Transformer fault determination method, device and system

Abstract

The invention provides a transformer fault determination method, device and system. According to the method, iterative clustering is carried out on a gas concentration data set by adopting a K-means clustering algorithm based on a genetic algorithm, an optimal clustering center is obtained, a diagnosis sub-model is obtained, the minimum Euclidean distance is determined, and the fault type of the transformer is determined at least according to the size of the minimum Euclidean distance. The purpose of combining a traditional transformer fault diagnosis method and an intelligent transformer fault diagnosis method is achieved, the operation difficulty is reduced on the premise that it is guaranteed that the precision is not affected, and therefore the problems that operation is complex and precision is poor when transformer faults are detected through an existing scheme are solved.

Description

Technical Field

[0001] The present application relates to the technical field of transformer diagnosis, and in particular to a method, device and system for determining transformer faults. Background Art

[0002] In the daily operation and maintenance of transformers, operation and maintenance personnel usually carry out regular maintenance. Although this method can effectively maintain transformers to avoid failures, the maintenance efficiency is limited by the scale of substations and the number of operation and maintenance personnel. At the same time, with the continuous expansion of the scale of power grids, the increasing number of connected devices, and the increasingly complex types of connected devices, regular maintenance relies on the professional skills of operation and maintenance personnel, the maintenance time arrangement is inflexible, and the coexistence of insufficient and excessive maintenance has gradually become apparent.

[0003] When the transformer is operating normally, the transformer oil and fiber insulation age and decompose due to the catalytic effects of moisture, oxygen, heat, and materials such as copper and iron. These chemical reactions produce gases, most of which dissolve in the oil, but the rate of gas generation is usually slow. However, once an initial fault or new fault conditions occur within the transformer, the rate and amount of gas generation increases dramatically. Most initial defects show early signs, so proper analysis of dissolved gases in the transformer oil can detect fault conditions in time.

[0004] At present, transformer fault diagnosis methods based on genetic algorithms are mainly divided into traditional methods and intelligent methods. Traditional methods are mainly based on work experience, and require the definition and description of the absolute concentration of gas, the concentration ratio between different gases, the proportion of gas dissolution, etc., including Duval triangle method, Gouda triangle method and other technologies. Intelligent methods usually use artificial intelligence to process and analyze dissolved gas, including artificial neural networks, supervised learning and unsupervised learning and other technologies.

[0005] Traditional methods and intelligent methods have their own advantages and disadvantages. Traditional methods are relatively simple and easy to understand and operate, but they have certain deficiencies in accuracy and uncertainty, and may cause misidentification in extreme cases. Intelligent methods have high fault diagnosis accuracy, but their operation is more complicated, and the diagnosis results often depend on the set parameters and the feature input vector used, resulting in the intelligent method's application scope being less broad than that of traditional methods.

[0006] That is, the existing scheme is more complicated to operate and has poor accuracy when detecting transformer faults. Summary of the invention

[0007] The main purpose of the present application is to provide a method, device and system for determining transformer faults, so as to at least solve the problem that the existing solutions are complex in operation and have poor accuracy when detecting transformer faults.

[0008] In order to achieve the above object, according to one aspect of the present application, a method for determining a transformer fault is provided, the method comprising:

[0009] Acquire a gas concentration data set, wherein the gas concentration data set includes concentrations of multiple dissolved gases in transformer oil, wherein the types of dissolved gases include H2, CH4, C2H6, C2H4, and C2H2;

[0010] Using a K-means clustering algorithm based on a genetic algorithm to iteratively cluster the gas concentration data set to obtain an optimal cluster center;

[0011] Obtaining a diagnostic sub-model, wherein the diagnostic sub-model represents a determination model of transformer faults generated using expert opinions, wherein the diagnostic sub-model includes a plurality of subsets, and each of the subsets includes a reference centroid;

[0012] A minimum Euclidean distance is determined, and a fault type of the transformer is determined at least according to the size of the minimum Euclidean distance, wherein the minimum Euclidean distance is the minimum value of the Euclidean distance between the optimal cluster center and the reference centroid.

[0013] Optionally, the gas concentration data set is iteratively clustered using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal clustering center, including:

[0014] The optimal cluster center is determined in a manner that minimizes the total variance of the cluster.

[0015] Optionally, in the process of iteratively clustering the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal cluster center, the method further includes:

[0016] according to Determine the total variance of the clusters;

[0017] in, SSE is the total variance of the cluster, d(x i ,m j ) is the Euclidean distance between them, n is the number of data in the gas concentration dataset, k is the number of non-overlapping subsets in the gas concentration dataset, m j is the cluster center of the jth non-overlapping subset of the gas concentration dataset, x i is the i-th data in the gas concentration data set.

[0018] Optionally, in the process of iteratively clustering the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal cluster center, the method further includes:

[0019] according to Determine the cluster center;

[0020] in, n is the number of data in the gas concentration data set, m j is the cluster center of the jth non-overlapping subset of the gas concentration dataset, x i is the i-th data in the gas concentration data set.

[0021] Optionally, determining the fault type of the transformer at least according to the minimum Euclidean distance includes:

[0022] When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the first subset of the diagnosis sub-model, it is determined that the fault type of the transformer is low-energy discharge.

