple terminals at the same moment may be limited by the time synchronization accuracy, resulting in misjudgment of the fault section. Summary of the Invention

[0006] Aiming at the deficiencies of the existing distribution network single-phase grounding fault section location technology, the purpose of the present invention is to provide a distribution network single-phase grounding fault section location method based on multi-source data fusion, aiming to solve the problems that the grounding method of the neutral point and the magnitude of the switch zero-sequence current are not fully considered and the terminal fault recording waves are not fully utilized in the existing technology.

[0007] To achieve the above purpose, the present invention provides the following technical solution: A distribution network single-phase grounding fault section location method based on multi-source data fusion, the location method includes the following steps:

[0008] Obtain the distribution cloud master station and dispatching telecommunication data, and through data cleaning, respectively obtain the grounding of distribution terminals, bus grounding, small current grounding line selection, arc suppression coil, and bus voltage information, and use any one of the distribution terminal grounding alarm signal, bus grounding, and small current grounding line selection information as the grounding discrimination trigger condition;

[0009] After any information such as the grounding alarm signal of the distribution terminal, bus grounding, or small current grounding line selection triggers the judgment of the grounding fault section, the distribution cloud master station uses three types of information, namely the grounding fault alarm information, zero-sequence current telemetry value, and grounding fault recording wave, to respectively judge the grounding fault section; for the system with the neutral point grounded through an arc suppression coil, the grounding fault alarm information and the grounding fault recording wave are used for the positioning and judgment of the grounding fault section, and the judgment result of the grounding fault recording wave by the distribution cloud master station is used as the main basis, and the grounding alarm information of the distribution terminal is used as an auxiliary criterion; if there is a conflict between the grounding fault information and the judgment result of the distribution cloud master station, the waveform analysis result of the distribution cloud master station shall prevail; for the ungrounded system, the three types of information, namely the grounding fault alarm information, zero-sequence current telemetry value, and grounding fault recording wave, are used to locate the section at the same time; for the system grounded through a small resistance, the section is located relying on the two types of information of the grounding fault alarm and the zero-sequence current telemetry value; among them, for the ungrounded system, the zero-sequence current telemetry value method and the grounding fault recording wave judgment are used as the main basis, and the confidence level of the zero-sequence current telemetry value method is higher than that of the grounding fault recording wave method.

[0010] In a preferred embodiment, using the zero-sequence current telemetry value to judge the grounding fault section includes the following steps: After triggering the judgment of the grounding fault section, the distribution cloud master station deploys an algorithm to extract information such as the relationship between the bus and the outgoing line, the switch topology relationship, the wire type, and the wire diameter and length, and calculates the zero-sequence current setting value when a grounding fault occurs in the switch area. I set_n , where n is the switch number, and it is compared with the zero-sequence current value of the recording wave sent by the distribution terminal. I 0n When I set_n is less than I 0n , mark that a grounding fault has occurred in this switch area; when I set_n is greater than I 0n , mark that a grounding fault has occurred outside this switch area; after triggering the judgment of the grounding fault section, the distribution cloud master station also takes the substation as a container, collects the zero-sequence current telemetry values of all switches, and obtains the grounding fault section by comparing the magnitudes of the zero-sequence current telemetry values at the fault moment and the magnitudes of the changes in the zero-sequence current telemetry values.

[0011] In a preferred embodiment, the calculation of the zero-sequence current setting value includes the following steps:

[0012] After obtaining the relationship between the bus and the outgoing line, the switch topology relationship, the line type, and the line diameter and line length information, calculate the zero-sequence current setting value of the zero-sequence current protection for each line. The zero-sequence current setting value of the zero-sequence current protection is set to be less than the sum of the capacitive currents of all lines on the same bus on the power supply side of the switch and greater than the sum of the capacitive currents of the lines on the load side of the switch. When a fault occurs on each automatic switch on the line, the zero-sequence current setting value I set The calculation formula is:

[0013] ,

[0014] wherein, L JK1 is the total length of the overhead lines of other 10kV feeder outgoing lines on the same bus, L JK2 is the total length of the overhead lines from the switch to the bus, L JK is the total length of the overhead lines on the power supply side of the switch and other feeder lines on the same bus, L 1 is the total length of the overhead lines on the load side of the switch, in kilometers, S 1 is the capacitive current value per kilometer of the overhead line; L DL1 is the total length of the cable lines of other 10kV feeder outgoing lines on the same bus, L DL2 is the total length of the cable lines from the switch to the bus, L DL is the total length of the cable lines on the power supply side of the switch and other feeder lines on the same bus, L 2 is the total length of the cable lines on the load side of the switch, in kilometers, S 2 is the capacitive current value per kilometer of the cable line; M is the margin coefficient; Calculate the capacitive current values of all lines on the same bus on the power supply side of the switch when a fault occurs on each automatic switch on the line I set1 and the capacitive current value of the line on the load side of the switch I set2 , and the zero-sequence current setting value I set is between I set1 and I set2 .

[0015] In a preferred embodiment, the steps for judging the grounding fault section by using the grounding fault recording wave are as follows: First, by comparing the correlation of the first half-wave waveforms of the zero-sequence currents of the outgoing line switches of each feeder, the fault line with a large deviation in correlation is the fault line where the grounding occurs. Then, according to the switch topology information, compare the deviation degree of the first half-wave waveforms of the zero-sequence currents of two adjacent switches. The grounding fault section is located between the two switches with the largest deviation degree.

[0016] In a preferred embodiment, after triggering the judgment of the grounded fault section, taking the substation as a container, collecting the switch grounded fault recording wave files, and recording the fault occurrence time as t 0, intercept t the first half-wave waveform of the zero-sequence current after time 0, perform non-linear fitting on the discrete points of each moment of the first half-wave waveform of the zero-sequence current, and the Levenberg-Marquardt algorithm is used for fitting. The fitting formula is as follows:

[0017] ,

[0018] In the formula, is the fitting curve of the first half-wave waveform of the zero-sequence current, I 0_max is the maximum amplitude of the first half-wave waveform of the zero-sequence current, t is the moment ，t max is the time value corresponding to the moment with the maximum amplitude, α , β is the fitting constant parameter;

[0019] After obtaining the fitting curve of the first half-wave waveform of the zero-sequence current, calculate the derivative of the fitting curve with respect to time , as shown in the following formula:

[0020] ,

[0021] In the formula 、 are the fitting zero-sequence current values at times t2 and t1 respectively;

[0022] Using the time derivative of the fitting curve of the first half-wave waveform of the switch zero-sequence current, calculate the Euclidean distance of the time derivatives of the first half-wave waveforms of the zero-sequence currents of two switches, as shown in the following formula:

[0023] ,

[0024] In the formula, represents the Euclidean distance of the time derivatives of the first half-wave waveforms of the zero-sequence currents of switch n and switch m and at time t, and T represents the duration of the first half-wave; is the sum of the Euclidean distances of the time derivatives of the first half-wave waveforms of the zero-sequence currents of two switches at each moment, and it is used as the correlation coefficient of the first half-wave waveforms of the zero-sequence currents of two switches; The larger it is, the lower the correlation; The smaller it is, the higher the correlation; when a grounded fault occurs, first calculate the correlation coefficient of the first half-wave of the zero-sequence current of each feeder outgoing switch or the first switch on the line , through pairwise comparison and clustering analysis, the feeder where the switch with the largest correlation coefficient between its zero-sequence current waveform's first half-wave and that of other switches is the faulty line; after determining the faulty line, through the switch topological relationship, compare the correlation coefficients of the first half-waves of the zero-sequence currents of the switches , a small correlation coefficient indicates that the first half-wave waveforms of the zero-sequence currents of the two switches are similar; if the fault point is between two switches, the correlation coefficient of the first half-waves of the zero-sequence currents of these two switches is much larger than that of the first half-waves of the zero-sequence currents of other adjacent switches; if the fault point is behind the last-stage switch and there is no branch line between this switch and the bus, the correlation coefficients of the first half-waves of the zero-sequence currents of the switches are all close, and it is judged that the grounding fault section is at the last stage.

[0025] In a preferred embodiment, using the grounding fault warning information to judge the grounding fault section includes the following steps:

[0026] When receiving the grounding warning information of the distribution terminal, first, taking the substation as a container, collect the grounding warning information of the distribution terminal, and correspond the distribution terminal that sends the grounding fault warning to the 10kV feeder, and verify whether the distribution terminals reporting the grounding fault remote signal are on the same feeder; if the distribution terminals reporting the grounding fault remote signal are on the same line and only one distribution terminal reports the grounding fault warning information, it is judged that the grounding fault section is between the switch reporting the grounding fault warning and its next-level switch; if more than one distribution terminal reports the grounding fault warning information, it is judged that the grounding fault section is between the last-stage switch reporting the grounding fault warning and its next-level switch.

