A method and system for fault location in distribution networks based on topology features

Abstract

This invention belongs to the field of power fault location technology. It provides a method and system for fault location in distribution networks based on topological features. The method statistically analyzes the cause types of each fault sub-region, constructs multiple anomaly sequence and main anomaly influence sequence, identifies coexisting anomaly points, and analyzes the coexisting anomaly points in multiple historical analysis periods to obtain anomaly tracing base points. By analyzing the anomaly tracing base points, fault tracing values ​​are obtained, and the fault source type is located. This not only enables more accurate determination of the fault source area, but also helps to deepen the understanding of the fault propagation mechanism in the distribution network, avoids tracing deviations caused by the dispersion and interference of fault signals, and greatly improves the accuracy and efficiency of fault source tracing.

Description

Technical Field

[0001] This invention belongs to the field of power fault location technology, specifically a method and system for fault location in distribution networks based on topological features. Background Technology

[0002] With the continuous expansion and increasing complexity of power systems, the stable operation of distribution networks has become crucial for ensuring electricity supply for social production and daily life. During operation, distribution networks frequently experience various faults due to factors such as equipment aging, external damage, and natural disasters. Quickly and accurately locating the source of the fault is of great significance for shortening power outage time and reducing economic losses.

[0003] In existing technologies, the identification of fault sub-regions in distribution networks often employs threshold-based judgment methods. This means that when electrical quantities such as current and voltage exceed preset thresholds, a fault area is identified. However, this method ignores the dynamic correlation and temporal characteristics between detection points within the fault sub-region, resulting in low identification accuracy in complex fault scenarios and an inability to accurately reflect the actual occurrence of the fault.

[0004] Secondly, in terms of tracing the source of distribution network faults, existing technologies often lack comprehensive analysis and mining of historical fault data, making it difficult to extract valuable fault feature information from massive amounts of data. Furthermore, existing technologies lack a deep understanding of the propagation mechanism of distribution network faults within the topology and lack systematic research on the dispersion and interference characteristics of fault signals in the distribution network. This makes the fault location process susceptible to the effects of fault signal dispersion and interference, leading to tracing errors and an inability to accurately determine the location of the fault source, thus affecting the targetedness and effectiveness of fault handling.

[0005] Therefore, the present invention provides a method and system for fault location in distribution networks based on topological features. Summary of the Invention

[0006] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0007] The technical solution adopted by this invention to solve its technical problem is:

[0008] Firstly, a method for fault location in distribution networks based on topological features includes:

[0009] Within multiple historical analysis periods, the fault points in the distribution network within each historical analysis period are analyzed to determine the influence relationship between the fault points in the distribution network.

[0010] Within each historical analysis period, fault sub-regions within the distribution network are acquired, and the abnormal current values ​​corresponding to each fault sub-region are extracted. These values ​​are then combined with the influence relationships between fault points within the distribution network to assess the fault sub-region type.

[0011] The cause types of each fault sub-region are statistically analyzed, a multi-abnormality fault sequence and a main abnormality impact sequence are constructed, coexisting anomaly points are identified, and coexisting anomaly points in multiple historical analysis periods are analyzed to obtain anomaly tracing base points;

[0012] Analyze the abnormal tracing baseline, obtain the fault tracing value, and locate the type of fault source.

[0013] As a preferred embodiment of the present invention, the analysis process for fault points in the distribution network within each historical analysis period is as follows:

[0014] The historical analysis period is divided into several historical analysis time periods, and the historical analysis time periods are divided into several historical analysis points. The historical current value of any detection point in the distribution network at each historical analysis point is obtained. The difference between the historical current values ​​at two adjacent historical analysis points is obtained to obtain the adjacent change range.

[0015] If the adjacent change amplitude is greater than the adjacent change amplitude threshold, it is displayed as an amplitude change anomaly signal, and the analyzed detection point is recorded as an amplitude change anomaly point, and the analyzed historical analysis period is marked as an amplitude change anomaly period;

[0016] Within the distribution network, all detection points adjacent to the amplitude anomaly point are extracted as target detection points, and the adjacent change amplitude of each target detection point within the historical analysis period is obtained. If the adjacent change amplitude is greater than the adjacent change amplitude threshold, the analyzed target detection point is recorded as the disturbed detection point.

[0017] A preferred embodiment of the present invention is as follows: determining the influence relationship between fault points within the distribution network, the process of which is as follows:

[0018] Extract all the interference detection points in each amplitude anomaly period and compare them for overlap. Count the number of interference detection points that overlap in different amplitude anomaly periods and calculate the ratio with the number of all detection points in the distribution network. Output the ratio of overlapping interference detection points.

[0019] During the period of amplitude variation anomaly, the difference between the adjacent change amplitude corresponding to the interference detection point and the adjacent change amplitude threshold is calculated, and the interference difference value is output. The difference between the adjacent change amplitude value corresponding to the amplitude variation anomaly point and the adjacent change amplitude threshold is calculated, and the amplitude variation difference value is output.

[0020] The amplitude variation difference value is calculated by comparing it with the interference difference value corresponding to the interference detection point, and the amplitude variation difference coefficient is output.

[0021] The standard deviation of the amplitude variation difference coefficient corresponding to each amplitude variation anomaly period of the interference detection point is calculated, and the standard deviation of amplitude variation difference is output.

[0022] The ratio of the number of overlapping interference detections to the standard deviation of amplitude variation is calculated, and the influence analysis value is output.

[0023] If the impact analysis value is greater than the impact analysis threshold, it is displayed as a high-impact signal; if the impact analysis value is less than or equal to the impact analysis threshold, it is displayed as a low-impact signal.