[0023] Optionally, determining the fault type of the transformer at least according to the minimum Euclidean distance includes:

[0024] When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the second subset of the diagnostic submodel, obtaining a first concentration ratio, a second concentration ratio, a third concentration ratio and a fourth concentration ratio in the gas concentration data set, wherein the first concentration ratio is C2H6 / (CH4+C2H4), the second concentration ratio is C2H2 / C2H4, the third concentration ratio is C2H4 / C2H6, and the fourth concentration ratio is (C2H6+C2H2) / (C2H4);

[0025] When the second concentration ratio is greater than or equal to the first threshold value, and the fourth concentration ratio is greater than or equal to the second threshold value, determining that the fault type of the transformer is low-energy discharge;

[0026] When the second concentration ratio is greater than or equal to the first threshold value, and the fourth concentration ratio is less than the second threshold value, determining that the fault type of the transformer is partial discharge;

[0027] When the second concentration ratio is less than the first threshold, the first concentration ratio is greater than or equal to a third threshold, and the third concentration ratio is less than a fourth threshold, determining that the fault type of the transformer is low-energy discharge;

[0028] When the second concentration ratio is less than the first threshold, the first concentration ratio is greater than or equal to the third threshold, and the third concentration ratio is greater than or equal to the fourth threshold, determining that the fault type of the transformer is high-energy discharge;

[0029] When the second concentration ratio is smaller than the first threshold value and the first concentration ratio is smaller than the third threshold value, it is determined that the fault type of the transformer is partial discharge.

[0030] Optionally, determining the fault type of the transformer at least according to the minimum Euclidean distance includes:

[0031] When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the third subset of the diagnostic submodel, obtaining a first concentration ratio, a second concentration ratio and a fourth concentration ratio in the gas concentration data set, wherein the first concentration ratio is C2H6 / (CH4+C2H4), the second concentration ratio is C2H2 / C2H4, and the fourth concentration ratio is (C2H6+C2H2) / (C2H4);

[0032] When the first concentration ratio is less than a third threshold, determining that the fault type of the transformer is a thermal fault;

[0033] When the first concentration ratio is greater than or equal to the third threshold, the second concentration ratio is less than the third threshold, and the fourth concentration ratio is greater than a fifth threshold, determining that the fault type of the transformer is a thermal fault;

[0034] When the first concentration ratio is greater than or equal to the third threshold, the second concentration ratio is less than the third threshold, and the fourth concentration ratio is less than or equal to the fifth threshold, determining that the fault type of the transformer is a thermal fault;

[0035] When the first concentration ratio is greater than or equal to the third threshold and the second concentration ratio is greater than or equal to the third threshold, it is determined that the fault type of the transformer is a thermal fault.

[0036] Optionally, determining the fault type of the transformer at least according to the minimum Euclidean distance includes:

[0037] When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of a preset subset of the diagnostic submodel, it is determined that the fault type of the transformer includes a thermal fault and a discharge fault, and the preset subset is a subset that is not the first subset, not the second subset, and not the third subset.

[0038] According to another aspect of the present application, a transformer fault determination device is provided, the transformer fault determination device comprising:

[0039] A first acquisition unit is used to acquire a gas concentration data set, wherein the gas concentration data set includes concentrations of multiple dissolved gases in transformer oil, wherein the types of dissolved gases include H2, CH4, C2H6, C2H4, and C2H2;

[0040] A first processing unit is used to iteratively cluster the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal clustering center;

[0041] A second acquisition unit is used to acquire a diagnosis sub-model, wherein the diagnosis sub-model represents a determination model of transformer faults generated by expert opinions, wherein the diagnosis sub-model includes a plurality of subsets, and each of the subsets includes a reference centroid;

[0042] The second processing unit is used to determine a minimum Euclidean distance and determine the fault type of the transformer at least according to the size of the minimum Euclidean distance, wherein the minimum Euclidean distance is the minimum value of the Euclidean distance between the optimal cluster center and the reference centroid.

[0043] According to another aspect of the present application, a transformer fault determination system is provided, the system comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include methods for executing any one of the methods described.

[0044] By applying the technical solution of the present application, the gas concentration data set is iteratively clustered using a K-means clustering algorithm based on a genetic algorithm to obtain the optimal cluster center, obtain a diagnostic sub-model, determine the minimum Euclidean distance, and determine the fault type of the transformer at least based on the size of the minimum Euclidean distance. This achieves the purpose of combining the traditional method and the intelligent method for transformer fault diagnosis, reduces the difficulty of operation while ensuring that the accuracy is not affected, and thus solves the problem that the existing scheme is complex to operate and has poor accuracy when detecting transformer faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The drawings constituting part of the present application are used to provide a further understanding of the present application. The exemplary embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0046] Figure 1 A schematic flow chart of a method for determining a transformer fault provided in accordance with an embodiment of the present application is shown;

[0047] Figure 2 A structural block diagram of a transformer fault determination device provided according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0048] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0049] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0050] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0051] For the convenience of description, some nouns or terms involved in the embodiments of the present application are explained below:

[0052] Roger four-ratio method: By comparing the ratios of different transformer parameters, it is possible to determine whether the transformer is faulty.

[0053] Improved IEC ratio method: Based on the International Electrotechnical Commission standard, it compares different parameters of the transformer to determine whether a fault exists.

