Defect index determination method and device of transformer bushing and electronic equipment

Abstract

The invention discloses a defect index determination method and device of a transformer bushing and electronic equipment. The method comprises the following steps: acquiring casing data of a to-be-detected casing; determining a reference defect set corresponding to the casing pipe type; determining a first defect feature according to the bushing operation data; determining a second defect feature according to the reference defect features corresponding to the plurality of reference sleeves; and determining a target defect index corresponding to the to-be-detected sleeve according to the first defect feature and the second defect feature. According to the method and the device, the technical problem that the defect index is determined inaccurately when the defect index of the transformer bushing is determined in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power systems, in particular to a method and device for determining a defect index of a transformer bushing and an electronic device. BACKGROUND

[0002] In related art, the transformer bushing is a key component in the power system, and its performance directly affects the normal operation of the transformer. The defect index of the transformer bushing can reflect the defect state of the transformer bushing, so as to find potential fault hazards in time and avoid power system interruption or equipment damage caused by bushing failure. However, in related art, when determining the defect index of the transformer bushing, there is a technical problem of inaccurate defect index determination.

[0003] For the above problems, no effective solution has been proposed so far. SUMMARY

[0004] The embodiments of the present application provide a method and device for determining a defect index of a transformer bushing and an electronic device, to at least solve the technical problem of inaccurate defect index determination in related art when determining the defect index of the transformer bushing.

[0005] According to an aspect of an embodiment of the present application, a method for determining a defect index of a transformer bushing is provided, comprising: obtaining bushing data of a to-be-detected bushing, wherein the bushing data comprises a bushing type and bushing operation data; determining a reference defect set corresponding to the bushing type, wherein the reference defect set comprises reference defect features corresponding to a plurality of reference bushings respectively; determining a first defect feature according to the bushing operation data, wherein the first defect feature represents a defect feature of the to-be-detected bushing; determining a second defect feature according to the reference defect features corresponding to the plurality of reference bushings respectively, wherein the second defect feature is a defect feature corresponding to the reference defect set; and determining a target defect index corresponding to the to-be-detected bushing according to the first defect feature and the second defect feature.

[0006] Optionally, the determining the second defect feature according to the reference defect features corresponding to the plurality of reference bushings respectively comprises: determining a production time period of the to-be-detected bushing; determining a production adjustment parameter corresponding to the bushing type within the production time period; determining an adjustment time point corresponding to the production adjustment parameter; dividing the production time period according to the adjustment time point to obtain a plurality of sub-time periods; and determining the second defect feature according to the plurality of sub-time periods and the reference defect set.

[0007] Optionally, the determining the second defect feature according to the plurality of sub-time periods and the reference defect set comprises: determining, from the plurality of reference casings, first target casings corresponding to the plurality of sub-time periods respectively; determining, according to the reference defect features corresponding to the first target casings in a first sub-time period, a first sub-population feature corresponding to the first sub-time period according to an execution order of the plurality of sub-time periods; determining, according to the first sub-population feature and the reference defect features corresponding to the first target casings in a next sub-time period, a next sub-population feature corresponding to the next sub-time period, until the plurality of sub-time periods are processed to obtain sub-population features corresponding to the plurality of sub-time periods respectively; and determining the second defect feature according to the sub-population features corresponding to the plurality of sub-time periods respectively.

[0008] Optionally, the determining the second defect feature according to the reference defect features corresponding to the plurality of reference casings respectively comprises: determining a target production batch of the casing to be detected; determining, according to the target production batch, a plurality of second target casings from the plurality of reference casings, wherein the plurality of second target casings are reference casings in the plurality of reference casings whose adjacent indexes between a reference production batch and the target production batch are greater than an adjacent threshold, and the reference production batch is a production batch of the corresponding reference casing; and determining the second defect feature according to the reference defect features corresponding to the plurality of second target casings respectively.

[0009] Optionally, the determining the second defect feature according to the reference defect features corresponding to the plurality of reference casings respectively comprises: determining a target function corresponding to the casing type, wherein the target function comprises a structure feature item, a life feature item and a defect distribution item, the structure feature item is an item corresponding to a casing structure feature, the life feature item is an item corresponding to a casing life feature, and the defect distribution item is an item corresponding to a casing defect distribution; and determining the second defect feature corresponding to the casing to be detected according to the reference defect set and the target function.