[0024] As a preferred embodiment of the present invention, the process for obtaining the fault sub-region is as follows:

[0025] Arbitrarily extract one amplitude variation anomaly period within each historical analysis period as the target analysis period, extract all amplitude variation anomaly points of the distribution network within the target analysis period, and take the distance between any two amplitude variation anomaly points as the amplitude variation anomaly distance.

[0026] If the amplitude variation anomaly distance is within the range of adjacent anomaly distances, then it belongs to the adjacent amplitude variation anomaly point;

[0027] If the amplitude anomaly distance does not fall within the range of adjacent anomaly distances, then it belongs to a non-adjacent amplitude anomaly point;

[0028] From the adjacent amplitude variation anomaly points, one amplitude variation anomaly point is randomly selected as the central anomaly point, and the other amplitude variation anomaly point is selected as the secondary anomaly point;

[0029] Based on secondary anomalies, and after removing adjacent amplitude anomalies, the remaining amplitude anomalies are arbitrarily combined, and iterative analysis is performed in accordance with the method of judging adjacent amplitude anomalies until the amplitude anomaly distance is no longer within the range of adjacent anomaly distances.

[0030] As a preferred embodiment of the present invention, the analysis process for fault sub-region types is as follows:

[0031] Obtain the target analysis period corresponding to each amplitude variation anomaly point in the fault sub-region, and count the number of amplitude variation anomalies in the same target analysis period and the proportion of the total number of amplitude variation anomalies, and output the ratio of the number of anomalies in the same period.

[0032] Extract the amplitude variation difference values ​​corresponding to the amplitude variation anomaly points within the same target analysis period, calculate the standard deviation, and output the amplitude variation anomaly standard deviation.

[0033] The summation of the ratio of the number of anomalies in the same period and the standard deviation of the amplitude anomaly is used to calculate the fault sub-region type value.

[0034] If the fault sub-region type value is greater than or equal to the fault sub-region type threshold, it is marked as a multi-origin fault sub-region;

[0035] If the fault sub-region type value is less than the fault sub-region type threshold, it is marked as a primary affected sub-region.

[0036] A preferred embodiment of the present invention is as follows: the process of obtaining multiple anomaly sequences and the main anomaly influence sequence, and determining coexisting anomaly points is as follows:

[0037] All fault sub-regions that generate multiple abnormal fault signals are sorted from largest to smallest according to the fault sub-region type value and integrated into a multiple abnormal fault sequence.

[0038] All fault sub-regions that generate main anomaly signals are sorted in ascending order of fault sub-region type value and integrated into the main anomaly sequence.

[0039] Within the multiple anomaly sequence and the main anomaly influence sequence, the multiple anomaly sub-regions with the same detection point of amplitude variation anomaly are combined with the main anomaly influence sub-regions to obtain multiple sub-region overlap analysis groups;

[0040] Within the sub-region overlap analysis group, detection points with the same amplitude variation anomaly are extracted as coexisting anomaly points, resulting in multiple coexisting anomaly points.

[0041] As a preferred embodiment of the present invention, the process for obtaining the anomaly tracing baseline is as follows:

[0042] Arbitrarily select a coexisting anomaly as the target anomaly, obtain the total number of times the target anomaly appears in each historical analysis period, and calculate the ratio with the total number of amplitude anomalies in the historical analysis period to output the target anomaly frequency ratio.

[0043] Calculate the standard deviation of the target anomaly frequency ratio for each historical analysis period and output the target anomaly frequency standard deviation.

[0044] The average number of target anomalies is calculated by averaging the ratio of the number of target anomalies for each historical analysis period, and the average number of target anomalies is output.

[0045] Substitute the mean and standard deviation of the number of target anomalies into the coefficient of variation formula to obtain the target anomaly stability value. Select the target anomaly point corresponding to the minimum target anomaly stability value as the anomaly tracing base point.

[0046] As a preferred embodiment of the present invention, the fault tracing value is obtained in the following way:

[0047] Extract the period of amplitude variation anomaly that occurs in each historical analysis period as the anomaly tracing base point, and obtain the ranking of the anomaly tracing period in the time series of each historical analysis period as the unit time series ranking. Calculate the mean of the corresponding unit time series ranking in each historical analysis period to obtain the anomaly time series ranking.

[0048] The distribution network is transformed into a grid-based coordinate system, and the coordinates of the anomaly tracing base points in each historical analysis period are obtained as the unit base point coordinates.

[0049] Extract the center coordinates of the multiple heterogeneous daughter region and the main heterogeneous influence daughter region in each historical analysis period;

[0050] Within each historical analysis period, the distance between the unit base point coordinates and the center coordinates of the main heterogeneous influence sub-region is calculated using the coordinate point distance formula, and then the average value is calculated to obtain the main heterogeneous center distance.

[0051] The ratio of the main heterocentric center distance to the perimeter of the main heterocentric influence sub-region is calculated, and the main heterocentric distance ratio is output.

[0052] The fault tracing value is obtained by summing the abnormal time sequence ranking with the main distance ratio.

[0053] As a preferred embodiment of the present invention, the process for determining the type of fault source is as follows:

[0054] If the fault traceability value is greater than the fault traceability threshold, then the fault source type is the multiple anomaly type.

[0055] If the fault traceability value is less than or equal to the fault traceability threshold, then the fault source type is the primary and secondary impact type.

[0056] Secondly, a distribution network fault location system based on topology features includes:

[0057] Fault Relationship Analysis Module: Analyzes fault points in the distribution network within each historical analysis period to determine the influence relationships between fault points in the distribution network.