[0054] IEC 60599 method: Based on the IEC 60599 standard, the condition of the transformer is diagnosed by detecting the gas in the transformer insulating oil.

[0055] Three-ratio method: Diagnose the condition of the transformer by comparing the ratios of three parameters of the transformer.

[0056] Gouda triangulation: Diagnose the condition of the transformer by drawing a triangle diagram between different parameters.

[0057] Duval triangulation: Similar to Gouda triangulation, the condition of the transformer is diagnosed by drawing a triangle diagram.

[0058] HYOSUN ratio method: Diagnose the condition of the transformer by calculating the ratio of different parameters.

[0059] Conditional probability method: Determine the state of the transformer by calculating the conditional probability of different parameters.

[0060] CSUS ANN method: A method based on artificial neural network, which diagnoses the status of the transformer by training the neural network.

[0061] Self-organizing map clustering method: The transformer data is clustered through the self-organizing map algorithm to identify different states.

[0062] As introduced in the background technology, the current transformer fault diagnosis methods based on genetic algorithms are mainly divided into traditional methods and intelligent methods. The traditional method is mainly based on work experience, and it is necessary to define and explain the absolute concentration of the gas, the concentration ratio between different gases, the proportion of gas dissolution, etc., including Duval triangulation, Gouda triangulation and other technologies. In the intelligent method, artificial intelligence is usually used to process and analyze the dissolved gas, including artificial neural networks, supervised learning and unsupervised learning and other technologies. In order to solve the problem that the existing scheme is more complicated to operate and has poor accuracy when detecting transformer faults, the embodiments of the present application provide a method, device and system for determining transformer faults.

[0063] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0064] In this embodiment, a method for determining a transformer fault running on a mobile terminal, a computer terminal or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0065] Figure 1 FIG. 1 is a flow chart of a method for determining a transformer fault according to an embodiment of the present application. Figure 1 As shown, the method comprises the following steps:

[0066] Step S101, obtaining a gas concentration data set, the gas concentration data set including the concentrations of multiple dissolved gases in transformer oil, the types of the dissolved gases including H2, CH4, C2H6, C2H4, and C2H2;

[0067] Step S102, using a K-means clustering algorithm based on a genetic algorithm to iteratively cluster the gas concentration data set to obtain an optimal cluster center;

[0068] In one embodiment of the present application, in the process of iteratively clustering the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal clustering center, the method further includes:

[0069] according to Determine the total variance of the clusters;

[0070] in, SSE is the total variance of the above clusters, d(x i ,m j ) is the Euclidean distance between them, n is the number of data in the above gas concentration dataset, k is the number of non-overlapping subsets in the above gas concentration dataset, m j is the cluster center of the jth non-overlapping subset in the above gas concentration dataset, x i is the i-th data in the above gas concentration data set.

[0071] The implementation steps of the K-means clustering algorithm are as follows:

[0072] 1) Determine the number of non-overlapping subset sets k, randomly assign k points as the centroids of the clusters and initialize the clusters.

[0073] 2) Assign each point to the cluster closest to the centroid.

[0074] 3) Use the total variance calculation formula to update the centroid based on the given data.

[0075] Repeat 2) and 3) until no data changes its cluster membership or the criterion function does not improve in multiple iterations.

[0076] In one embodiment of the present application, in the process of iteratively clustering the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal clustering center, the method further includes:

[0077] according to Determine the cluster center;

[0078] in, n is the number of data in the above gas concentration data set, m j is the cluster center of the jth non-overlapping subset in the above gas concentration dataset, x i is the i-th data in the above gas concentration data set.

[0079] In one embodiment of the present application, the K-means clustering algorithm is used based on the genetic algorithm to iteratively cluster the above gas concentration data set to obtain the optimal cluster center, including:

[0080] The optimal cluster center is determined in a way that minimizes the total variance of the cluster.

[0081] Step S103, obtaining a diagnosis sub-model, wherein the diagnosis sub-model represents a transformer fault determination model generated by expert opinions, wherein the diagnosis sub-model includes a plurality of subsets, and each of the subsets includes a reference centroid;

[0082] Step S104, determining a minimum Euclidean distance, and determining the fault type of the transformer at least according to the size of the minimum Euclidean distance, wherein the minimum Euclidean distance is the minimum value of the Euclidean distance between the optimal cluster center and the reference centroid.

[0083] In the above steps, the K-means clustering algorithm is used based on the genetic algorithm to iteratively cluster the above gas concentration data set to obtain the optimal cluster center, obtain the diagnostic sub-model, determine the minimum Euclidean distance, and at least determine the fault type of the transformer based on the size of the minimum Euclidean distance, thereby achieving the purpose of combining the traditional method and the intelligent method of transformer fault diagnosis, reducing the difficulty of operation while ensuring that the accuracy is not affected, thereby solving the problem that the existing scheme is more complicated to operate and has poor accuracy when detecting transformer faults.

[0084] First case:

[0085] In one embodiment of the present application, determining the fault type of the transformer at least according to the size of the minimum Euclidean distance includes:

[0086] When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the first subset of the diagnosis sub-model, it is determined that the fault type of the transformer is low-energy discharge.

[0087] The Euclidean distance threshold of the reference centroid of the first subset is the Euclidean distance between the first centroid in the centroid matrix M and the optimal cluster center.