[0010] Optionally, the determining the target defect index corresponding to the casing to be detected according to the first defect feature and the second defect feature comprises: calling a target model, wherein the target model is trained according to a target matrix and sample data, the target matrix is used for dimensionality reduction processing of a non-key feature and dimensionality increase processing of a key feature, the key feature is a defect feature with a key index greater than a key threshold, and the non-key feature is a defect feature with a key index less than or equal to the key threshold; and inputting the first defect feature and the second defect feature into the target model to obtain the target defect index corresponding to the casing to be detected.

[0011] Optionally, the determining the reference defect set corresponding to the bushing type comprises: determining a plurality of reference bushings corresponding to the bushing type; determining reference data corresponding to the plurality of reference bushings respectively, wherein the corresponding reference data comprises text data, image data; and determining reference defect features corresponding to the plurality of reference bushings respectively according to the reference data corresponding to the plurality of reference bushings respectively, to obtain the reference defect set.

[0012] According to an aspect of an embodiment of the present application, there is provided a transformer bushing defect index determination apparatus, comprising: an acquisition module configured to acquire bushing data of a to-be-detected bushing, wherein the bushing data comprises a bushing type and bushing operation data; a first determination module configured to determine a reference defect set corresponding to the bushing type, wherein the reference defect set comprises reference defect features corresponding to a plurality of reference bushings respectively; a second determination module configured to determine a first defect feature according to the bushing operation data, wherein the first defect feature represents a defect feature of the to-be-detected bushing; a third determination module configured to determine a second defect feature according to the reference defect features corresponding to the plurality of reference bushings respectively, wherein the second defect feature is a defect feature corresponding to the reference defect set; and a fourth determination module configured to determine a target defect index corresponding to the to-be-detected bushing according to the first defect feature and the second defect feature.

[0013] According to an aspect of an embodiment of the present application, there is provided an electronic device, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement the transformer bushing defect index determination method of any one of the above.

[0014] According to an aspect of an embodiment of the present application, there is provided a computer-readable storage medium, comprising: instructions stored in the computer-readable storage medium, when executed by a processor of an electronic device, cause the electronic device to perform the transformer bushing defect index determination method of any one of the above.

[0015] In this embodiment of the invention, bushing data of the bushing to be inspected is acquired, wherein the bushing data includes bushing type and bushing operation data; a reference defect set corresponding to the bushing type is determined, wherein the reference defect set includes reference defect features corresponding to multiple reference bushings; based on the bushing operation data, a first defect feature is determined, wherein the first defect feature represents the defect feature of the bushing to be inspected; based on the reference defect features corresponding to multiple reference bushings, a second defect feature is determined, wherein the second defect feature is the defect feature corresponding to the reference defect set; based on the first defect feature and the second defect feature, a target defect index corresponding to the bushing to be inspected is determined. By combining the individual operation data of the bushing to be inspected with the group defect patterns of bushings of the same type, a two-dimensional analysis of individual features + group reference is achieved, comprehensively analyzing and quantifying the defect status of the bushing to be inspected, avoiding the one-sidedness of defect judgment, effectively improving the accuracy of defect assessment, and thus solving the technical problem in related technologies where the defect index determination of transformer bushings is inaccurate. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0017] Figure 1 This is a flowchart of a method for determining the defect index of a transformer bushing according to an embodiment of the present invention;

[0018] Figure 2 This is a structural block diagram of a transformer bushing defect index determination device according to an embodiment of the present invention. Detailed Implementation

[0019] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0020] It is to be understood that the terms "first", "second", and the like, used in the description and the claims of the application and the above description of the drawings merely refer to different categories and do not necessarily imply a sequence or order of execution. It is to be understood that the data thus used in the embodiments of the application described herein can be interchanged, where appropriate, so that the embodiments of the application described herein can be carried out in orders other than those illustrated or described herein. Also, the terms "comprise" and "have" and any variations thereof are intended to cover a non-exclusive inclusion, for example, a process, method, system, product, or apparatus that comprises a list of steps or units is not necessarily limited to those steps or units that are clearly listed, but can include other steps or units not clearly listed or inherent to such processes, methods, products, or apparatuses.

[0021] Embodiment 1

[0022] According to an embodiment of the application, an embodiment of a method for determining a defect index of a transformer bushing is provided. It should be noted that the steps shown in the flowchart of the 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 herein can be executed in an order different from that shown herein.