[0058] Fault Sub-area Identification Module: In each historical analysis period, it acquires fault sub-areas within the distribution network, extracts the abnormal current value corresponding to each fault sub-area, and combines it with the influence relationship between fault points within the distribution network to evaluate the fault sub-area type.

[0059] Anomaly tracing and analysis module: Statistically analyzes the cause types of each fault sub-area, constructs multiple anomaly generation sequences and main anomaly impact sequences, identifies coexisting anomaly points, and analyzes coexisting anomaly points in multiple historical analysis periods to obtain anomaly tracing base points;

[0060] Fault source location module: Analyzes the abnormal tracing base point, obtains the fault tracing value, and locates the fault source type.

[0061] The beneficial effects of this invention are as follows:

[0062] This invention analyzes fault points in the distribution network within each historical analysis period, determines the influence relationship between fault points in the distribution network, obtains fault sub-regions in the distribution network within each historical analysis period, extracts the abnormal current value corresponding to each fault sub-region, and combines it with the influence relationship between fault points in the distribution network to evaluate the fault sub-region type. This enables a more accurate delineation of the fault location area, avoids the fault sub-region being too large or too small due to inaccurate judgment of adjacent relationships, and improves the accuracy of fault location in the spatial dimension.

[0063] This invention statistically analyzes the cause types of each fault sub-region, constructs multiple anomaly sequences and main anomaly influence sequences, identifies coexisting anomaly points, and analyzes these coexisting anomaly points across multiple historical analysis periods to obtain anomaly tracing baselines. Analyzing these baselines yields fault tracing values ​​and pinpoints the fault source type. This not only enables more accurate determination of the fault source region but also facilitates a deeper understanding of the fault propagation mechanism in the distribution network, avoiding tracing deviations caused by the dispersion and interference of fault signals, and significantly improving the accuracy and efficiency of fault source tracing. Attached Figure Description

[0064] The invention will now be further described with reference to the accompanying drawings.

[0065] Figure 1 This is a flowchart of the steps of a distribution network fault location method based on topology features according to the present invention;

[0066] Figure 2 This is a schematic diagram of a power distribution network fault location system based on topological features according to the present invention. Detailed Implementation

[0067] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments. Example 1

[0068] Please see Figure 1 As shown in the embodiment of the present invention, a method for fault location in a distribution network based on topology features includes the following steps:

[0069] Step 1: Analyze the fault points in the distribution network within each historical analysis period to determine the influence relationship between the fault points in the distribution network.

[0070] In a preferred embodiment, the historical analysis period is divided into several equal historical analysis periods, and the duration of each historical analysis period is equal.

[0071] The historical analysis period is divided into several historical analysis points, and the historical current value of any detection point in the distribution network at each historical analysis point is obtained.

[0072] Within the historical analysis period, the difference between the historical current values ​​at two adjacent historical analysis points is used to obtain the adjacent change range;

[0073] If the amplitude of adjacent changes is less than or equal to the threshold of adjacent changes, it indicates that the amplitude change of the analyzed detection point is small between two adjacent historical analysis points, which is a normal amplitude signal.

[0074] If the adjacent change amplitude is greater than the adjacent change amplitude threshold, it indicates that the amplitude change of the analyzed detection point is large between two adjacent historical analysis points, which is displayed as an amplitude change anomaly signal. The analyzed detection point is recorded as an amplitude change anomaly point, and the analyzed historical analysis period is marked as an amplitude change anomaly period.

[0075] Within the distribution network, all detection points adjacent to the amplitude anomaly point are extracted as target detection points;

[0076] Obtain the adjacent change amplitude of each target detection point within the historical analysis period. If the adjacent change amplitude is greater than the adjacent change amplitude threshold, it indicates that the amplitude of the analyzed target detection point changes significantly between two adjacent historical analysis points, which is displayed as an adjacent amplitude change abnormal signal. The analyzed target detection point is recorded as the interfered detection point.

[0077] Extract all the interference detection points in each amplitude anomaly period and compare them for overlap. Count the number of interference detection points that overlap in different amplitude anomaly periods and calculate the ratio with the number of all detection points in the distribution network. Output the ratio of overlapping interference detection points.

[0078] During the period of amplitude anomaly, the difference between the adjacent change amplitude corresponding to the interference detection point and the adjacent change amplitude threshold is calculated, and the interference difference value is output.

[0079] The difference between the adjacent amplitude values ​​corresponding to the amplitude anomaly point and the adjacent amplitude threshold is calculated to obtain the amplitude difference value.

[0080] The amplitude variation difference value is calculated by comparing it with the interference difference value corresponding to the interference detection point, and the amplitude variation difference coefficient is output.

[0081] The standard deviation of the amplitude variation difference coefficient corresponding to each amplitude variation anomaly period of the interference detection point is calculated, and the standard deviation of amplitude variation difference is output.

[0082] The ratio of the number of overlapping interference detections to the standard deviation of amplitude variation is calculated, and the influence analysis value is output.

[0083] It is understandable that the impact analysis values ​​represent the strength of the influence relationship between fault points within the distribution network. Specifically, the purpose is to:

[0084] Objective 1: To obtain the influence relationship between different detection points of the fault, which is helpful to determine the degree of interference and stability of the adjacent detection points when an amplitude anomaly occurs at a certain detection point, based on the influence analysis value. In combination with the line connection topology of the distribution network, the path of the fault propagation from the source to other areas can be inferred.

[0085] Objective 2: In the distribution network topology, faults often have an impact within a certain range. By analyzing overlapping interference detection points, we can identify areas that are more concentrated and stable in terms of fault impact. Combined with the topological characteristics of the distribution network, we can improve the accuracy of tracing the location of the fault source.