[0088] The subsequent Euclidean distance threshold is similar and will not be elaborated here.

[0089] Second case:

[0090] In one embodiment of the present application, determining the fault type of the transformer at least according to the size of the minimum Euclidean distance includes:

[0091] When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the second subset of the diagnostic submodel, obtaining a first concentration ratio, a second concentration ratio, a third concentration ratio and a fourth concentration ratio in the gas concentration data set, wherein the first concentration ratio is C2H6 / (CH4+C2H4), the second concentration ratio is C2H2 / C2H4, the third concentration ratio is C2H4 / C2H6, and the fourth concentration ratio is (C2H6+C2H2) / (C2H4);

[0092] When the second concentration ratio is greater than or equal to the first threshold value, and the fourth concentration ratio is greater than or equal to the second threshold value, determining that the fault type of the transformer is low-energy discharge;

[0093] When the second concentration ratio is greater than or equal to the first threshold value, and the fourth concentration ratio is less than the second threshold value, determining that the fault type of the transformer is partial discharge;

[0094] When the second concentration ratio is less than the first threshold, the first concentration ratio is greater than or equal to the third threshold, and the third concentration ratio is less than the fourth threshold, determining that the fault type of the transformer is low-energy discharge;

[0095] When the second concentration ratio is less than the first threshold, the first concentration ratio is greater than or equal to the third threshold, and the third concentration ratio is greater than or equal to the fourth threshold, determining that the fault type of the transformer is high-energy discharge;

[0096] When the second concentration ratio is smaller than the first threshold value and the first concentration ratio is smaller than the third threshold value, it is determined that the fault type of the transformer is partial discharge.

[0097] Specifically, the first threshold is 25, the second threshold is 250, the third threshold is 0.1, and the fourth threshold is 3.5.

[0098] The third case:

[0099] In one embodiment of the present application, determining the fault type of the transformer at least according to the size of the minimum Euclidean distance includes:

[0100] When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the third subset of the diagnostic submodel, obtaining a first concentration ratio, a second concentration ratio and a fourth concentration ratio in the gas concentration data set, wherein the first concentration ratio is C2H6 / (CH4+C2H4), the second concentration ratio is C2H2 / C2H4, and the fourth concentration ratio is (C2H6+C2H2) / (C2H4);

[0101] When the first concentration ratio is less than a third threshold, determining that the fault type of the transformer is a thermal fault;

[0102] When the first concentration ratio is greater than or equal to the third threshold, the second concentration ratio is less than the third threshold, and the fourth concentration ratio is greater than the fifth threshold, determining that the fault type of the transformer is a thermal fault;

[0103] When the first concentration ratio is greater than or equal to the third threshold, the second concentration ratio is less than the third threshold, and the fourth concentration ratio is less than or equal to the fifth threshold, determining that the fault type of the transformer is a thermal fault;

[0104] When the first concentration ratio is greater than or equal to the third threshold value, and the second concentration ratio is greater than or equal to the third threshold value, it is determined that the fault type of the transformer is a thermal fault.

[0105] Specifically, the fifth threshold is 0.25.

[0106] In one embodiment of the present application, determining the fault type of the transformer at least according to the size of the minimum Euclidean distance includes:

[0107] When the above-mentioned minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the preset subset of the above-mentioned diagnostic sub-model, it is determined that the fault type of the above-mentioned transformer includes thermal fault and discharge fault, and the above-mentioned preset subset is a subset that is not the first subset, not the second subset, and not the third subset.

[0108] Specifically, when the preset subset is not a subset of the first subset, the second subset, and the third subset, it is determined that the fault type of the transformer includes a thermal fault and a discharge fault, and if the first situation, the second situation, and the third situation are not met, it is determined that the fault type of the transformer includes a thermal fault and a discharge fault.

[0109] The standard fault classification is shown in Table 1.

[0110] Table 1

[0111]

[0112] Different fault types and fault locations will produce different dissolved gases, such as hydrogen (H2), methane (CH4), ethane (C2H6), ethylene (C2H4), acetylene (C2H2), propane (C3H8) and propylene (C3H6), carbon monoxide (CO), carbon dioxide (CO2), etc. Different fault types correspond to different main dissolved gases and small dissolved gases as shown in Table 2.

[0113] Table 2

[0114] Fault type Main gas Small amount of gas PD <![CDATA[H2,CH4,CO]]> <![CDATA[C2H6,C2H2,CO2]]> <![CDATA[D1]]> <![CDATA[H2,C2H2]]> / <![CDATA[D2]]> <![CDATA[H2,C2H2,CO,CO2]]> <![CDATA[CH4,C2H4,C2H6]]> <![CDATA[T1]]> <![CDATA[CH4,C2H6,CO,CO2]]> <![CDATA[H2,C2H4]]> <![CDATA[T2]]> <![CDATA[CH4,C2H4]]> <![CDATA[H2]]> <![CDATA[T3]]> <![CDATA[C2H4]]> <![CDATA[H2,C2H6]]>

[0115] The data were preprocessed and K-means clustering was performed on the dissolved gas data using genetic algorithm.