[0023] Figure 1 is a flowchart of a method for determining a defect index of a transformer bushing according to an embodiment of the application, as shown in Figure 1 the method comprises the following steps:

[0024] S102, obtaining bushing data of a bushing to be detected, wherein the bushing data comprises a bushing type and bushing operation data.

[0025] In the step S102 provided in the present application, the bushing data of the bushing to be detected is obtained, wherein the bushing data comprises the bushing type and the bushing operation data.

[0026] wherein the bushing to be detected is involved, which is a transformer bushing that needs to be evaluated for a defect index.

[0027] wherein the bushing data is involved, which is data reflecting the characteristics of the bushing to be detected.

[0028] wherein the bushing type is involved, which is the type corresponding to the bushing to be detected, which can be divided according to the model, function, scene, etc.

[0029] wherein the bushing operation data is involved, which is a data set for reflecting the operation state of the bushing.

[0030] The casing data of the casing to be detected is data reflecting casing characteristics of the casing to be detected. By obtaining the casing data, a data basis is provided for subsequent defect index evaluation.

[0031] In step S104, a reference defect set corresponding to the casing type is determined, wherein the reference defect set includes reference defect characteristics corresponding to a plurality of reference casings.

[0032] In step S104 provided in the present application, a reference defect set corresponding to the casing type is determined, wherein the reference defect set includes reference defect characteristics corresponding to a plurality of reference casings.

[0033] The reference defect set is a set of defect characteristic data of a plurality of reference casings, which are of the same type as the casing to be detected, and the defect characteristics thereof are used as a reference for evaluating the defect state of the casing to be detected. The purpose of the reference defect set is to provide a group characteristic benchmark to more accurately evaluate the defect index of the casing to be detected.

[0034] The plurality of reference casings are casings of the same type as the casing to be detected.

[0035] The reference defect characteristics are the defect characteristics of the corresponding reference casings.

[0036] The reference defect set includes reference defect characteristics corresponding to a plurality of reference casings corresponding to the casing type, which can provide a group defect characteristic benchmark of the same type as the casing to be detected, thereby enabling comprehensive analysis of the group characteristics of casings of the same type to avoid one-sidedness of single evaluation.

[0037] In step S106, a first defect characteristic is determined according to the casing operation data, wherein the first defect characteristic represents the defect characteristic of the casing to be detected.

[0038] In step S106 provided in the present application, a first defect characteristic is determined according to the casing operation data, wherein the first defect characteristic represents the defect characteristic of the casing to be detected.

[0039] The first defect characteristic is a specific characteristic extracted by analyzing the operation data of the casing to be detected, which can represent the current defect state of the casing, and reflects the abnormality or potential situation that may occur in the casing to be detected.

[0040] By analyzing the operation data (such as temperature, current, voltage, etc.), the first defect characteristic reflecting the current operation state of the casing is extracted, which provides a direct basis for subsequent defect evaluation based on individual operation state, ensures the accuracy and pertinence of defect judgment, and avoids misjudgment due to neglect of individual differences.

[0041] S108, determine the second defect feature according to the reference defect features corresponding to the plurality of reference casings.

[0042] In step S108 provided in the present application, the second defect feature is determined according to the reference defect features corresponding to the plurality of reference casings.

[0043] The second defect feature is obtained by comprehensive analysis of the reference defect features corresponding to the plurality of reference casings, and is used to reflect the group defect feature of the casing of the same type as the casing to be detected.

[0044] The second defect feature reflects the common defect feature of the casing group of the same type as the casing to be detected, thereby providing a group level reference benchmark for defect evaluation.

[0045] S110, determine the target defect index corresponding to the casing to be detected according to the first defect feature and the second defect feature.

[0046] In step S110 provided in the present application, the target defect index corresponding to the casing to be detected is determined according to the first defect feature and the second defect feature.

[0047] The target defect index is an index for quantifying the defect degree of the casing to be detected, which is obtained by comprehensively considering the first defect feature (reflecting the defect state of the individual casing to be detected) and the second defect feature (reflecting the defect law of the casing group of the same type as the casing to be detected).

[0048] By combining individual features and group experience, a more comprehensive and objective defect evaluation basis is provided, thereby improving the accuracy of defect judgment.