[0086] The impact analysis values ​​are compared with the impact analysis thresholds, as follows;

[0087] If the impact analysis value is greater than the impact analysis threshold, it indicates that the proportion of overlapping interference detection points in different fault periods is relatively high, and the fluctuation of amplitude difference is small, which shows a highly impactful signal.

[0088] If the impact analysis value is less than or equal to the impact analysis threshold, it indicates that the proportion of overlapping interference detection points in different fault periods is relatively low, and the amplitude difference fluctuates greatly, indicating a low-level impact signal.

[0089] Step 2: Within each historical analysis period, obtain the fault sub-regions in the distribution network, extract the abnormal current value corresponding to each fault sub-region, and combine it with the influence relationship between fault points in the distribution network to evaluate the fault sub-region type;

[0090] Among them, the mainstream fault cause types are multiple abnormal fault signals or main abnormal influence signals;

[0091] It should be noted that the meaning of multiple abnormal current signals is: within the historical analysis period, the current at multiple monitoring nodes in the distribution network shows a small amplitude abnormality. Due to the large number of such abnormalities, the cumulative effect causes a fault in the distribution network.

[0092] The meaning of the main anomaly impact signal is: within the historical analysis period, the current at a monitoring node in the distribution network is significantly abnormal, which in turn affects the current at many other monitoring nodes in the distribution network to be slightly abnormal.

[0093] In a preferred embodiment, the process of obtaining the faulty sub-region is as follows:

[0094] Arbitrarily extract one period of amplitude variation anomaly within each historical analysis period as the target analysis period;

[0095] During the target analysis period, all amplitude transformer anomaly points in the distribution network are extracted, and the distance between any two amplitude transformer anomaly points is taken as the amplitude transformer anomaly distance.

[0096] If the amplitude change anomaly distance is within the range of adjacent anomaly distances, it means that the two amplitude change anomaly points analyzed are close in location in the spatial dimension of the distribution network and belong to adjacent amplitude change anomaly points;

[0097] If the amplitude anomaly distance is not within the range of adjacent anomaly distances, it means that the two amplitude anomaly points analyzed are far apart in the spatial dimension of the distribution network and belong to non-adjacent amplitude anomaly points.

[0098] From the adjacent amplitude variation anomaly points, one amplitude variation anomaly point is randomly selected as the central anomaly point, and the other amplitude variation anomaly point is selected as the secondary anomaly point;

[0099] Based on secondary anomalies, and after removing adjacent amplitude anomalies, the remaining amplitude anomalies are arbitrarily combined, and iterative analysis is performed in accordance with the method of judging adjacent amplitude anomalies until the amplitude anomaly distance is no longer within the range of adjacent anomaly distances.

[0100] The method for obtaining the distance range between adjacent anomalies is as follows:

[0101] Within the power distribution network, arbitrarily select adjacent detection points and obtain the distance between adjacent detection points as the distance between adjacent detection points;

[0102] Compare the distances between all adjacent detection points, and select the largest and smallest distances between adjacent detection points to form the range of adjacent anomaly distances;

[0103] It should be noted that the purpose of obtaining the distance range between adjacent anomalies is:

[0104] Objective 1: From a spatial perspective, selecting the central and secondary anomalies from adjacent amplitude transformer anomalies and conducting iterative analysis based on these anomalies can more accurately delineate the fault location area, avoid overly large or small fault sub-region locations due to inaccurate judgment of adjacency relationships, improve the spatial accuracy of fault location, gain a deeper understanding of the fault propagation mechanism in the distribution network topology, and provide a basis for taking targeted fault isolation and repair measures.

[0105] Objective 2: From a data perspective, when analyzing the current anomaly data of numerous monitoring nodes in the distribution network, the distance range between adjacent anomalies helps to filter out data closely related to fault location, reduce the interference of irrelevant data on the analysis results, improve data processing efficiency and analysis accuracy, and in the iterative analysis process, after removing adjacent amplitude transformer anomaly points, the remaining points are combined for analysis, avoiding repeated calculations and invalid analysis, so that the data can serve fault location more effectively;

[0106] Objective 3: In the process of acquiring fault sub-regions, iterative analysis is performed based on the distance range between adjacent anomalies, providing clear termination conditions and judgment criteria for the iterative algorithm, enabling the algorithm to run more efficiently. Furthermore, since the distance range between adjacent anomalies is determined based on the actual distance between adjacent detection points in the distribution network, it can adapt to distribution networks with different topologies, making the fault location algorithm more versatile and adaptable.

[0107] The analysis process for fault sub-region types is as follows:

[0108] Obtain the target analysis period corresponding to each amplitude variation anomaly point in the fault sub-region, and count the number of amplitude variation anomalies in the same target analysis period and the proportion of the total number of amplitude variation anomalies, and output the ratio of the number of anomalies in the same period.

[0109] Extract the amplitude variation difference values ​​corresponding to the amplitude variation anomaly points within the same target analysis period, calculate the standard deviation, and output the amplitude variation anomaly standard deviation.

[0110] The summation of the ratio of the number of anomalies in the same period and the standard deviation of the amplitude anomaly is used to calculate the fault sub-region type value.

[0111] It is understandable that the fault sub-region type value represents the following meaning: by reflecting the spatial distribution density and amplitude consistency of abnormal current within the fault sub-region, specifically, if the fault sub-region type value is larger, it indicates that there are more detection points with abnormal amplitude changes in the same period, and the amplitude change difference value corresponding to each detection point with abnormal amplitude changes has a larger deviation; if the fault sub-region type value is smaller, it indicates that there are fewer detection points with abnormal amplitude changes in the same period, and the amplitude change difference value corresponding to each detection point with abnormal amplitude changes has a smaller deviation.