[0116] When generating subsets, the training data needs to be preprocessed, and the relative proportion of dissolved gas in transformer oil is used instead of gas concentration:

[0117]

[0118] p i is the relative proportion of the i-th dissolved gas in the transformer oil, C1 to C5 represent the concentrations of H2, CH4, C2H6, C2H4, and C2H2, respectively, in parts per million.

[0119] The clustering problem can be summarized as follows: Let X be a set of n data samples, where each data sample is an m-dimensional feature vector, then the cluster C of X can be expressed as a set of k non-overlapping subsets, that is:

[0120]

[0121] The K-means clustering algorithm is based on partitioning clustering technology, which divides the data set into several clusters by optimizing a criterion function. The principle is that for a data set X = {x1, x2, ... x n}, find the set of k non-overlapping subsets C = {C1, C2, ..., C k}, so that the total variance of the cluster is minimized;

[0122] The implementation steps of the K-means clustering algorithm are as follows:

[0123] 1) Determine the number of non-overlapping subset sets k, randomly assign k points as the centroids of the clusters and initialize the clusters.

[0124] 2) Assign each point to the cluster closest to the centroid.

[0125] 3) Use the total variance calculation formula to update the centroid based on the given data.

[0126] Repeat 2) and 3) until no data changes its cluster membership or the criterion function does not improve in multiple iterations.

[0127] Genetic algorithm is an evolutionary algorithm inspired by genetics and evolution theory, which uses genetic operations such as natural selection, crossover and mutation to generate new populations. In each iteration, the selection operator selects individuals from the current population for the next generation of reproduction based on the fitness level of the individuals. In K-means clustering based on genetic algorithm, a population of individuals containing candidate centroids is first created, and each individual is evaluated by calculating the total variance of each individual. During the initialization process, feature vectors are randomly selected from the data set to form the initial population.

[0128] For the total population N pop After the individuals have been evaluated, the population is sorted and half N is identified and marked. pop The best individuals are obtained by selecting parent individuals from these best individuals, applying the crossover operator to generate offspring individuals, applying the mutation operator to generate variants, re-evaluating the combined population consisting of the current population, offspring individuals, and variants, and selecting N from the combined population through natural selection. pop The best individual performs the next iteration to update the population. The best cluster set is obtained by the candidate centroid of the best individual in the previous iteration.

[0129] Through clustering, a traditional diagnosis sub-model is proposed by experts to separate different faults related to the subsets.

[0130] After clustering, a cluster may contain one or more faults, so a traditional diagnostic sub-model based on gas ratio is proposed to separate different faults related to the subset. The diagnostic model contains 15 gas ratios mainly H2, CH4, C2H6, C2H4, and C2H2, and the final diagnostic model is obtained by combining different sub-models.

[0131] The failure ratio in the same cluster is shown in Table 3.

[0132] Table 3

[0133]

[0134]

[0135] The diagnostic process is as follows:

[0136] Hydrogen (H2), methane (CH4), ethane (C2H6), ethylene (C2H4), acetylene (C2H2), propane (C3H8) and propylene (C3H6), carbon monoxide (CO), carbon dioxide (CO2);

[0137] Set the number of clusters to 120 and load the centroid matrix M;

[0138] Enter the dissolved gas sample concentration and calculate the ratio R1 to R 15 ;

[0139] Calculate the input feature vector consisting of the five gas concentrations;

[0140] Calculate the Euclidean distance between the sample and the centroid;

[0141] Determine a subset of the sample:

[0142] If d min =d1, then the sample is subset 1;

[0143] If d min =d2, then the sample is subset 2;

[0144] This process is repeated in this way until the number of samples reaches the subset 120.

[0145] d min It is the minimum distance between the input sample and each centroid in each centroid matrix.

[0146] P and U represent transformers of on-load tap changers with and without communication, respectively, I and B represent transformers and bushings, R represents reactors, and C represents cables. As shown in Tables 4 and 5, it can be seen that the method of the present invention has better diagnostic accuracy for different devices than other methods.

[0147] Table 4

[0148]

[0149]

[0150] Table 5

[0151]

[0152]

[0153] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0154] The embodiment of the present application also provides a device for determining a transformer fault. It should be noted that the device for determining a transformer fault in the embodiment of the present application can be used to execute the method for determining a transformer fault provided in the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions thereof will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0155] The following is an introduction to the transformer fault determination device provided in an embodiment of the present application.

[0156] Figure 2 1 is a structural block diagram of a transformer fault determination device provided according to an embodiment of the present application. Figure 2 As shown, the device comprises:

[0157] A first acquisition unit 21 is used to acquire a gas concentration data set, wherein the gas concentration data set includes concentrations of multiple dissolved gases in transformer oil, wherein the types of the dissolved gases include H2, CH4, C2H6, C2H4, and C2H2;

[0158] A first processing unit 22 is used to iteratively cluster the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal clustering center;

[0159] A second acquisition unit 23 is used to acquire a diagnosis sub-model, wherein the diagnosis sub-model represents a transformer fault determination model generated by expert opinions, wherein the diagnosis sub-model includes a plurality of subsets, and each of the subsets includes a reference centroid;

[0160] The second processing unit 24 is used to determine a minimum Euclidean distance, and determine the fault type of the transformer at least according to the size of the minimum Euclidean distance, wherein the minimum Euclidean distance is the minimum value of the Euclidean distance between the optimal cluster center and the reference centroid.

