Power distribution network branch tripping fault analysis method, device, equipment and medium

Abstract

The invention discloses a power distribution network branch tripping fault analysis method, device and equipment and a medium, and relates to the field of power distribution network fault analysis, and the method comprises the steps: determining a feature set of tripping time sequence variation according to historical operation data; selecting m abnormal operation conditions in combination with set time to form an initial state matrix; inputting the initial state matrix into a working condition similarity operator for iterative optimization to obtain an optimal state matrix; acquiring real-time operation data of the power distribution network branch, inputting the real-time operation data and the optimal state matrix into a working condition similarity operator for calculation to obtain an estimated value, and solving a first relative error between the estimated value and the real-time operation data; and performing weighted average on the first relative error, determining a first average error of the current operation state of the power distribution network branch, and analyzing whether the power distribution network branch has a tripping fault based on the first average error. According to the invention, the problem of low accuracy of branch circuit tripping fault analysis at present is solved.

Description

Technical Field

[0001] The present invention relates to the field of distribution network fault analysis, and more specifically, to a method, device, equipment and medium for analyzing the tripping fault of a branch circuit in a distribution network. Background Technique

[0002] As the "nerve endings" of the power system, the distribution network directly provides services to users, and the safety and reliability of its operating state are particularly important. However, with the acceleration of the urbanization process and the continuous growth of electricity demand, problems such as equipment aging, load surges, and complex and changeable external environments are likely to occur, which greatly increase the risk of faults in the branch circuits of the distribution network.

[0003] The branch circuit is the last station in the power grid and is directly connected to users. Corresponding to the branch circuit is the distribution substation area, and the problems of low voltage and heavy overload in the substation area have always been important influencing factors for the power quality of the distribution network substation area. The troubles caused by it to the power consumption stability and safety are the biggest concerns of users. Moreover, compared with other parts of the power grid, the power equipment in the substation area has problems such as a more complex environment, more diverse problems, and more difficult fault detection and solution. One of the important tasks of tripping detection for the low-voltage branch circuits of the distribution network is based on the monitoring data of the branch circuit, which is the basis for tripping fault diagnosis and analysis. It takes the branch circuit data as the analysis object and realizes the tripping detection of the low-voltage branch circuit in the distribution network data based on information such as branch circuit voltage, phase current, alarm, and archives.

[0004] However, the current related technologies have insufficient correlation analysis of factors such as the three-phase voltage, phase current, active power, reactive power, and load rate change characteristics of the distribution network branch circuit, so the tripping fault of the distribution network branch circuit cannot be effectively analyzed. Summary of the Invention

[0005] The purpose of the present invention is to provide a method, device, equipment and medium for analyzing the tripping fault of a branch circuit in a distribution network, and the present invention solves the problem of low accuracy in analyzing the tripping fault of the branch circuit at present.

[0006] In the first aspect of the present invention, a method for analyzing the tripping fault of a branch circuit in a distribution network is provided, and the method includes:

[0007] Obtain the historical operation data of the branch circuit in the distribution network, and determine the feature set of the tripping time sequence change amount according to the historical operation data;

[0008] Based on the K-means clustering algorithm, cluster n related features from the feature set, and form an operating condition by the n related features. Combine the set time to select m abnormal operating conditions to form an initial state matrix; where each column vector in the initial state matrix represents an abnormal operating condition of the branch circuit, and n and m are both positive integers;

[0009] Input the initial state matrix into the operating condition similarity operator for iterative optimization to obtain the optimal state matrix;

[0010] Obtain the real-time operating data of the distribution network branch circuit, input the real-time operating data and the optimal state matrix into the operating condition similarity operator for calculation to obtain the predicted value, and calculate the first relative error between the predicted value and the real-time operating data;

[0011] Perform weighted averaging on the first relative error to determine the first average error of the current operating state of the distribution network branch circuit, and analyze whether there is a tripping fault in the distribution network branch circuit based on the first average error.