[0112] If the fault sub-region type value is greater than or equal to the fault sub-region type threshold, it indicates that there are a large number of detection points with abnormal amplitude changes in the same period, and the amplitude difference value corresponding to each detection point with abnormal amplitude changes has a large deviation. The fault sub-region type is evaluated as a multi-abnormal fault signal and marked as a multi-abnormal fault sub-region.

[0113] If the fault sub-region type value is less than the fault sub-region type threshold, it indicates that the number of detection points with abnormal amplitude changes in the same period is small, and the amplitude difference value corresponding to each detection point with abnormal amplitude changes has a small deviation. The fault sub-region type is evaluated as the main abnormal influence signal, and it is marked as the main abnormal influence sub-region.

[0114] The purpose of obtaining the fault sub-area type is to: on the one hand, by clarifying the fault type, we can analyze the topological relationship and electrical connection between these nodes, improve the accuracy of locating the fault source area; on the other hand, if it is a multi-abnormal fault signal, it means that the fault is caused by the accumulation of small abnormalities of multiple nodes, which may require a comprehensive inspection of the lines and equipment in the area, replacement of aging or damaged parts, and optimization of the electrical connection in the area to prevent similar faults from recurring. If it is a main abnormal influence signal, the key node should be identified, and measures should be taken immediately to cut off the power to the node to prevent the fault from expanding further. Then, the node should be inspected and repaired in detail.

[0115] The specific solution in this embodiment is as follows: Within multiple historical analysis periods, the fault points in the distribution network within each historical analysis period are analyzed to determine the influence relationship between the fault points in the distribution network. Within each historical analysis period, the fault sub-regions in the distribution network are obtained, and the abnormal current values ​​corresponding to each fault sub-region are extracted. These values ​​are then combined with the influence relationship between the fault points in the distribution network to evaluate the type of fault sub-region. This allows for a more accurate delineation of the fault location area, avoiding the fault sub-region being too large or too small due to inaccurate judgment of adjacent relationships, and improving the accuracy of fault location in the spatial dimension. Example 2

[0116] Please see Figure 1 As shown in the embodiment of the present invention, a method for fault location in a distribution network based on topology features includes the following steps:

[0117] Step 3: Statistically analyze the cause types of each fault sub-region, construct the multi-abnormality fault sequence and the main abnormality impact sequence, identify coexisting anomalies, and analyze the coexisting anomalies in multiple historical analysis periods to obtain the anomaly tracing base point;

[0118] In a preferred embodiment, the process of obtaining the multiple heterogeneous sequences is as follows:

[0119] All fault sub-regions that generate multiple fault signals are sorted from largest to smallest according to the fault sub-region type value and integrated into a multiple fault sequence.

[0120] The process of obtaining the main heterogeneous influence sequence is as follows:

[0121] All fault sub-regions that generate main anomaly signals are sorted in ascending order of fault sub-region type value and integrated into the main anomaly sequence.

[0122] The process for determining coexisting outliers is as follows:

[0123] Within the multiple anomaly sequence and the main anomaly influence sequence, the multiple anomaly sub-regions with the same detection point of amplitude variation anomaly are combined with the main anomaly influence sub-regions to obtain multiple sub-region overlap analysis groups;

[0124] Within the sub-region overlap analysis group, the detection points with the same amplitude variation anomaly are extracted as coexisting anomaly points, resulting in multiple coexisting anomaly points;

[0125] The process of obtaining the anomaly tracing baseline is as follows:

[0126] Arbitrarily select a coexisting anomaly as the target anomaly, obtain the total number of times the target anomaly appears in each historical analysis period, and calculate the ratio with the total number of amplitude anomalies in the historical analysis period to output the target anomaly frequency ratio.

[0127] Calculate the standard deviation of the target anomaly frequency ratio for each historical analysis period and output the target anomaly frequency standard deviation.

[0128] The average number of target anomalies is calculated by averaging the ratio of the number of target anomalies for each historical analysis period, and the average number of target anomalies is output.

[0129] Substituting the mean and standard deviation of the target anomalies into the coefficient of variation formula, the stable value of the target anomalies is obtained. ;

[0130] Specifically, the formula for the coefficient of variation is: ,in, Expressed as the standard deviation of the number of target anomalies. Represented as the average number of target anomalies;

[0131] It is understandable that the target anomaly stability value means that it reflects the stability of the frequency of amplitude anomalies of the target anomaly point in different historical analysis periods. Specifically, the standard deviation of the number of target anomalies reflects the fluctuation of the target anomaly point in different historical analysis periods, while the mean of the number of target anomalies reflects the average frequency of amplitude anomalies of the target anomaly point in all historical analysis periods, thus providing a data foundation for subsequent tracing of the fault source node.

[0132] Compare the target anomaly stability values ​​corresponding to each target anomaly point, and select the target anomaly point corresponding to the smallest target anomaly stability value as the anomaly tracing base point;

[0133] The purpose of determining the anomaly tracing baseline is as follows:

[0134] Objective 1: Anomaly tracing baselines reflect points where amplitude anomalies are stable in frequency across different historical analysis periods. These points not only enable more accurate identification of the fault source region but also help to gain a deeper understanding of the fault propagation mechanism in the distribution network, avoiding tracing deviations caused by the dispersion and interference of fault signals, and greatly improving the accuracy of fault source tracing.

[0135] Objective 2: The number and distribution of anomaly tracing points can reflect the overall health status of the distribution network. By timely detecting and handling faults related to anomaly tracing points, the number of faults and power outage time can be reduced, thereby improving the power supply reliability and stability of the distribution network.