[0049] Through the steps S102-S110, the casing data of the casing to be detected is acquired, wherein the casing data comprises a casing type and casing operation data; a reference defect set corresponding to the casing type is determined, wherein the reference defect set comprises reference defect features corresponding to a plurality of reference casings respectively; a first defect feature is determined according to the casing operation data, wherein the first defect feature represents a defect feature of the casing to be detected; a second defect feature is determined according to the reference defect features corresponding to the plurality of reference casings respectively, wherein the second defect feature is a defect feature corresponding to the reference defect set; and a target defect index corresponding to the casing to be detected is determined according to the first defect feature and the second defect feature. By combining the individual operation data of the casing to be detected with the group defect regularity of the casings of the same type, two-dimensional analysis of individual features and group references is realized, the defect state of the casing to be detected is comprehensively analyzed and quantified, defect judgment one-sidedness is avoided, the accuracy of defect evaluation is effectively improved, and thus the technical problem that the defect index is not accurately determined when determining the defect index of the transformer casing in the related art is solved.

[0050] As an optional embodiment, determining the second defect feature according to the reference defect features corresponding to the plurality of reference casings comprises: determining a production time period of the casing to be detected; determining a production adjustment parameter corresponding to the casing type in the production time period; determining an adjustment time point corresponding to the production adjustment parameter; dividing the production time period according to the adjustment time point to obtain a plurality of sub-time periods; and determining the second defect feature according to the plurality of sub-time periods and the reference defect set.

[0051] In this embodiment, the specific steps of determining the second defect feature according to the reference defect features corresponding to the plurality of reference casings are described.

[0052] The production time period is a time interval from the production of the casing to be detected to the detection period.

[0053] The production adjustment parameter is a parameter for reflecting the adjustment of production process, material, quality control, etc. of the casing of the casing type in the production time period.

[0054] The adjustment time point is a specific time point of production adjustment according to the production adjustment parameter.

[0055] The sub-time period is a time period obtained by dividing the production time period according to the adjustment time point, and the production adjustment parameter remains unchanged in each sub-time period.

[0056] By subdividing the casing under different production adjustment parameters, the influence of different production stages on the casing quality can be accurately captured, providing a more detailed population reference benchmark for defect evaluation, enabling the evaluation to fully consider factors such as production process changes, thereby more accurately reflecting the defect state of the casing to be detected.

[0057] Further, production adjustment parameters (such as process optimization, material replacement) and adjustment time points are included in the population division basis, allowing the reference population to be further focused on the same production stage and the same process condition, avoiding population characteristic deviations caused by production process changes, significantly improving the relevance and pertinence of the population defect characteristics to the casing to be detected, and further helping to improve the accuracy of casing defect evaluation.

[0058] As an optional embodiment, the second defect feature is determined according to the plurality of sub-time periods and the reference defect set, comprising: determining a first target casing corresponding to each of the plurality of sub-time periods from the plurality of reference casings; determining a first sub-population feature corresponding to a first sub-time period according to the reference defect feature corresponding to the first target casing in the first sub-time period, in the order of execution of the plurality of sub-time periods; determining a next sub-population feature corresponding to a next sub-time period according to the first population feature and the reference defect feature corresponding to the first target casing in the next sub-time period, until the plurality of sub-time periods are processed, to obtain a sub-population feature corresponding to each of the plurality of sub-time periods; and determining the second defect feature according to the sub-population features corresponding to the plurality of sub-time periods.

[0059] In this embodiment, the specific steps of determining the second defect feature according to the plurality of sub-time periods and the reference defect set are described.

[0060] Among them, the first target casing is a reference casing produced in each sub-time period. For example, when there is only one adjustment time point, the production time period is divided according to the adjustment time point to obtain a first time period and a second time period, wherein the first time period is the time period before the adjustment time point, and the second time period is the time period after the adjustment time point; a plurality of first casings are determined from the plurality of reference casings, wherein the plurality of first casings are reference casings in the plurality of reference casings whose production time belongs to the first time period; a plurality of second casings are determined from the plurality of reference casings, wherein the plurality of second casings are reference casings in the plurality of reference casings whose production time belongs to the second time period.

[0061] Among them, the first sub-time period is the first sub-time period in the plurality of sub-time periods according to the order of execution.

[0062] Among them, the first sub-population feature is the population defect feature of the same type of casing in the first sub-time period.