[0012] In one implementation, the historical operating data includes three-phase voltage, three-phase current, load rate, active power, and reactive power;

[0013] Determine the feature set of the tripping time sequence change amount according to the historical operating data, including:

[0014] Take the phase A current, phase A voltage, phase B current, phase B voltage, phase C current, phase C voltage, load rate, active power, and reactive power as features

[0015] Sort the features in time sequence, subtract the feature at the latter moment from the feature at the previous moment in the sorting result to obtain the time sequence change amount corresponding to each feature;

[0016] Determine the feature set of the tripping time sequence change amount based on the features and the time sequence change amount corresponding to each feature.

[0017] In one implementation, input the initial state matrix into the operating condition similarity operator for iterative optimization to obtain the optimal state matrix. Specifically:

[0018] Input the initial state matrix into the operating condition similarity operator to calculate the reconstructed predicted value;

[0019] Take the three-phase voltage, three-phase current, load rate, and active power as features, calculate the second relative error between the predicted value of each feature and the true state value, perform weighted averaging on the historical relative error, and calculate the second average error of each feature dimension;

[0020] If the error with an absolute value greater than 10 in the second average error, eliminate the corresponding operating condition from the initial state matrix to obtain a sub-state matrix;

[0021] If the mean value of the second average error is greater than 0.1 or the variance is greater than 1, and the length of the sub-state matrix is less than 80, add features to the sub-state matrix to obtain the optimal state matrix.

[0022] In one implementation, adding features to the sub-state matrix is specifically as follows:

[0023] Calculate the predicted value obtained by the real-time operation data and the sub-state matrix through the pattern similarity operator, find the error between the predicted value and the real state value, locate the feature with the largest error, and add the feature to the sub-state matrix.

[0024] In one implementation, analyzing whether there is a tripping fault in the distribution network branch based on the first average error includes:

[0025] Judge whether the first average error exceeds the preset error threshold. If it does not exceed the error threshold, the distribution network branch operates normally; if it exceeds the error threshold, judge whether the current change amount of the branch is greater than 0.1. If the current change amount of the branch is greater than 0.1, the distribution network branch trips.

[0026] In one implementation, the process of presetting the error threshold is as follows: calculate the normal average error of the normal state matrix composed of a normal operation condition;

[0027] Calculate the normal variance according to the normal average error of each normal operation condition;

[0028] Calculate the error threshold according to the number of normal operation conditions, the normal average error, and the normal variance.

[0029] In one implementation, the calculation formula of the error threshold is: where G is the error threshold, error is the average error of the nth normal operation condition, and std(·) is the variance calculation function.

[0030] In the second aspect of the present invention, a device for analyzing tripping faults in a distribution network branch is provided. The device includes:

[0031] A data processing module, configured to obtain the historical operation data of the distribution network branch, and determine the feature set of the tripping time sequence change amount according to the historical operation data;

[0032] A clustering module, configured to cluster n associated features from the feature set based on the K-means clustering algorithm, form an operation condition by the n associated features, and combine the set time to select m abnormal operation conditions to form an initial state matrix; where each column vector in the initial state matrix represents an abnormal operation condition of the branch, and n and m are both positive integers;

[0033] A matrix optimization module, configured to input the initial state matrix into the operation condition similarity operator for iterative optimization to obtain the optimal state matrix;

[0034] A relative error calculation module is configured to obtain the real-time operation data of the distribution network branch circuit, input the real-time operation data and the optimal state matrix into the operating condition similarity operator for calculation to obtain an estimated value, and calculate the first relative error between the estimated value and the real-time operation data;

[0035] A fault analysis module is configured to perform weighted averaging on the first relative error to determine the first average error of the current operating state of the distribution network branch circuit, and analyze whether there is a tripping fault in the distribution network branch circuit based on the first average error.

[0036] In a third aspect of the present invention, an electronic device is provided. The electronic device includes a processor, a memory, and a computer program stored on the memory and executable by the processor. When the computer program is executed by the processor, the steps of a method for analyzing tripping faults in a distribution network branch circuit provided in the first aspect of the present invention are implemented.