[0027] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0028] 1. The present invention obtains the distribution cloud master station and dispatching remote signal data through KAFKA, and through data cleaning, respectively obtains the distribution terminal grounding, bus grounding, small current grounding line selection, arc suppression coil, and bus voltage information, and uses any one of the distribution terminal grounding warning signal, bus grounding, and small current grounding line selection information as the grounding judgment trigger condition, avoiding problems such as missed reporting of grounding faults caused by inconsistent relevant equipment or setting of fixed values in the main and distribution networks.

[0029] 2. After the present invention obtains the relationship between the bus and the outgoing line, the switch topological relationship, the line type, the line diameter and line length information through the business middleware, the distribution cloud master station uniformly calculates the zero-sequence current setting value of each line's automatic switch in the neutral-point non-grounding system when a fault occurs I set , I set Compare it with the zero-sequence current value at the fault moment to determine the grounding fault section, avoiding problems such as misreporting and missed reporting of grounding fault information caused by incorrect setting of the in-service fixed value, and greatly reducing the work pressure of a large number of grass-roots personnel to modify the in-service distribution terminal fixed value.

[0030] 3. The present invention deploys a grounding fault section location algorithm based on the correlation of the first half-wave waveform of zero-sequence current in the distribution cloud master station. By analyzing the similarity of the first half-wave waveform of zero-sequence current, the faulty line and the fault section are determined, unifying the credibility of the grounding fault recording and analysis of the fault section, and avoiding the problem of inconsistent grounding judgment credibility caused by different terminal devices and different program versions.

[0031] 4. The present invention uses three kinds of information, namely, the grounding alarm signal of the distribution terminal, the telemetry value of zero-sequence current, and the grounding fault recording, as the basis for judging the grounding fault section. Select a suitable section location method according to the system neutral grounding method, and clarify the primary and secondary of using multiple information for judgment. Fully consider factors such as the failure and absence of information upload, and use the substation as a container to push the comprehensively analyzed grounding fault section through the enterprise-level message platform. Description of the Drawings

[0032] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings.

[0033] Figure 1 It is a schematic circuit topology diagram of the embodiment of the present invention;

[0034] Figure 2 It is a schematic diagram for calculating the setting value of the zero-sequence current of the switch in the ungrounded neutral system of the present invention;

[0035] Figure 3 It is a schematic diagram for locating the grounding fault section by using the grounding alarm information of the distribution terminal;

[0036] Figure 4 It is a schematic diagram for locating the grounding fault section by using the telemetry value of zero-sequence current;

[0037] Figure 5 It is a schematic diagram for locating the grounding fault section by using the grounding fault recording;

[0038] Figure 6 It is a schematic diagram of the grounding fault section location information push logic proposed by the present invention. Detailed Embodiment

[0039] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0040] This embodiment provides a method for locating single-phase grounding fault sections in a distribution network based on multi-source data fusion. The location method needs to apply feeder switches, distribution terminals (FTUs, DTUs), and a distribution network cloud master station. The distribution terminals in this embodiment have the functions of recording and uploading zero-sequence current and grounding fault recording waveforms, and have a grounding fault judgment algorithm, and can generate and upload grounding fault alarm signals according to preset values. The distribution network cloud master station has functions such as accessing substation information, bus information, switch information, grounding fault recording waveforms of distribution terminals, and homologous diagram models, and can obtain real-time telemetry and telecontrol information of distribution terminals and access grounding fault recording waveforms uploaded by distribution terminals.

[0041] The implementation of the present invention needs to obtain the single-line diagram of the line and the special topic diagram of the substation in the business middle platform through a data interface, and parse to obtain the switch topology relationship, the lengths of cables and overhead lines, the relationship between the bus and the 10 kV feeder line, and the relationship between buses. Through KAFKA, obtain the telecontrol data of the distribution network cloud master station and the dispatching, and through data cleaning, respectively obtain the grounding of distribution terminals, bus grounding, small current grounding line selection, arc suppression coil, and bus voltage information, and use any information of the grounding alarm signal of the distribution terminal, bus grounding, and small current grounding line selection as the grounding judgment trigger condition.