[0161] In the above-mentioned device, a K-means clustering algorithm is used based on a genetic algorithm to iteratively cluster the above-mentioned gas concentration data set to obtain the optimal cluster center, obtain a diagnostic sub-model, determine the minimum Euclidean distance, and determine the fault type of the transformer at least based on the size of the above-mentioned minimum Euclidean distance, thereby achieving the purpose of combining the traditional method of transformer fault diagnosis with the intelligent method, reducing the difficulty of operation while ensuring that the accuracy is not affected, thereby solving the problem that the existing scheme is more complicated to operate and has poor accuracy when detecting transformer faults.

[0162] In one embodiment of the present application, the first processing unit includes a first determining module.

[0163] The first determination module is used to determine the above-mentioned optimal cluster center in a manner of minimizing the total variance of the cluster.

[0164] In one embodiment of the present application, the first processing unit includes a second determining module, in the process of iteratively clustering the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal clustering center,

[0165] The second determining module is used to Determine the total variance of the clusters;

[0166] in, SSE is the total variance of the above clusters, d(x i ,m j ) is the Euclidean distance between them, n is the number of data in the above gas concentration dataset, k is the number of non-overlapping subsets in the above gas concentration dataset, m j is the cluster center of the jth non-overlapping subset in the above gas concentration dataset, x i is the i-th data in the above gas concentration data set.

[0167] In one embodiment of the present application, the first processing unit includes a third determination module, in the process of iteratively clustering the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal clustering center,

[0168] The third determination module is used to Determine the cluster center;

[0169] in, n is the number of data in the above gas concentration data set, m j is the cluster center of the jth non-overlapping subset in the above gas concentration dataset, x i is the i-th data in the above gas concentration data set.

[0170] In one embodiment of the present application, the second processing unit includes a fourth determining module,

[0171] The fourth determination module is used to determine that the fault type of the transformer is low-energy discharge when the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the first subset of the diagnosis sub-model.

[0172] In one embodiment of the present application, the second processing unit includes an acquisition module, a fifth determination module, a sixth determination module, a seventh determination module, an eighth determination module and a ninth determination module.

[0173] The acquisition module is used for acquiring a first concentration ratio, a second concentration ratio, a third concentration ratio and a fourth concentration ratio in the gas concentration data set when the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the second subset of the diagnostic submodel, wherein the first concentration ratio is C2H6 / (CH4+C2H4), the second concentration ratio is C2H2 / C2H4, the third concentration ratio is C2H4 / C2H6, and the fourth concentration ratio is (C2H6+C2H2) / (C2H4);

[0174] The fifth determination module is used to determine that the fault type of the transformer is low-energy discharge when the second concentration ratio is greater than or equal to the first threshold and the fourth concentration ratio is greater than or equal to the second threshold;

[0175] The sixth determination module is used to determine that the fault type of the transformer is partial discharge when the second concentration ratio is greater than or equal to the first threshold and the fourth concentration ratio is less than the second threshold;

[0176] The seventh determination module is used to determine that the fault type of the transformer is low-energy discharge when the second concentration ratio is less than the first threshold, the first concentration ratio is greater than or equal to the third threshold, and the third concentration ratio is less than the fourth threshold;

[0177] The eighth determination module is used to determine that the fault type of the transformer is high-energy discharge when the second concentration ratio is less than the first threshold, the first concentration ratio is greater than or equal to the third threshold, and the third concentration ratio is greater than or equal to the fourth threshold;

[0178] The ninth determination module is used to determine that the fault type of the transformer is partial discharge when the second concentration ratio is less than the first threshold value and the first concentration ratio is less than the third threshold value.

[0179] In one embodiment of the present application, the second processing unit includes a second acquisition module, a tenth determination module, an eleventh determination module, a twelfth determination module and a thirteenth determination module.

[0180] The second acquisition module is used for acquiring the first concentration ratio, the second concentration ratio and the fourth concentration ratio in the gas concentration data set when the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the third subset of the diagnostic submodel, the first concentration ratio is C2H6 / (CH4+C2H4), the second concentration ratio is C2H2 / C2H4, and the fourth concentration ratio is (C2H6+C2H2) / (C2H4);

[0181] The tenth determination module is used to determine that the fault type of the transformer is a thermal fault when the first concentration ratio is less than a third threshold value;

[0182] The eleventh determination module is used to determine that the fault type of the transformer is a thermal fault when the first concentration ratio is greater than or equal to the third threshold, the second concentration ratio is less than the third threshold, and the fourth concentration ratio is greater than the fifth threshold;

[0183] The twelfth determination module is used to determine that the fault type of the transformer is a thermal fault when the first concentration ratio is greater than or equal to the third threshold, the second concentration ratio is less than the third threshold, and the fourth concentration ratio is less than or equal to the fifth threshold;

[0184] The thirteenth determination module is used to determine that the fault type of the transformer is a thermal fault when the first concentration ratio is greater than or equal to the third threshold and the second concentration ratio is greater than or equal to the third threshold.

[0185] In one embodiment of the present application, the second processing unit includes a fourteenth determining module,

[0186] The fourteenth determination module is used to determine that the fault type of the above-mentioned transformer includes thermal fault and discharge fault when the above-mentioned minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the preset subset of the above-mentioned diagnostic sub-model, and the above-mentioned preset subset is a subset that is not the first subset, not the second subset, and not the third subset.