[0037] In a fourth aspect of the present invention, a computer-readable storage medium is provided. A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the steps of a method for analyzing tripping faults in a distribution network branch circuit provided in the first aspect of the present invention are implemented.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] In a method for analyzing tripping faults in a distribution network branch circuit provided by the present invention, a fault judgment algorithm for low-voltage branch circuits based on the similarity theory (SBM) is proposed to obtain the historical operation data of the distribution network branch circuit, and a feature set of the tripping time sequence change amount is determined according to the historical operation data; n related features are clustered from the feature set based on the K-means clustering algorithm, and an operating condition is composed of the n related features. Combining the set time, m abnormal operating conditions are selected to form an initial state matrix; the initial state matrix is input into the operating condition similarity operator for iterative optimization to obtain an optimal state matrix; the real-time operation data of the distribution network branch circuit is obtained, and the real-time operation data and the optimal state matrix are input into the operating condition similarity operator for calculation to obtain an estimated value, and the first relative error between the estimated value and the real-time operation data is calculated; the first relative error is weighted averaged to determine the first average error of the current operating state of the distribution network branch circuit, and whether there is a tripping fault in the distribution network branch circuit is analyzed based on the first average error. At the same time, when analyzing whether there is a tripping fault in the distribution network branch circuit based on the first average error, the correlation relationship between the tripping event and the metering time sequence data is comprehensively considered, and an error threshold with different first average errors is designed as the judgment criterion, so as to reduce the contradiction between false alarms and missed alarms and improve the accuracy of fault analysis. Description of the Drawings

[0040] The accompanying drawings described herein are used to provide a further understanding of the embodiments of the present invention, form a part of this application, and do not limit the embodiments of the present invention. In the drawings:

[0041] Figure 1 It is a schematic flowchart of a method for analyzing the tripping fault of a distribution network branch provided by an embodiment of the present invention;

[0042] Figure 2 It is a principle block diagram of a device for analyzing the tripping fault of a distribution network branch provided by an embodiment of the present invention. Detailed implementation manners

[0043] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with the embodiments and the accompanying drawings. The illustrative embodiments and descriptions thereof of the present invention are only used to explain the present invention and do not limit the present invention.

[0044] It should be noted that the term "including" or "may include" that can be used in various embodiments of this application indicates the existence of the claimed functions, operations, or elements, and does not limit the addition of one or more functions, operations, or elements. In addition, as used in various embodiments of this application, the terms "including", "having", and their cognates are only intended to represent specific features, numbers, steps, operations, elements, components, or combinations of the foregoing items, and should not be construed as first excluding the existence or addition of the possibility of one or more other features, numbers, steps, operations, elements, components, or combinations of the foregoing items.

[0045] In various embodiments of this application, the expression "or" or "at least one of B or / and C" includes any combination or all combinations of the listed words. For example, the expression "B or C" or "at least one of B or / and C" may include B, may include C, or may include both B and C.

[0046] It should be understood that terms such as "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, "a plurality of" means two or more unless otherwise specifically defined.

[0047] Please refer to Figure 1 , Figure 1 It is a schematic flowchart of a method for analyzing the tripping fault of a distribution network branch provided by an embodiment of the present invention. As Figure 1 shown, the method includes:

[0048] S101. Obtain the historical operation data of the distribution network branch circuit, and determine the feature set of the tripping time series variation based on the historical operation data.

[0049] In this embodiment, the historical operation data of the branch circuit includes tripping record data, historical feature data, and lightly loaded branch data. Among them, the data content of the tripping record data includes feeder name, distribution transformer name, distribution transformer GIS number, substation type (low-voltage total / operation and distribution 2.0 substation / operation and distribution 2.0 branch substation / ordinary substation), actual low-voltage switch fault condition, system department judgment condition (location field + fault type field), judgment accuracy condition (false alarm / correct alarm), and whether tripping is judged. The data content of the historical feature data includes metering point number, user number, user name, GISID, time, year, month, day, second, minute, rated capacity, phase A current, phase A voltage, phase B current, phase B voltage, phase C current, phase C voltage, apparent power, load rate, comprehensive multiple, active power, and reactive power. The data content of the lightly loaded branch data includes: feeder name, affiliated district bureau, current-carrying capacity (A), maximum current value, highest load rate, attribute, occurrence month, maximum load occurrence time, reason for light load, specific reason (unstructured data).