[0042] To reduce the basic operation and maintenance pressure, the present invention takes the grounding alarm information of the distribution terminal as a supplement, and deploys a method for judging single-phase grounding fault sections based on the line and switch topology relationship and the correlation of zero-sequence current waveforms in the distribution network cloud master station. As Figure 4 shown, the specific implementation method is as follows:

[0043] After obtaining the relationship between the bus and the outgoing line, the switch topology relationship, the line type, the wire diameter and the line length through the business middle platform, calculate the zero-sequence current setting value of the zero-sequence current protection of each line automation switch. The zero-sequence current setting value of the zero-sequence current protection is set to be less than the sum of the capacitive currents of all lines on the same bus on the power supply side of the switch and greater than the sum of the capacitive currents of the lines on the load side of the switch. Figure 2 The figure shows a schematic diagram of calculating the zero-sequence current setting value of a switch in an isolated neutral system. The zero-sequence current setting value of each automation switch on the line when a fault occurs I set The calculation expression is:

[0044] ,

[0045] Among them, L JK1 is the total length of the overhead lines of other 10kV feeder lines on the same busbar, L JK2 is the total length of the overhead lines from the switch to the busbar, L JK is the total length of the overhead lines on the power supply side of the switch and other feeder lines on the same busbar, L 1 is the total length of the overhead lines on the load side of the switch, in kilometers, S 1 is the capacitive current value per kilometer of the overhead lines; L DL1 is the total length of the cable lines of other 10kV feeder lines on the same busbar, L DL2 is the total length of the cable lines from the switch to the busbar, L DL is the total length of the cable lines on the power supply side of the switch and other feeder lines on the same busbar, L 2 is the total length of the cable lines on the load side of the switch, in kilometers, S 2 is the capacitive current value per kilometer of the cable lines; M is the margin coefficient, taking 1.2 - 1.5. Through the above formula, the capacitive current values of all lines on the same busbar on the power supply side of each automatic switch on the line during a fault can be calculated I set1 and the capacitive current value of the line on the load side of the switch I set2 . The setting value of the zero-sequence current I set is between I set1 and I set2 . If it is an ungrounded neutral system, after the judgment of the grounding fault section is triggered, the distribution cloud master station deploys an algorithm to extract information such as the line type, wire diameter, switch topology, and location of the feeder lines of the busbar, and calculates the setting value of the zero-sequence current when a grounding fault occurs in the switch area I set_n (n is the switch number), and compares it with the zero-sequence current value I 0n sent by the distribution terminal for wave recording. When I set_n is less than I 0n , mark that a grounding fault has occurred in this switch area; when I set_n is greater than I 0nWhen a ground fault occurs outside this switch area, the position relationship between the ground fault point and the switch can be obtained. In addition, after triggering the judgment of the ground fault section, the distribution cloud master station can also use the substation as a container to collect the telemetry values of the zero-sequence current of all switches, and obtain the ground fault section by comparing the magnitudes of the zero-sequence current telemetry values at this moment. As Figure 1 shown, when a ground fault occurs between Switch 2 and Switch 3 on Feeder 3, the setting values of the zero-sequence current of Switch 1 and Switch 2 I set_1 、 I set_2 are respectively less than I 01 and I 02 , the setting values of the zero-sequence current of other switches I set_n are all greater than I 0n , the zero-sequence current telemetry values of Switch 1 and Switch 2 are much larger than those of other switches, and the change value Δ I of the zero-sequence current telemetry values of other switches before and after the ground fault is much smaller than that of Switch 1 and Switch 2. Combining with the switch topology relationship, it can be obtained that the ground fault section is located between Switch 2 and Switch 3. Due to the compensation effect of the arc suppression coil on the grounding zero-sequence current, it is impossible to rely on the magnitude of the zero-sequence current telemetry value of the distribution terminal to distinguish whether this switch is in the ground fault area. Therefore, the method based on the magnitude of the zero-sequence current is not used to judge the ground fault section in the neutral point grounded through an arc suppression coil system. By online comparing the recorded wave zero-sequence current I 0 with the setting value of the zero-sequence current when a ground fault occurs in the switch area I set , as well as comparing the magnitude and change value Δ I of the zero-sequence current telemetry value, the ground fault section of the neutral point ungrounded system can be obtained, which greatly reduces the working pressure of grass-roots operation and maintenance personnel to modify the setting values of in-service distribution terminals.