[0187] The transformer fault determination device includes a processor and a memory, wherein the first acquisition unit, the first processing unit, the second acquisition unit, and the second processing unit are all stored in the memory as program units, and the processor executes the program units stored in the memory to implement corresponding functions. The modules are all located in the same processor; or, the modules are located in different processors in any combination.

[0188] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and the kernel parameters can be adjusted to solve the problem that the existing scheme is more complicated to operate and has poor accuracy when detecting transformer faults.

[0189] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0190] An embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the transformer fault determination method.

[0191] An embodiment of the present invention provides a processor, and the processor is used to run a program, wherein the method for determining a transformer fault is executed when the program is run.

[0192] The embodiment of the present invention provides a device, which includes a processor, a memory, and a program stored in the memory and can be run on the processor. When the processor executes the program, at least the following steps are implemented: obtaining a gas concentration data set, the gas concentration data set includes the concentrations of multiple dissolved gases in transformer oil, and the types of the dissolved gases include H2, CH4, C2H6, C2H4, and C2H2; iteratively clustering the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal clustering center; obtaining a diagnostic submodel, the diagnostic submodel represents a transformer fault determination model generated using expert opinions, the diagnostic submodel includes multiple subsets, and each of the subsets includes a reference centroid; determining a minimum Euclidean distance, and determining the fault type of the transformer at least based on the size of the minimum Euclidean distance, the minimum Euclidean distance being the minimum value of the Euclidean distance between the optimal clustering center and the reference centroid. The device in this article can be a server, a PC, a PAD, a mobile phone, etc.

[0193] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that is initialized with at least the following method steps: obtaining a gas concentration data set, the gas concentration data set including the concentrations of multiple dissolved gases in transformer oil, the types of dissolved gases including H2, CH4, C2H6, C2H4, and C2H2; iteratively clustering the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal clustering center; obtaining a diagnostic submodel, the diagnostic submodel representing a determination model of a transformer fault generated using expert opinions, the diagnostic submodel including multiple subsets, each of the subsets including a reference centroid; determining a minimum Euclidean distance, and determining the fault type of the transformer at least based on the size of the minimum Euclidean distance, the minimum Euclidean distance being the minimum value of the Euclidean distance between the optimal clustering center and the reference centroid.

[0194] The present application also provides a transformer fault determination system, the system comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include methods for executing any of the above methods. Based on the genetic algorithm, the K-means clustering algorithm is used to iteratively cluster the gas concentration data set to obtain the optimal cluster center, obtain the diagnostic sub-model, determine the minimum Euclidean distance, and determine the fault type of the transformer at least based on the size of the minimum Euclidean distance, thereby achieving the purpose of combining the traditional transformer fault diagnosis method with the intelligent method, reducing the difficulty of operation while ensuring that the accuracy is not affected, thereby solving the problem that the existing scheme is more complicated to operate and has poor accuracy when detecting transformer faults.

[0195] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0196] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0197] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0198] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0199] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0200] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0201] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0202] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0203] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0204] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0205] 1) The transformer fault determination method of the present application uses a K-means clustering algorithm based on a genetic algorithm to iteratively cluster the above-mentioned gas concentration data set to obtain the optimal cluster center, obtain a diagnostic sub-model, determine the minimum Euclidean distance, and determine the fault type of the transformer at least based on the size of the above-mentioned minimum Euclidean distance, thereby achieving the purpose of combining the traditional transformer fault diagnosis method and the intelligent method, reducing the operation difficulty while ensuring that the accuracy is not affected, thereby solving the problem that the existing scheme is more complicated to operate and has poor accuracy when detecting transformer faults.

[0206] 2) The transformer fault determination device of the present application uses a K-means clustering algorithm based on a genetic algorithm to iteratively cluster the above-mentioned gas concentration data set to obtain the optimal cluster center, obtain a diagnostic sub-model, determine the minimum Euclidean distance, and determine the fault type of the transformer at least based on the size of the above-mentioned minimum Euclidean distance, thereby achieving the purpose of combining the traditional transformer fault diagnosis method and the intelligent method, reducing the operation difficulty while ensuring that the accuracy is not affected, thereby solving the problem that the existing scheme is more complicated to operate and has poor accuracy when detecting transformer faults.

[0207] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for determining a transformer fault, characterized in that: include: Acquire a gas concentration data set, wherein the gas concentration data set includes concentrations of multiple dissolved gases in transformer oil, wherein the types of dissolved gases include H2, CH4, C2H6, C2H4, and C2H2; Using a K-means clustering algorithm based on a genetic algorithm to iteratively cluster the gas concentration data set to obtain an optimal cluster center; Obtaining a diagnostic sub-model, wherein the diagnostic sub-model represents a determination model of transformer faults generated using expert opinions, wherein the diagnostic sub-model includes a plurality of subsets, and each of the subsets includes a reference centroid; A minimum Euclidean distance is determined, and a fault type of the transformer is determined at least according to the size of the minimum Euclidean distance, wherein the minimum Euclidean distance is the minimum value of the Euclidean distance between the optimal cluster center and the reference centroid.