[0050] Specifically, the historical operation data includes three-phase voltage, three-phase current, load rate, active power, and reactive power.

[0051] Take phase A current, phase A voltage, phase B current, phase B voltage, phase C current, phase C voltage, load rate, active power, and reactive power as features; sort the features in time series, subtract the feature at the previous moment from the feature at the next moment in the sorting result to obtain the time series variation corresponding to each feature; based on the features and the time series variation corresponding to each feature, determine the feature set of the tripping time series variation.

[0052] Specifically, select phase A current, phase A voltage, phase B current, phase B voltage, phase C current, phase C voltage, load rate, active power, and reactive power as features and sort them in time series; subtract the feature at the previous moment from the feature at the next moment to find the variation corresponding to each feature. The 18 feature dimensions of the features and the variations of the features form an n * 18 matrix, where n is the number of moments.

[0053] S102. Cluster n associated features from the feature set based on the K-means clustering algorithm, form an operating condition with the n associated features, and combine with the set time to select m abnormal operating conditions to form an initial state matrix; where each column vector in the initial state matrix represents an abnormal operating condition of the branch circuit, and n and m are both positive integers.

[0054] In this embodiment, the K-means++ clustering algorithm is selected to perform clustering analysis on the tags, and a complete power consumption portrait system for the substation area is established. The K-means++ algorithm optimizes the step of initializing the clustering center on the basis of the K-means algorithm. The K-means++ algorithm first randomly selects a sample from the sample set as the first clustering center, and then calculates the distance D(x) between each sample in the sample set and its nearest clustering center for each sample, and then selects a sample point as the new clustering center according to the principle that the greater the D(x), the higher the probability of being selected as the new clustering center. Repeat this process until K clustering centers are selected to complete the initialization of the clustering center. The steps of the K-means algorithm are as follows:

[0055] (1) Randomly select k samples μ1, μ2,..., μk as the initial clustering centers;

[0056] (2) Calculate the distance between each sample x(i) and all clustering centers, and assign it to the class c(m) corresponding to the nearest clustering center, c(m) = min k ∣∣x(i) - μ m ∣∣ 2 ;

[0057] (3) For each category, take its centroid as the new clustering center, that is Loop through steps (2) and (3) until a preset termination condition is reached. The termination condition can be any of the following: 1) No (or the minimum number of) objects are reassigned to different clusters. 2) No (or the minimum number of) clustering centers change anymore. 3) The sum of squared errors is locally minimized. The following formula can be used to determine whether convergence has occurred. When E reaches the preset termination condition or the oscillation is very small, it indicates that the algorithm has become stable.

[0058] Select the initial state matrix. Specifically: Select samples from the given historical operation data X and form the state matrix D. That is, a certain process or device has n associated features. Suppose a sample is taken at a certain moment i, and the n features collected are selected as an operating condition, that is, X(i) = [x1, x2,..., x n T

[0059] Combined with the changes in the operating conditions in the set time, select m operating conditions and form the state matrix D, Each column vector in the initial state matrix represents an operating condition of a branch circuit.

[0060] ​Specifically, based on the k-means clustering method, m clustering centers are selected from the operation data of the past month to form an initial state matrix of m * 18, where m ranges from 30 to 50.

[0061] S103, input the initial state matrix into the operating condition similarity operator for iterative optimization to obtain the optimal state matrix.