[0046] If it is a neutral point grounded through an arc suppression coil system, first, by comparing the correlation of the first half-wave waveforms of the zero-sequence current of each feeder outgoing switch, the fault line with a large deviation in correlation is the grounded fault line. Then, according to the switch topology information, compare the deviation degree of the first half-wave waveforms of the zero-sequence current of adjacent two switches, and the ground fault section is located between the two switches with the largest deviation degree. As Figure 5 shown, the specific implementation method is as follows:

[0047] After triggering the judgment of the ground fault section, use the substation as a container to collect the switch ground fault recording files, record the fault moment as t 0, and intercept tZero-sequence current first half-wave waveform after the 0 moment. To avoid the influence brought by sampling mutation points or jitter errors, non-linear fitting is performed on the discrete points at each moment of the zero-sequence current first half-wave waveform. The Levenberg-Marquardt algorithm is used for fitting, and the fitting formula is as follows:

[0048] ,

[0049] In the formula, is the fitting curve of the zero-sequence current first half-wave waveform, I 0_max is the maximum amplitude of the zero-sequence current first half-wave waveform, t is the moment ，t max is the time value corresponding to the moment with the maximum amplitude, α , β are the fitting constant parameters;

[0050] After obtaining the fitting curve of the zero-sequence current first half-wave waveform, calculate the derivative of the fitting curve with respect to time , as shown in the following formula:

[0051] ,

[0052] In the formula 、 are the fitted zero-sequence current values at times t2 and t1 respectively;

[0053] Using the time derivative of the fitting curve of the zero-sequence current first half-wave waveform of the switch, calculate the Euclidean distance between the time derivatives of the zero-sequence current first half-waves of two switches, as shown in the following formula:

[0054] ,

[0055] In the formula, represents the Euclidean distance between the time derivatives of the zero-sequence current first half-waves of switch n and switch m and at time t, T represents the duration of the first half-wave; is the sum of the Euclidean distances between the time derivatives of the zero-sequence current first half-waves of two switches at each moment, and it is used as the correlation coefficient between the zero-sequence current first half-waves of two switches; The larger it is, the lower the correlation; The smaller it is, the higher the correlation; When a ground fault occurs, first calculate the correlation coefficient of the zero-sequence current first half-wave of each feeder outgoing switch or the first switch on the line. Through pairwise comparison and cluster analysis, the feeder where the switch with the largest correlation coefficient with the zero-sequence current waveform first half-wave of other switches is located is the fault line; After determining the fault line, through the switch topology relationship, compare the correlation coefficients of the zero-sequence current first half-waves of the switches , a small correlation coefficient indicates that the first half-wave waveforms of the zero-sequence currents of the two switches are similar; if the fault point is located between the two switches, the correlation coefficient of the first half-wave of the zero-sequence currents of these two switches is much larger than that of the first half-wave of the zero-sequence currents of other adjacent switches; if the fault point is located behind the last-stage switch and there is no branch line between this switch and the bus, the correlation coefficients of the first half-wave of the zero-sequence currents of the switches are close, and it is judged that the grounding fault section is at the last stage. For example Figure 1 As shown, if a grounding fault occurs between switch 2 and switch 3 on feeder 2, then and is much larger than , the similarity of the first half-wave waveforms of the zero-sequence currents of switch 2 and switch 3 is low, and it is judged that the grounding fault is located between switch 2 and switch 3; if a grounding fault occurs behind switch 3 on feeder 2, the correlation coefficient between the switches is small and close to 0. At this time, it is judged that the grounding fault point is located at the last section, that is, behind switch 3. By comparing the correlation coefficients of the first half-wave of the zero-sequence currents, the grounding fault section of the neutral point grounded through an arc suppression coil system can be determined, avoiding the problem of insufficient judgment accuracy caused by the traditional algorithm considering the magnitude and phase of the zero-sequence voltage and zero-sequence current at the same time.