[0063] wherein the next sub-time period is the next time period in the execution order.

[0064] wherein the next sub-population feature is the population defect feature of the same type of casing in the next sub-time period.

[0065] wherein the sub-population feature is the population defect feature of the same type of casing in each sub-time period.

[0066] By determining the first target casings corresponding to each sub-time period from the plurality of reference casings and sequentially analyzing the population defect features in each sub-time period, the second defect feature is ultimately determined, which can fully consider the influence of changes in the production process (such as process adjustment, material replacement, etc.) on the casing defects, thereby providing more accurate and representative population reference benchmarks for defect evaluation through phased and refined analysis of the defect rules of the casing population in different production stages.

[0067] As an optional embodiment, determining the second defect feature according to the reference defect features corresponding to the plurality of reference casings comprises: determining a target production batch of the casing to be detected; determining a plurality of second target casings from the plurality of reference casings according to the target production batch, wherein the plurality of second target casings are reference casings in the plurality of reference casings whose adjacent index between a reference production batch and the target production batch is greater than an adjacent threshold, and the reference production batch is the production batch of the corresponding reference casing; and determining the second defect feature according to the reference defect features corresponding to the plurality of second target casings.

[0068] In this embodiment, the specific steps of determining the second defect feature according to the reference defect features corresponding to the plurality of reference casings are described.

[0069] wherein the target production batch is the specific production batch to which the casing to be detected belongs. The production batch is usually used to identify the same batch of products produced in a specific time period, and has similar production process, material and quality control standards.

[0070] wherein the plurality of second target casings are reference casings selected from the plurality of reference casings and having high similarity with the target production batch of the casing to be detected. The adjacent index between the production batch of these casings and the target production batch is greater than the set adjacent threshold, indicating that they are close enough to the casing to be detected in terms of production conditions and can be used as a reference for evaluation.

[0071] The reference production batch is the specific production batch to which each reference sleeve belongs, and is used to identify the production time, process conditions, and quality standards of the reference sleeve.

[0072] The proximity index is used to measure the proximity between two production batches.

[0073] The proximity threshold is used to determine whether two production batches are close enough, thereby determining whether a certain reference sleeve is included in the range of the target sleeve.

[0074] Determining the target production batch of the to-be-detected sleeve and selecting a second target sleeve from the plurality of reference sleeves with a proximity index greater than the proximity threshold ensures that the selected second target sleeve is a reference sleeve with similar production conditions, thereby providing a more targeted and representative group reference for the to-be-detected sleeve.

[0075] As an optional embodiment, the second defect feature is determined according to the reference defect features corresponding to the plurality of reference sleeves, including: determining a target function corresponding to the sleeve type, wherein the target function includes a structure feature item, a service life feature item, and a defect distribution item, the structure feature item is an item corresponding to the structure feature of the sleeve, the service life feature item is an item corresponding to the service life feature of the sleeve, and the defect distribution item is an item corresponding to the defect distribution of the sleeve; and determining the second defect feature corresponding to the to-be-detected sleeve according to the reference defect set and the target function.

[0076] In this embodiment, the specific steps of determining the second defect feature according to the reference defect features corresponding to the plurality of reference sleeves are described.

[0077] The target function is a function used to determine the second defect feature, including a structure feature item, a service life feature item, and a defect distribution item.

[0078] The structure feature item is an item used to reflect the structure feature of the sleeve.

[0079] The service life feature item is an item used to reflect the service life feature of the sleeve.

[0080] The defect distribution item is a feature item used to reflect the distribution of defects of the sleeve.

[0081] The sleeve structure feature is a specific characteristic of the sleeve in physical structure.

[0082] The sleeve service life feature is a feature related to the service life of the sleeve.

[0083] wherein the casing defect distribution refers to the distribution of defects that occur in the casing during use.

[0084] By determining the objective function corresponding to the casing type (including the structure feature item, the service life feature item, and the defect distribution item), the structure characteristics, service life, and defect distribution of the casing can be comprehensively considered, so that the defect state of the casing to be detected can be more accurately reflected, multi-dimensional reference can be provided for defect evaluation, and the accuracy and reliability of the evaluation result can be improved.