[0062] In this embodiment, the tripping risk judgment mainly includes two parts, the operating condition mode similarity and the operating condition time similarity based on continuity. In the operating condition mode similarity module, the real-time data is reconstructed into an estimated state, and the real-time data will be compared with the estimated state. The "AND" logic is used to connect the operating mode similarity module and the operating time similarity module. Based on the comparison of the reconstructed and predicted data, the missing and misidentification of abnormal states will be reduced.

[0063] Select the subspace (D) composed of m historical operating conditions in the state matrix, which can represent the entire dynamic process of the normal operation of the equipment; the composition of the entire state matrix is the learning of the operating characteristics of the equipment, and the combination of these modes can be used to generate an estimation model;

[0064] At a certain time point, an input pattern xin is composed of a single reading of each sensor in the model:

[0065] x in =[x 1in x 2in …x nin T

[0066] Compare the similarity degree between the input pattern xin and each pattern in the initial state matrix (D), which will generate a similarity vector (a), and the number of elements it contains is the same as the number of elements of the training matrix (pattern) stored in the state matrix;

[0067]

[0068] Convert the similarity vector representing the similarity degree into a weight vector (w)

[0069] w0=G -1 ·a;

[0070]

[0071] In the formula: is a non-linear operator, selected as the (Euclidean distance) EUCLIDEAN between two vectors, that is:

[0072] Generate an estimated value through the linear combination of the sample and the weight: x out ​= D·w

[0073] That is:

[0074] A residual value is generated by subtracting the estimated pattern from the input pattern, and the variable with a lower residual value is taken as the normal variable and directly displayed in the system without triggering a warning signal; RES = x in -x out .

[0075] In one embodiment, the initial state matrix is input into the operating condition similarity operator for iterative optimization to obtain the optimal state matrix. Specifically: The initial state matrix is input into the operating condition similarity operator to calculate the reconstructed predicted value; taking the three-phase voltage, three-phase current, load rate, and active power as features, calculating the second relative error between the predicted value of each feature and the true state value, performing weighted averaging on the historical relative error, and calculating the second average error of each feature dimension; if the absolute value of the error in the second average error is greater than 10, the corresponding operating condition is removed from the initial state matrix to obtain a sub-state matrix; if the mean value of the second average error is greater than 0.1 or the variance is greater than 1, and the length of the sub-state matrix is less than 80, features are added to the sub-state matrix to obtain the optimal state matrix.

[0076] More specifically, adding features to the sub-state matrix is as follows: calculating the predicted value obtained from the real-time operation data and the sub-state matrix through the pattern similarity operator, finding the error between the predicted value and the true state value, locating the feature with the largest error, and adding the feature to the sub-state matrix.

[0077] The optimization process of the initial state matrix is as follows:

[0078] (a) Select the historical data of 96 moments per day with a 15-minute interval in the most recent month and divide it into a training set and a test set; among them, the test set is the feature set of the tripping day where tripping is known to exist, and the training set is the normal operation feature set;

[0079] (b) Input the initial state matrix into the pattern similarity operator, calculate the reconstructed predicted value, and according to the power grid operation mechanism, analyze 8 feature dimensions including phase A current, phase A voltage, phase B current, phase B voltage, phase C current, phase C voltage, load rate, and active power, calculate the relative error between the predicted value of each feature dimension and the true state value, and perform weighted averaging to obtain the error of each feature dimension;

[0080] (c) Locate the errors with an absolute value greater than 10 in the average error, and correspondingly remove these operating conditions from the original data and the initial state matrix, regarded as outliers mixed in the normal operating conditions;

[0081] (d) When the mean of the error is greater than 0.1 or the variance is greater than 1, and the length of the state matrix is less than 80, features are added to the initial state matrix, and the calculation method is as follows: Calculate the predicted value obtained by the pattern similarity operator for the original data and the state matrix, find the error between the predicted value and the true value, locate the feature with the largest error, add the feature to the state matrix, and delete it from the original data;

[0082] (e) According to the final state matrix, calculate the average value and variance of the relative error between the predicted value and the true state value, save the mean value, variance, and the final state matrix, and calculate the error threshold.