2. The method according to claim 1, characterized in that The gas concentration data set is iteratively clustered using a K-means clustering algorithm based on a genetic algorithm to obtain the optimal cluster center, including: The optimal cluster center is determined in a manner that minimizes the total variance of the cluster.

3. The method according to claim 1, characterized in that In the process of iteratively clustering the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal cluster center, the method further includes: according to Determine the total variance of the clusters; in, SSE is the total variance of the cluster, d(x i ,m j ) is the Euclidean distance between them, n is the number of data in the gas concentration dataset, k is the number of non-overlapping subsets in the gas concentration dataset, m j is the cluster center of the jth non-overlapping subset of the gas concentration dataset, x i is the i-th data in the gas concentration data set.

4. The method according to claim 1, characterized in that: In the process of iteratively clustering the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal cluster center, the method further includes: according to Determine the cluster center; in, n is the number of data in the gas concentration data set, m j is the cluster center of the jth non-overlapping subset of the gas concentration dataset, x i is the i-th data in the gas concentration data set.

5. The method according to claim 1, characterized in that: Determining the fault type of the transformer at least according to the size of the minimum Euclidean distance includes: When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the first subset of the diagnosis sub-model, it is determined that the fault type of the transformer is low-energy discharge.

6. The method according to claim 1, characterized in that Determining the fault type of the transformer at least according to the size of the minimum Euclidean distance includes: When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the second subset of the diagnostic submodel, obtaining a first concentration ratio, a second concentration ratio, a third concentration ratio and a fourth concentration ratio in the gas concentration data set, wherein the first concentration ratio is C2H6 / (CH4+C2H4), the second concentration ratio is C2H2 / C2H4, the third concentration ratio is C2H4 / C2H6, and the fourth concentration ratio is (C2H6+C2H2) / (C2H4); When the second concentration ratio is greater than or equal to the first threshold value, and the fourth concentration ratio is greater than or equal to the second threshold value, determining that the fault type of the transformer is low-energy discharge; When the second concentration ratio is greater than or equal to the first threshold value, and the fourth concentration ratio is less than the second threshold value, determining that the fault type of the transformer is partial discharge; When the second concentration ratio is less than the first threshold, the first concentration ratio is greater than or equal to a third threshold, and the third concentration ratio is less than a fourth threshold, determining that the fault type of the transformer is low-energy discharge; When the second concentration ratio is less than the first threshold, the first concentration ratio is greater than or equal to the third threshold, and the third concentration ratio is greater than or equal to the fourth threshold, determining that the fault type of the transformer is high-energy discharge; When the second concentration ratio is smaller than the first threshold value and the first concentration ratio is smaller than the third threshold value, it is determined that the fault type of the transformer is partial discharge.

7. The method according to claim 1, characterized in that Determining the fault type of the transformer at least according to the size of the minimum Euclidean distance includes: When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of the third subset of the diagnostic submodel, obtaining a first concentration ratio, a second concentration ratio and a fourth concentration ratio in the gas concentration data set, wherein the first concentration ratio is C2H6 / (CH4+C2H4), the second concentration ratio is C2H2 / C2H4, and the fourth concentration ratio is (C2H6+C2H2) / (C2H4); When the first concentration ratio is less than a third threshold, determining that the fault type of the transformer is a thermal fault; When the first concentration ratio is greater than or equal to the third threshold, the second concentration ratio is less than the third threshold, and the fourth concentration ratio is greater than a fifth threshold, determining that the fault type of the transformer is a thermal fault; When the first concentration ratio is greater than or equal to the third threshold, the second concentration ratio is less than the third threshold, and the fourth concentration ratio is less than or equal to the fifth threshold, determining that the fault type of the transformer is a thermal fault; When the first concentration ratio is greater than or equal to the third threshold and the second concentration ratio is greater than or equal to the third threshold, it is determined that the fault type of the transformer is a thermal fault.

8. The method according to claim 1, characterized in that Determining the fault type of the transformer at least according to the size of the minimum Euclidean distance includes: When the minimum Euclidean distance is equal to the Euclidean distance threshold of the reference centroid of a preset subset of the diagnostic submodel, it is determined that the fault type of the transformer includes a thermal fault and a discharge fault, and the preset subset is a subset that is not the first subset, not the second subset, and not the third subset.

9. A device for determining transformer fault, characterized in that: include: A first acquisition unit is used to acquire a gas concentration data set, wherein the gas concentration data set includes concentrations of multiple dissolved gases in transformer oil, wherein the types of dissolved gases include H2, CH4, C2H6, C2H4, and C2H2; A first processing unit is used to iteratively cluster the gas concentration data set using a K-means clustering algorithm based on a genetic algorithm to obtain an optimal clustering center; A second acquisition unit is used to acquire a diagnosis sub-model, wherein the diagnosis sub-model represents a determination model of transformer faults generated by expert opinions, wherein the diagnosis sub-model includes a plurality of subsets, and each of the subsets includes a reference centroid; The second processing unit is used to determine a minimum Euclidean distance and determine the fault type of the transformer at least according to the size of the minimum Euclidean distance, wherein the minimum Euclidean distance is the minimum value of the Euclidean distance between the optimal cluster center and the reference centroid.

10. A transformer fault determination system, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include methods for executing any one of claims 1 to 8.

Images