[0056] Figure 3 The figure shows a schematic diagram of the relevant logic flow for locating the fault section based on the grounding alarm information of the distribution terminal. When the grounding alarm information of the distribution terminal is received, first, taking the substation as a container, collect the grounding alarm information of the distribution terminal, and correspond the distribution terminal that sends the grounding fault alarm to the 10 kV feeder, and verify whether the distribution terminal reporting the grounding fault remote signal is on the same feeder. If the distribution terminals reporting the grounding fault remote signal are on the same line and only one distribution terminal reports the grounding fault information, then use this method to judge that the grounding fault section is located between this switch and its next-level switch; if more than one distribution terminal reports the grounding fault information, then use this method to judge that the grounding fault section is located between the last-stage switch reporting the grounding fault alarm and its next-level switch. For example, if the distribution terminal of switch 3 on feeder 1 sends the grounding fault alarm information, then it is judged that the grounding fault section is located behind switch 3; if the distribution terminals of switch 1 and switch 2 send the grounding fault alarm information and the distribution terminal of switch 3 does not send it, then the grounding fault section is located in the section between switch 2 and switch 3. If the distribution terminals sending the grounding alarm are not on the same 10 kV feeder, then judge the grounding fault sections of each line separately according to the above method to avoid the risk of missed reporting of multi-point grounding faults.

[0057] After any of the information such as the grounding alarm signal of the distribution terminal, bus grounding, and small current grounding line selection triggers the judgment of the grounding fault section, the distribution cloud master station uses three types of information, namely the grounding fault alarm information, zero-sequence current telemetry value, and grounding fault recording wave, to respectively judge the grounding fault section. For the system with the neutral point grounded through an arc suppression coil, the grounding fault alarm information and the grounding fault recording wave are used for the positioning and judgment of the grounding fault section, and the judgment result of the grounding fault recording wave by the distribution cloud master station is used as the main basis, and the grounding alarm information of the distribution terminal is used as an auxiliary criterion. When the judgment algorithm deployed by the distribution cloud master station can accurately locate the smallest fault section relying on the grounding fault recording wave, the final judgment shall be based on this result; when the grounding fault recording wave file fails to be uploaded or is missing, etc., the grounding fault alarm information shall be relied on as an auxiliary criterion to assist in making a decision on the grounding fault section location; if there is a conflict between the grounding fault information and the judgment result of the distribution cloud master station, the waveform analysis result of the distribution cloud master station shall prevail. For the ungrounded neutral system, the three types of information, namely the grounding fault alarm information, zero-sequence current telemetry value, and grounding fault recording wave, can be used to locate the section; for the system grounded through a small resistance, the grounding fault alarm and zero-sequence current telemetry value can be relied on to locate the section. Among them, for the ungrounded neutral system, the zero-sequence current telemetry value method and the grounding fault recording wave judgment are used as the main basis, and the confidence level of the zero-sequence current value method is higher than that of the grounding fault recording wave method. For example, when Figure 1 as shown in the grounding fault occurs on feeder 1, relying on the zero-sequence current values of the distribution terminals of switch 1 and switch 3 to analyze that the grounding fault section is between switch 1 and switch 3. If the distribution cloud master station judges that the rear end of switch 2 is grounded through the grounding fault recording wave of the distribution terminal of switch 2, it is given that the grounding fault section is between switch 2 and switch 3; if it is judged that the rear end of switch 4 is grounded through the grounding fault recording wave of the distribution terminal of switch 4, the judgment result of the zero-sequence current value shall prevail, and the grounding fault section is located between switch 1 and switch 3, skipping the smallest fault section location algorithm. When the grounding fault recording wave file fails to be uploaded, is missing, etc., the grounding alarm information is used as an auxiliary basis, which is the same as the above system with the neutral point grounded through an arc suppression coil. If the grounding fault information can help locate the smallest fault section and does not conflict with the judgment results of the grounding fault recording wave and zero-sequence current telemetry value, the grounding fault information is valid and can assist in locating the fault section; if there is a conflict, the grounding fault information is invalid and there is no need to give the smallest grounding fault section.

[0058] To sum up, the three grounding fault section judgment methods can be used to locate the smallest grounding fault section. If there is a conflict in the information, the method with a higher confidence level as described above shall prevail, and the smallest fault section location result can be abandoned to prevent incorrect section location and increase the work pressure of grass-roots operation and maintenance personnel. As Figure 6 shown, after the distribution cloud master station locates the grounding fault section, taking the substation as a container, information can be pushed to the specified group through the enterprise-level message push platform to assist grass-roots operation and maintenance personnel to find the fault point more quickly and efficiently, isolate the fault section, and restore power supply.