[0083] S104. Obtain the real-time operation data of the distribution network branch, input the real-time operation data and the optimal state matrix into the operating condition similarity operator for calculation to obtain the predicted value, and calculate the first relative error between the predicted value and the real-time operation data.

[0084] In this embodiment, (a) Read the real-time data of the substation area; Read the state matrix, variance, mean value, and error threshold calculated and saved by the pattern similarity calculation under normal operating conditions;

[0085] (b) Calculate the predicted value of the current state based on the SBM function, and calculate the relative error between the predicted value and the current value;

[0086] (c) Based on the relative error calculated in step (b), weighted average to obtain the average error of the current state, and determine whether the average error of the current state exceeds the error threshold. If it does not exceed, it is judged as normal; if it exceeds, further determine whether there is a situation where the current increases abnormally, that is, determine whether the current change amount > 0.1; If there is a situation where the current increases abnormally, filter this state and do not issue a tripping alarm. If not, it is judged as tripping;

[0087] The tripping risk analysis method uses accuracy and recall for performance evaluation. Specifically, it is defined as the following formula:

[0088] Among them, TP is the number of actual tripping events in the predicted tripping events, TN is the number of predictions of actual non-tripping events in the predicted non-tripping events, FN is the number of predictions of actual non-tripping events in the predicted tripping events, and FP is the number of actual tripping events in the predicted non-tripping events.

[0089] S105. Perform a weighted average on the first relative error to determine the first average error of the current operating state of the distribution network branch, and analyze whether there is a tripping fault in the distribution network branch based on the first average error.

[0090] In this embodiment, a strategy for adaptively setting the judgment threshold is proposed. The setting of the branch risk judgment threshold is the key to risk judgment. According to the comparison between the relative error of the predicted value and the true value of the algorithm and the threshold, if the error is higher than the threshold, it means that the operating condition is abnormal; if it is lower than the threshold, it means that the operating condition is similar to the historical normal operating condition. If the threshold is set too high, the number of missed trip reports will increase; if the threshold is set too low, the number of false trip reports will increase. At the same time, due to the large number of distribution transformer areas and the significant differences in the operating conditions and environments of each area, if a unified threshold multiple is used, it is impossible to flexibly adjust according to the characteristics of the distribution transformer areas.

[0091] According to the optimal state matrix, calculate the average value and variance of the relative error between the predicted value and the true state value, and save the mean value, variance, and optimal state matrix. Calculate the error threshold, and the calculation formula is as follows:

[0092] where G is the error threshold, error is the average error of the nth normal operating condition, and std(·) is the variance calculation function.

[0093] Judge whether the first average error exceeds the preset error threshold. If it does not exceed the error threshold, the distribution network branch is operating normally; if it exceeds the error threshold, then judge whether the current change in current of the branch is greater than 0.1. If the current change of the branch is greater than 0.1, the distribution network branch trips.

[0094] Specifically, based on the first average error, judge whether the average error of the current state exceeds the error threshold. If it does not exceed, judge it as normal; if it exceeds, further judge whether there is a situation where the current abnormally increases, that is, judge whether the current change > 0.1; if there is a situation where the current abnormally increases, filter this state and do not issue a trip alarm. If not, judge it as a trip.

[0095] Please refer to Figure 2 , Figure 2 which is the principle block diagram of a device for analyzing the trip fault of a distribution network branch provided by an embodiment of the present invention. As Figure 2 shown, the device includes:

[0096] A data processing module 210, configured to obtain historical operation data of a distribution network branch, and determine a feature set of the trip time sequence change amount according to the historical operation data;

[0097] A clustering module 220, configured to cluster n associated features from the feature set based on the K-means clustering algorithm. An operating condition is composed of the n associated features, and m abnormal operating conditions are combined with the set time to form an initial state matrix; wherein, each column vector in the initial state matrix represents an abnormal operating condition of the branch, and n and m are both positive integers;

[0098] A matrix optimization module 230 is configured to input an initial state matrix into a working condition similarity operator for iterative optimization to obtain an optimal state matrix;

[0099] A relative error calculation module 240 is configured to obtain real-time operation data of a distribution network branch, input the real-time operation data and the optimal state matrix into a working condition similarity operator for calculation to obtain a predicted value, and calculate a first relative error between the predicted value and the real-time operation data;

[0100] A fault analysis module 250 is configured to perform weighted averaging on the first relative error to determine a first average error of the current operation state of the distribution network branch, and analyze whether there is a tripping fault in the distribution network branch based on the first average error.