[0136] Step 4: Analyze the anomaly tracing baseline, obtain the fault tracing value, and locate the type of fault source;

[0137] In a preferred embodiment, the fault traceability value is obtained as follows:

[0138] For example, the abnormality tracing base point is extracted from the period of amplitude abnormality in each historical analysis period, and used as the abnormality tracing period;

[0139] Obtain the ranking of the anomaly tracing period in the time series within each historical analysis period, as the unit time series ranking;

[0140] It should be noted that if the abnormal time series ranking is smaller, it means that the abnormal tracing base point occurred earlier in the historical analysis period. If the abnormal time series ranking is larger, it means that the abnormal tracing base point occurred later in the historical analysis period.

[0141] The average of the time series rankings of the corresponding units within each historical analysis period is used to obtain the abnormal time series rankings.

[0142] The distribution network is transformed into a grid-based coordinate system, and the coordinates of the anomaly tracing base points in each historical analysis period are obtained as the unit base point coordinates.

[0143] Extract the center coordinates of the multiple heterogeneous daughter region and the main heterogeneous influence daughter region in each historical analysis period;

[0144] Within each historical analysis period, the distance between the unit base point coordinates and the center coordinates of the main heterogeneous influence sub-region is calculated using the coordinate point distance formula, and the average value is calculated to obtain the main heterogeneous center distance.

[0145] The ratio of the main heterocentric center distance to the perimeter of the main heterocentric influence sub-region is calculated, and the main heterocentric distance ratio is output.

[0146] The fault tracing value is obtained by summing the abnormal time sequence ranking with the main distance ratio.

[0147] It is understandable that the fault tracing value represents the probability that the fault type leading to the fault at the fault tracing baseline is a primary anomaly-affected fault. Specifically, on the one hand, the anomaly time series ranking reflects the stability of the anomaly tracing baseline in the early stages of the historical analysis period, indicating that the more likely the anomaly is to occur in the early stages of the fault cycle, the higher the degree of matching with the characteristic of a primary anomaly-affected source: "a strong anomaly appears first, then the influence spreads." On the other hand, the primary anomaly distance ratio reflects the relative position of the fault tracing baseline and the center of the primary anomaly-affected sub-region. The smaller the value, the closer the point is to the center of the primary anomaly sub-region, meaning that it is more likely to be the fault source in terms of topological structure.

[0148] The fault tracing value is compared with the fault tracing threshold, as follows:

[0149] If the fault tracing value is greater than the fault tracing threshold, it indicates that the fault occurred relatively late in each historical analysis period and is far from the center of the main abnormal influence sub-region, and the fault source type is the multi-abnormal fault type.

[0150] If the fault tracing value is less than or equal to the fault tracing threshold, it indicates that the fault occurred earlier in each historical analysis period and is closer to the center of the main anomaly sub-region, and the fault source type is the main anomaly type.

[0151] The specific scheme of this embodiment is as follows: statistically analyze the cause types of each fault sub-region, construct multiple anomaly sequence and main anomaly influence sequence, determine coexisting anomaly points, analyze coexisting anomaly points in multiple historical analysis periods to obtain anomaly tracing base points, analyze the anomaly tracing base points to obtain fault tracing values, and locate the fault source type. This not only enables more accurate determination of the fault source area, but also helps to deeply understand the fault propagation mechanism in the distribution network, avoids tracing deviations caused by the dispersion and interference of fault signals, and greatly improves the accuracy and efficiency of fault source tracing.

[0152] Example 3

[0153] Please see Figure 2 As shown, this embodiment of the invention provides both a method for locating distribution network faults based on topology features and a corresponding system for locating distribution network faults based on topology features, including the following modules:

[0154] Fault Relationship Analysis Module: Analyzes fault points in the distribution network within each historical analysis period to determine the influence relationships between fault points in the distribution network.

[0155] Fault Sub-area Identification Module: In each historical analysis period, it acquires fault sub-areas within the distribution network, extracts the abnormal current value corresponding to each fault sub-area, and combines it with the influence relationship between fault points within the distribution network to evaluate the fault sub-area type.

[0156] Anomaly tracing and analysis module: Statistically analyzes the cause types of each fault sub-area, constructs multiple anomaly generation sequences and main anomaly impact sequences, identifies coexisting anomaly points, and analyzes coexisting anomaly points in multiple historical analysis periods to obtain anomaly tracing base points;

[0157] Fault source location module: Analyzes the abnormal tracing base point, obtains the fault tracing value, and locates the fault source type.