[0085] As an optional embodiment, the target defect index corresponding to the casing to be detected is determined according to the first defect feature and the second defect feature, including: calling a target model, wherein the target model is obtained by training according to a target matrix and sample data, the target matrix is used for dimensionality reduction processing of non-key features and dimensionality reduction processing of key features, the key features are defect features with a key index greater than a key threshold, and the non-key features are defect features with a key index less than or equal to the key threshold; the first defect feature and the second defect feature are input into the target model to obtain the target defect index corresponding to the casing to be detected.

[0086] In this embodiment, the specific steps of determining the target defect index corresponding to the casing to be detected according to the first defect feature and the second defect feature are described.

[0087] wherein the target model is obtained by training according to a target matrix and sample data, and is a model used for determining the casing defect index.

[0088] wherein the target matrix is a matrix used for enhancing or weakening different features, specifically, dimensionality reduction processing is performed on key features and features that have less impact on defect evaluation.

[0089] wherein the dimensionality reduction processing refers to enhancing the key features to improve their weights in the model. Through dimensionality reduction processing, these features will be given higher weights when calculating the target defect index, so as to more significantly affect the final evaluation result.

[0090] wherein the key features are those features that have a greater impact on defect evaluation, and their key indexes are greater than a key threshold. These features usually include defect features directly related to the safe operation of the casing.

[0091] Among them, non-critical features are involved, which are features that have less impact on defect evaluation, and their critical index is less than or equal to the critical threshold. These features, although to some extent reflect the running state of the casing, have less impact on the final defect evaluation.

[0092] Among them, the dimension reduction processing is involved, which is to weaken the non-critical features to reduce their weight in the model. Through dimension reduction processing, the influence of these features on the calculation of the target defect index will be weakened, so as to avoid unnecessary interference with the evaluation results.

[0093] Among them, the critical index is involved, which is a quantitative index for measuring the importance of defect features to defect evaluation. The higher the critical index, the greater the impact of the feature on defect evaluation.

[0094] Among them, the critical threshold is involved, which is a preset threshold for determining whether a defect feature is a critical feature. If the critical index of a feature is greater than the critical threshold, the feature is considered a critical feature; otherwise, it is considered a non-critical feature. The setting of the critical threshold needs to be adjusted according to the specific application scenario and evaluation requirements.

[0095] By making the critical features (critical index greater than critical threshold) in the model have a higher weight according to the target matrix, so as to more significantly affect the final evaluation results, while the non-critical features (critical index less than or equal to critical threshold) reduce their interference with the evaluation results through dimension reduction processing, thereby optimizing the accuracy of the model defect evaluation.

[0096] As an optional embodiment, determining the reference defect set corresponding to the casing type comprises: determining a plurality of reference casings corresponding to the casing type; determining reference data corresponding to the plurality of reference casings respectively, wherein the corresponding reference data comprises text data and image data; and determining reference defect features corresponding to the plurality of reference casings respectively according to the reference data corresponding to the plurality of reference casings respectively, to construct the reference defect set.

[0097] In this embodiment, the specific steps of determining the reference defect set corresponding to the casing type are described.

[0098] Among them, the reference data is involved, which is various data used to analyze and determine the reference casing defect features, including but not limited to text data and image data, which provides detailed information about the reference casing for constructing the reference defect set.

[0099] Among them, the text data is involved, which is text form data related to the casing defects of the reference casing.

[0100] The image data is data in the form of an image related to a sleeve defect of a reference sleeve.

[0101] By comprehensively analyzing the reference data, the defect characteristics of the reference sleeve can be comprehensively understood, and an accurate reference defect set can be constructed, thereby providing a reliable reference for subsequent defect evaluation.

[0102] Based on the above embodiments and optional embodiments, an optional implementation is provided, which is specifically described as follows.

[0103] In the related art, the transformer sleeve is a key component in the power system, and its performance directly affects the normal operation of the transformer. The defect index of the transformer sleeve can reflect the defect state of the transformer sleeve, so as to discover potential fault hidden dangers in time and avoid power system interruption or equipment damage caused by sleeve failure. However, in the related art, when determining the defect index of the transformer sleeve, there is a technical problem of inaccurate defect index determination.

[0104] At present, no effective solution has been proposed for the above problems.

[0105] Therefore, in the optional embodiment of the present application, a defect index determination method for a transformer sleeve is provided, which can effectively solve the above technical problems.