[0101] An embodiment of the present invention further provides an electronic device, which includes a processor, a memory, a communication interface, and at least one communication bus for connecting the processor, the memory, and the communication interface. The memory includes, but is not limited to, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (PROM), or a portable read-only memory (CD-ROM), and the memory is used for storing relevant instructions and data.

[0102] The communication interface is used for receiving and sending data. The processor may be one or more CPUs. When the processor is a single CPU, the CPU may be a single-core CPU or a multi-core CPU. The processor in the electronic device is configured to read one or more programs stored in the memory and perform the following operations: obtaining historical operation data of a distribution network branch, determining a feature set of a tripping time sequence change amount according to the historical operation data; clustering n associated features from the feature set based on the K-means clustering algorithm, forming an operation condition by the n associated features, and combining a set time to select m abnormal operation conditions to form an initial state matrix; wherein each column vector in the initial state matrix represents an abnormal operation condition of a branch, and both n and m are positive integers; inputting the initial state matrix into a working condition similarity operator for iterative optimization to obtain an optimal state matrix; obtaining real-time operation data of the distribution network branch, inputting the real-time operation data and the optimal state matrix into a working condition similarity operator for calculation to obtain a predicted value, and calculating a first relative error between the predicted value and the real-time operation data; performing weighted averaging on the first relative error to determine a first average error of the current operation state of the distribution network branch, and analyzing whether there is a tripping fault in the distribution network branch based on the first average error.

[0103] It should be noted that the specific implementation of each operation can be as described above Figure 1For the corresponding description of the method embodiments shown, the electronic device can be used to execute a method for analyzing the tripping fault of a distribution network branch described in the above method embodiments of the present application, which will not be elaborated here in detail.

[0104] In the embodiments of the present disclosure, a computer-readable storage medium is further provided. The computer-readable storage medium is a memory device in a computer device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and, of course, the extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space that stores the operating system of the terminal. And in this storage space, one or more instructions suitable for being loaded and executed by the processor are also stored. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the method for analyzing the tripping fault of a distribution network branch in the above embodiments. Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0105] The specific embodiments described above further elaborate the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above description is only the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for analyzing the tripping fault of a distribution network branch, characterized in that, The method includes: Obtaining the historical operation data of the distribution network branch circuit, and determining the feature set of the tripping time series variation amount according to the historical operation data; Based on the K-means clustering algorithm, clustering n associated features from the feature set, forming an operating condition by the n associated features, and combining the set time to select m abnormal operating conditions to form an initial state matrix; wherein, each column vector in the initial state matrix represents an abnormal operating condition of the branch circuit, and both n and m are positive integers; Inputting the initial state matrix into the operating condition similarity operator for iterative optimization to obtain the optimal state matrix; Obtaining the real-time operation data of the distribution network branch circuit, inputting the real-time operation data and the optimal state matrix into the operating condition similarity operator for calculation to obtain a predicted value, and calculating the first relative error between the predicted value and the real-time operation data; Performing weighted average on the first relative error to determine the first average error of the current operating state of the distribution network branch circuit, and analyzing whether there is a tripping fault in the distribution network branch circuit based on the first average error.

2. The method for analyzing the tripping fault of a distribution network branch according to claim 1, wherein The historical operation data includes three-phase voltage, three-phase current, load rate, active power, and reactive power; Determining the feature set of the tripping time series variation amount according to the historical operation data, including: Taking the A-phase current, A-phase voltage, B-phase current, B-phase voltage, C-phase current, C-phase voltage, load rate, active power, and reactive power as features Sorting the features in time series, subtracting the feature at the latter moment from the feature at the former moment in the sorting result to obtain the time series variation amount corresponding to each feature; Based on the features and the time series variation amounts corresponding to each feature, determining the feature set of the tripping time series variation amount.