[0059] In the description of this specification, the descriptions referring to terms such as "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0060] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not elaborate all the details and do not limit the present invention to only the specific implementation manners. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A single-phase grounding fault section location method for a distribution network based on multi-source data fusion, characterized in that: The positioning method includes the following steps: Obtain the distribution network cloud master station and dispatching tele-signaling data. Through data cleaning, obtain the information of distribution terminal grounding, bus grounding, small current grounding line selection, arc suppression coil, and bus voltage respectively, and use any one of the distribution terminal grounding alarm signal, bus grounding, and small current grounding line selection information as the grounding judgment trigger condition; After any one of the distribution terminal grounding alarm signal, bus grounding, and small current grounding line selection information triggers the judgment of the grounding fault section, the distribution network cloud master station uses three kinds of information, namely, grounding fault alarm information, zero-sequence current telemetry value, and grounding fault recording wave, to judge the grounding fault section respectively; for the system with neutral point grounded through arc suppression coil, use the grounding fault alarm information and grounding fault recording wave to locate and judge the grounding fault section, and take the judgment result of the distribution network cloud master station on the grounding fault recording wave as the main basis, and the distribution terminal grounding alarm information as the auxiliary criterion; if the grounding fault information conflicts with the judgment result of the distribution network cloud master station, take the waveform analysis result of the distribution network cloud master station as the standard; for the ungrounded neutral system, use the grounding fault alarm information, zero-sequence current telemetry value, and grounding fault recording wave to locate the section at the same time; the system grounded through small resistance relies on the grounding fault alarm and zero-sequence current telemetry value to locate the section; among them, the ungrounded neutral system takes the zero-sequence current telemetry value method and grounding fault recording wave judgment as the main basis, and the confidence level of the zero-sequence current telemetry value method is higher than that of the grounding fault recording wave method; Using the grounding fault recording wave to judge the grounding fault section includes the following steps: First, calculate the correlation coefficient of the first half-wave of the zero-sequence current of each feeder outgoing switch or the first switch on the line. Through pairwise comparison and clustering analysis, the feeder where the switch with the largest correlation coefficient with the first half-wave of the zero-sequence current waveform of other switches is located is the fault line. After determining the fault line, according to the switch topology information, compare the correlation coefficients of the first half-waves of the zero-sequence current waveforms of two adjacent switches, and the grounding fault section is located between the two switches with the largest correlation coefficient.

2. The single-phase grounding fault section location method for a distribution network based on multi-source data fusion according to claim 1, wherein: Judging the section of the grounding fault by using the telemetered zero-sequence current value includes the following steps: After triggering the judgment of the grounding fault section, the distribution cloud master station deploys an algorithm to extract the relationship between the bus and the outgoing line, the switch topology relationship, the line type, the line diameter and the line length information, and calculates the setting value of the zero-sequence current when a grounding fault occurs in the switch area. I set_n , where n is the switch number, and it is compared with the zero-sequence current value of the oscillogram sent by the distribution terminal. I 0n When I set_n is less than I 0n , it is marked that a grounding fault occurs in this switch area; when I set_n is greater than I 0n , it is marked that a grounding fault occurs outside this switch area; After triggering the judgment of the grounding fault section, the distribution cloud master station also takes the substation as a container to collect the telemetered zero-sequence current values of all switches, and obtains the grounding fault section by comparing the magnitudes of the telemetered zero-sequence current values at the fault moment and the change values of the telemetered zero-sequence current values.