[0158] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for fault location in a power distribution network based on topological features, characterized in that: Comprise: In multiple historical analysis periods, analyze the fault points in the power distribution network in each historical analysis period, and determine the influence relationship between the fault points in the power distribution network; In each historical analysis period, obtain the fault sub-area in the power distribution network, and extract the abnormal current value corresponding to each fault sub-area, and combine the influence relationship between the fault points in the power distribution network to evaluate the fault sub-area type; Statistical analysis of each fault sub-area cause type, construction of multiple hetero genesis fault sequence and main hetero influence sequence, determination of coexisting abnormal point, and analysis of coexisting abnormal point in multiple historical analysis periods to obtain abnormal tracing base point; Analyze the abnormal tracing base point to obtain the fault tracing value and locate the fault source type; The acquisition process of the fault sub-area is as follows: Arbitrarily extract a amplitude variation abnormal period in each historical analysis period as a target analysis period, extract all amplitude variation abnormal points in the target analysis period, and take the distance between any two amplitude variation abnormal points as the amplitude variation abnormal distance; If the amplitude variation abnormal distance exists within the adjacent abnormal distance range, it belongs to the adjacent amplitude variation abnormal point; If the amplitude variation abnormal distance does not exist within the adjacent abnormal distance range, it belongs to the non-adjacent amplitude variation abnormal point; Arbitrarily select one amplitude variation abnormal point from the adjacent amplitude variation abnormal point as the center abnormal point, and the other as the secondary abnormal point; Based on the secondary abnormal point, after excluding the adjacent amplitude variation abnormal point, the remaining amplitude variation abnormal points are arbitrarily combined, and the iteration analysis is carried out according to the method of judging the adjacent amplitude variation abnormal point until the amplitude variation abnormal distance does not exist within the adjacent abnormal distance range; The analysis process of the fault sub-area type is as follows: Get the target analysis period corresponding to each amplitude variation abnormal point in the fault sub-area, and count the number of amplitude variation abnormal points in the same target analysis period, the proportion of the total number of amplitude variation abnormal points, and output the simultaneous period abnormal number ratio; Extract the amplitude variation difference value corresponding to the amplitude variation abnormal point in the same target analysis period, and calculate the standard deviation, and output the amplitude variation abnormal standard deviation; Sum the simultaneous period abnormal number ratio and the amplitude variation abnormal standard deviation to obtain the fault sub-area type value; If the fault sub-area type value is greater than or equal to the fault sub-area type threshold, it is marked as a multiple hetero genesis fault sub-area; If the fault sub-area type value is less than the fault sub-area type threshold, it is marked as a main hetero influence sub-area; The process of determining the coexisting abnormal point is as follows: Sort all the fault sub-areas that generate multiple hetero genesis fault signals according to the fault sub-area type value from large to small, and integrate them into a multiple hetero genesis fault sequence; Sort all the fault sub-areas that generate main hetero influence signals according to the fault sub-area type value from small to large, and integrate them into a main hetero influence sequence; In the multiple hetero genesis fault sequence and the main hetero influence sequence, combine the multiple hetero genesis fault sub-areas and the main hetero influence sub-areas that exist in the same detection point of amplitude variation abnormality to obtain multiple sub-area overlap analysis groups; In the sub-area overlap analysis group, extract the amplitude variation abnormality same detection point as the coexisting abnormal point to obtain multiple coexisting abnormal points; The acquisition process of the abnormal tracing base point is as follows: Optionally selecting one coexisting abnormal point as a target abnormal point, obtaining the total number of times that the target abnormal point appears in each historical analysis period, and calculating the ratio with the total number of times that the amplitude variation abnormal point appears in the historical analysis period, and outputting the target abnormal number ratio; Calculating the standard deviation of the target abnormal number ratio corresponding to each historical analysis period, and outputting the target abnormal number standard deviation; The target abnormal number mean value is calculated by the mean value calculation of the target abnormal number ratio corresponding to each historical analysis period, and the target abnormal number mean value is obtained; The target abnormal stability value is obtained by substituting the target abnormal number mean value and the target abnormal number standard deviation into the coefficient of variation formula, and the target abnormal point corresponding to the minimum target abnormal stability value is selected as the abnormal backtracking base point; The fault backtracking value is obtained in the following way: Extracting the amplitude variation abnormal period of the abnormal backtracking base point in each historical analysis period as the abnormal backtracking period, obtaining the ranking of the abnormal backtracking period on the time sequence in each historical analysis period as the unit time sequence ranking, and calculating the mean value of the corresponding unit time sequence ranking in each historical analysis period to obtain the abnormal time sequence ranking; The distribution network is converted into a grid coordinate system, and the coordinate points of the abnormal backtracking base point in each historical analysis period are obtained as the unit base point coordinate points; The center coordinates of the multiple foreign fault subareas and the main foreign influence subareas in each historical analysis period are extracted respectively; In each historical analysis period, the distance between the unit base point coordinate point and the center coordinate of the main foreign influence subarea is calculated by the coordinate point distance formula, and the mean value is calculated to obtain the main foreign center distance; The main foreign center distance is calculated by the ratio of the main foreign center distance and the perimeter of the main foreign influence subarea, and the main foreign distance ratio is outputted; The fault backtracking value is obtained by summing the abnormal time sequence ranking and the main foreign distance ratio.

2. The method of claim 1, wherein: The analysis process of each fault point in the distribution network in each historical analysis period is as follows: The historical analysis period is equally divided into several historical analysis periods, and the historical analysis period is equally divided into several historical analysis points. The historical current value of any detection point in the distribution network at each historical analysis point is obtained, and the historical current values at adjacent two historical analysis points are subtracted to obtain the adjacent change amplitude; If the adjacent change amplitude is greater than the adjacent change amplitude threshold, it is displayed as an amplitude variation abnormal signal, and the analyzed detection point is marked as an amplitude variation abnormal point, and the analyzed historical analysis period is marked as an amplitude variation abnormal period; In the distribution network, all detection points adjacent to the amplitude variation abnormal point are extracted as target detection points, and the adjacent change amplitude of each target detection point in the historical analysis period is obtained. If the adjacent change amplitude is greater than the adjacent change amplitude threshold, the analyzed target detection point is marked as a disturbed detection point.