3. A method for analyzing the tripping fault of a distribution network branch, as described in claim 1, characterized in that Inputting the initial state matrix into the operating condition similarity operator for iterative optimization to obtain the optimal state matrix, specifically: Inputting the initial state matrix into the operating condition similarity operator to calculate the reconstructed predicted value; Taking the three-phase voltage, three-phase current, load rate, and active power as features, calculating the second relative error between the predicted value of each feature and the true state value, performing weighted average on the historical relative error, and calculating the second average error of each feature dimension; If the error with an absolute value greater than 10 in the second average error, removing the corresponding operating condition from the initial state matrix to obtain a sub-state matrix; If the mean value of the second average error is greater than 0.1 or the variance is greater than 1, and the length of the sub-state matrix is less than 80, adding features to the sub-state matrix to obtain the optimal state matrix.

4. The method for analyzing the tripping fault of a distribution network branch according to claim 3, characterized in that, Adding features to the sub-state matrix, specifically: Calculating the predicted value obtained by the real-time operation data and the sub-state matrix through the pattern similarity operator, calculating the error between the predicted value and the true state value, locating the feature with the largest error, and adding the feature to the sub-state matrix.

5. A method for analyzing the tripping fault of a distribution network branch, characterized in that, Analyzing whether there is a tripping fault in the distribution network branch circuit based on the first average error, including: Judging whether the first average error exceeds the preset error threshold. If it does not exceed the error threshold, the distribution network branch circuit operates normally; if it exceeds the error threshold, judging whether the current change amount of the branch circuit is greater than 0.

1. If the current change amount of the branch circuit is greater than 0.1, the distribution network branch circuit trips.

6. The method for analyzing the tripping fault of a distribution network branch according to claim 5, characterized in that, The process of presetting the error threshold is as follows: calculate the normal average error of the normal state matrix composed of a normal operating condition; Calculate the normal variance according to the normal average error of each normal operating condition; Calculate the error threshold according to the number of normal operating conditions, the normal average error, and the normal variance.

7. A method for analyzing the tripping fault of a distribution network branch, characterized in that, according to claim 6, The calculation formula for the error threshold is as follows: where G is the error threshold, error is the average error of the nth normal operating condition, and std(·) is the variance calculation function.

8. A device for analyzing the tripping fault of a distribution network branch circuit, characterized in that The device includes: A data processing module, configured to obtain historical operation data of a distribution network branch, and determine a feature set of the characteristic quantity of the tripping time sequence change according to the historical operation data; A clustering module, configured to cluster n associated features from the feature set based on the K-means clustering algorithm, form an operating condition from the n associated features, and combine with a set time to select m abnormal operating conditions to form an initial state matrix; wherein each column vector in the initial state matrix represents an abnormal operating condition of the branch, and both n and m are positive integers; A matrix optimization module, configured to input the initial state matrix into the operating condition similarity operator for iterative optimization to obtain an optimal state matrix; A relative error calculation module, configured to obtain real-time operation data of the distribution network branch, input the real-time operation data and the optimal state matrix into the operating condition similarity operator for calculation to obtain a predicted value, and calculate the first relative error between the predicted value and the real-time operation data; A fault analysis module, configured to perform weighted averaging on the first relative error to determine the first average error of the current operating state of the distribution network branch, and analyze whether there is a tripping fault in the distribution network branch based on the first average error.

9. An electronic device, characterized in that, The electronic device includes a processor, a memory, and a computer program stored on the memory and executable by the processor. When the computer program is executed by the processor, the steps of a method for analyzing tripping faults in a distribution network branch as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, the steps of a method for analyzing tripping faults in a distribution network branch as described in any one of claims 1 to 7 are implemented.

Images