3. The single-phase grounding fault section location method for a distribution network based on multi-source data fusion according to claim 2, characterized in that: The calculation of the zero-sequence current setting value includes the following steps: After obtaining the relationship between the bus and the outgoing line, the switch topology relationship, the line type, and the line diameter and length information, calculate the zero-sequence current setting value of the zero-sequence current protection for each line's automatic switch. The zero-sequence current setting value of the zero-sequence current protection is set to be less than the sum of the capacitive currents of all lines on the same bus on the source side of the switch and greater than the sum of the capacitive currents of the lines on the load side of the switch. The zero-sequence current setting value for each automatic switch on the line during a fault I set The calculation formula is: ， ， ， ， ， in, L JK1 The total length of overhead lines of other 10kV feeder lines on the same busbar, L JK2 is the total length of the overhead line from the switch to the busbar, L JK is the total length of overhead lines on the switching power supply side and other feeder lines on the same busbar, L 1 is the total length of overhead lines on the load side of the switch, in kilometers. S 1 is the capacitance current value per kilometer of overhead line; L DL1 The total length of other 10kV feeder cables on the same busbar. L DL2 is the total length of the cable from the switch to the busbar, L DL is the total length of the switch power supply side and other feeder cables on the same busbar, L 2 is the total length of the cable on the load side of the switch, in kilometers. S 2 is the capacitance current value per kilometer of the cable line; M is the margin coefficient; the capacitance current value of all lines on the same busbar on the switch power supply side of each automatic switch on the line when a fault occurs is calculated I set1 and the line capacitance current value on the load side of the switch I set2 , zero sequence current setting value I set Between I set1 and I set2 between.

4. The single-phase grounding fault section location method for a distribution network based on multi-source data fusion according to claim 1, characterized in that: After triggering the judgment of the grounding fault section, taking the substation as a container, collect the switch grounding fault recording files, and record the fault occurrence time as t 0, intercept t the first half-wave waveform of the zero-sequence current after time 0, perform non-linear fitting on the discrete points at each time of the first half-wave waveform of the zero-sequence current, and the Levenberg-Marquardt algorithm is used for fitting. The fitting formula is as follows: ， In the formula, is the fitting curve of the first half-wave waveform of the zero-sequence current, I 0_max is the maximum amplitude of the first half-wave waveform of the zero-sequence current, t is the moment ，t max is the time value corresponding to the moment with the maximum amplitude, α , β is the fitting constant parameter; After obtaining the fitting curve of the first half-wave waveform of the zero-sequence current, calculate the derivative of the fitting curve with respect to time , as shown in the following formula: ， wherein 、 are the fitted zero-sequence current values at times t2 and t1, respectively; Use the time derivative of the fitting curve of the first half-wave of the zero-sequence current of the switch to calculate the Euclidean distance of the time derivatives of the first half-waves of the zero-sequence currents of two switches, as shown in the following formula: ， ， In the formula, represents the time derivative of the first half-wave of the zero-sequence current of switch n and switch m and the Euclidean distance at time t, where T represents the duration of the first half-wave; is the sum of the Euclidean distances of the time derivatives of the first half-waves of the zero-sequence currents of the two switches, which is used as the correlation coefficient of the first half-waves of the zero-sequence currents of the two switches; The larger it is, the lower the correlation; The smaller it is, the higher the correlation; When a ground fault occurs, first calculate the correlation coefficient of the first half-wave of the zero-sequence current of each feeder outgoing switch or the first switch on the line , through pairwise comparison and cluster analysis, the feeder where the switch with the largest correlation coefficient of the first half-wave of the zero-sequence current waveform with other switches is located is the faulty line; after determining the faulty line, through the switch topology relationship, compare the correlation coefficient of the first half-wave of the zero-sequence current of the switches , a small correlation coefficient indicates that the first half-wave waveforms of the zero-sequence currents of the two switches are similar; If the fault point is located between two switches, the correlation coefficient of the first half-waves of the zero-sequence currents of these two switches is much larger than that of the first half-waves of the zero-sequence currents of other adjacent switches; If the fault point is located behind the last-stage switch and there is no branch line between this switch and the bus, the correlation coefficients of the first half-waves of the zero-sequence currents of the switches are all close, and it is judged that the grounding fault section is at the last stage.

5. The single-phase grounding fault section location method for a distribution network based on multi-source data fusion according to claim 1, characterized in that: Using the grounding fault alarm information to judge the grounding fault section includes the following steps: When receiving the grounding alarm information of the distribution terminal, first, taking the substation as a container, collect the grounding alarm information of the distribution terminal, and correspond the distribution terminal that sends the grounding fault alarm with the 10kV feeder to verify whether the distribution terminals reporting the grounding fault remote signal are on the same feeder; if the distribution terminals reporting the grounding fault remote signal are on the same line and only one distribution terminal reports the grounding fault alarm information, then judge that the grounding fault section is between the switch reporting the grounding fault alarm and its next-level switch; if more than one distribution terminal reports the grounding fault alarm information, then judge that the grounding fault section is between the last-level switch reporting the grounding fault alarm and its next-level switch.

Images