3. The method of claim 2, wherein: The influence relationship between the fault points in the distribution network is determined as follows: Extracting all disturbed detection points in each amplitude variation abnormal period, and performing coincidence comparison, counting the number of disturbed detection points that coincide in different amplitude variation abnormal periods, and calculating the ratio with the number of all detection points in the distribution network, and outputting the coincidence interference detection number ratio; In the amplitude variation abnormal period, the adjacent amplitude variation corresponding to the interference detection point is subtracted from the adjacent amplitude variation threshold, and the difference value is output to obtain the interference difference value. The adjacent amplitude variation value corresponding to the amplitude variation abnormal point is subtracted from the adjacent amplitude variation threshold, and the difference value is output to obtain the amplitude variation difference value; The amplitude variation difference value is calculated by the ratio of the interference difference value corresponding to the interference detection point, and the amplitude variation difference coefficient is output; The amplitude variation difference coefficient corresponding to the interference detection point in each amplitude variation abnormal period is calculated by the standard deviation, and the amplitude variation difference standard deviation is output; The influence analysis value is obtained by the ratio calculation of the coincidence interference detection number ratio and the amplitude variation difference standard deviation; If the influence analysis value is greater than the influence analysis threshold, it is displayed as a high impact signal, and if the influence analysis value is less than or equal to the influence analysis threshold, it is displayed as a low impact signal.

4. The method of claim 1, wherein: The determination process of the fault source type is as follows: If the fault tracing value is greater than the fault tracing threshold, the fault source type is a multi-foreign source type; If the fault tracing value is less than or equal to the fault tracing threshold, the fault source type is a main-foreign impact type.

5. A topology feature based power distribution network fault location system for performing the method of any of the preceding claims 1-4, characterized in that: It includes: Fault relationship analysis module: In multiple historical analysis periods, analyze the fault points in the power distribution network in each historical analysis period to determine the influence relationship between the fault points in the power distribution network; Fault sub-area identification module: In each historical analysis period, obtain the fault sub-area in the power distribution network, and extract the abnormal current value corresponding to each fault sub-area, and combine it with the influence relationship between the fault points in the power distribution network to evaluate the fault sub-area type; Abnormal tracing analysis module: Statistics of each fault sub-area cause type, construction of multi-foreign fault sequence and main-foreign impact sequence, determination of coexisting abnormal points, and analysis of coexisting abnormal points in multiple historical analysis periods to obtain abnormal tracing base points; Fault source positioning module: Analyze the abnormal tracing base point to obtain the fault tracing value and locate the fault source type; The acquisition process of the fault sub-area is as follows: Arbitrarily extract an amplitude variation abnormal period in each historical analysis period as a target analysis period, and extract all amplitude variation abnormal points in the power distribution network. The distance between any two amplitude variation abnormal points is taken as the amplitude variation abnormal distance; If the amplitude variation abnormal distance exists within the adjacent abnormal distance range, it belongs to the adjacent amplitude variation abnormal point; If the amplitude variation abnormal distance does not exist within the adjacent abnormal distance range, it belongs to the non-adjacent amplitude variation abnormal point; Arbitrarily select one amplitude variation abnormal point from the adjacent amplitude variation abnormal points as the center abnormal point, and the other as the secondary abnormal point; Based on the secondary abnormal point, after excluding the adjacent amplitude variation abnormal points, the remaining amplitude variation abnormal points are arbitrarily combined, and the iteration analysis is performed according to the method of judging the adjacent amplitude variation abnormal points until the amplitude variation abnormal distance does not exist within the adjacent abnormal distance range; The analysis process of the fault sub-area type is as follows: Get the target analysis period corresponding to each amplitude variation abnormal point in the fault sub-area, and count the number of amplitude variation abnormal points in the same target analysis period, and the proportion of the total number of amplitude variation abnormal points, and output to obtain the simultaneous period abnormal number ratio; Extract the amplitude variation difference value corresponding to the amplitude variation abnormal point in the same target analysis period, and perform standard deviation calculation to output the amplitude variation abnormal standard deviation; Sum the number of abnormal points in the same period and the amplitude variation abnormal standard deviation to output the fault sub-zone type value; If the fault sub-zone type value is greater than or equal to the fault sub-zone type threshold value, it is marked as a multi-heterogeneous fault sub-zone; If the fault sub-zone type value is less than the fault sub-zone type threshold value, it is marked as a main hetero-affected sub-zone; Get the multi-heterogeneous fault sequence and the main hetero-affected sequence, and the coexisting abnormal point determination process is as follows: Sort all fault sub-zones that generate multi-heterogeneous fault signals in descending order according to the fault sub-zone type value, and integrate them into a multi-heterogeneous fault sequence; Sort all fault sub-zones that generate main hetero-affected signals in ascending order according to the fault sub-zone type value, and integrate them into a main hetero-affected sequence; In the multi-heterogeneous fault sequence and the main hetero-affected sequence, combine the multi-heterogeneous fault sub-zone and the main hetero-affected sub-zone that exist in the same amplitude abnormal detection point to obtain multiple sub-zone overlap analysis groups; In the sub-zone overlap analysis group, extract the amplitude abnormal detection point as the coexisting abnormal point to obtain multiple coexisting abnormal points. The acquisition process of the abnormal trace base point is as follows: Arbitrarily select a coexisting abnormal point as a target abnormal point, get the total number of times the target abnormal point appears in each historical analysis period, and perform ratio calculation with the total number of amplitude variation abnormal points in the historical analysis period to output the target abnormal frequency ratio; Perform standard deviation calculation on the target abnormal frequency ratio corresponding to each historical analysis period to output the target abnormal frequency standard deviation; Perform mean value calculation on the target abnormal frequency ratio corresponding to each historical analysis period to output the target abnormal frequency mean value; Put the target abnormal frequency mean value and the target abnormal frequency standard deviation into the coefficient of variation formula to output the target abnormal stability value, and select the target abnormal point corresponding to the minimum target abnormal stability value as the abnormal trace base point.

Images