[0106] S1, obtaining sleeve data of a to-be-detected sleeve, wherein the sleeve data includes a sleeve type and sleeve operation data;

[0107] S2, determining a reference defect set corresponding to the sleeve type, wherein the reference defect set includes reference defect characteristics corresponding to a plurality of reference sleeves respectively;

[0108] Further, S2 can further include:

[0109] determining a plurality of reference sleeves corresponding to the sleeve type; determining reference data corresponding to the plurality of reference sleeves respectively, wherein the corresponding reference data includes text data and image data; determining reference defect characteristics corresponding to the plurality of reference sleeves respectively according to the reference data corresponding to the plurality of reference sleeves respectively, to construct the reference defect set.

[0110] Taking the construction of a fault defect candidate library (i.e., the reference defect set) as an example.

[0111] 1) Collecting sleeve fault data: collecting transformer sleeve defect cases from various sources (such as books, technical reports, etc.), the collected cases including defect phenomenon description, defect cause analysis, disposal measures, etc. information, and the defect cases should not have obvious errors.

[0112] 2) Case sorting and screening: classify and organize cases according to casing structure and defect causes, and screen high-quality cases through typicality, repeatability, and rationality, and delete repetitive and unreasonable cases.

[0113] 3) Defect case labeling: label and classify defect cases according to the type and characteristics of casing defects, and classify them according to defect location, defect type, and defect severity.

[0114] 4) Construction of defect case library: store the labeled defect cases in the database, each case including defect type, defect phenomenon, defect occurrence time, defect location, and other information, and can add pictures, test data, etc.

[0115] S3, determining a first defect feature according to casing operation data, wherein the first defect feature represents the defect feature of the casing to be detected;

[0116] S4, determining a second defect feature according to the reference defect features corresponding to the plurality of reference casings, wherein the second defect feature is a defect feature corresponding to the reference defect set;

[0117] Further, S4 can further include:

[0118] determining the production time period of the casing to be detected; determining the production adjustment parameter corresponding to the casing type in the production time period; determining the adjustment time point corresponding to the production adjustment parameter; dividing the production time period according to the adjustment time point to obtain a plurality of sub-time periods; and determining the second defect feature according to the plurality of sub-time periods and the reference defect set.

[0119] Further, determining the second defect feature according to the plurality of sub-time periods and the reference defect set includes:

[0120] determining target casings corresponding to the plurality of sub-time periods from the plurality of reference casings; determining a first sub-population feature corresponding to a first sub-time period according to the reference defect feature corresponding to the target casing in the first sub-time period according to the execution order of the plurality of sub-time periods; determining a next sub-population feature corresponding to a next sub-time period according to the first population feature and the reference defect feature corresponding to the target casing in the next sub-time period, until the plurality of sub-time periods are processed to obtain a sub-population feature corresponding to the plurality of sub-time periods; and determining the second defect feature according to the sub-population features corresponding to the plurality of sub-time periods.

[0121] Further, S4 can further include:

[0122] Determine the target production batch of the casing to be detected; according to the target production batch, determine a plurality of target casings from a plurality of reference casings, wherein the plurality of target casings are reference casings in the plurality of reference casings, the proximity index between the reference production batch and the target production batch being greater than the proximity threshold, and the reference production batch is the production batch of the corresponding reference casing; and determine the second defect feature according to the reference defect features corresponding to the plurality of target casings.

[0123] Further, a production evaluation index system can also be established, and the casing to be monitored can be scored according to the importance of the index (1 point for extremely unimportant, 2 points for unimportant, 3 points for general, 4 points for relatively important, and 5 points for very important). In order to make the selection of the index more reliable, the statistical results are made The reliability analysis of the reliability coefficient method, the formula is:

[0124]

[0125] Among them:

[0126] is the evaluation result;

[0127] is the total number of evaluation dimensions;

[0128] is the within-item variance of the ith item score;

[0129] is the variance of the total item score.

[0130] From the above formula, it can be seen that The coefficient evaluates the consistency between the scores of each evaluation dimension, which belongs to the internal consistency coefficient.

[0131] After obtaining the above production evaluation index, normalization processing can also be performed.

[0132] In order to prevent the increase of network training time and the problem that the network cannot converge caused by the existence of singular sample data, the sample data set is normalized before training. Linear function normalization is used, and the formula is:

[0133]

[0134] Among them:

[0135] is the normalized data;

[0136] is the original data;

[0137] is the minimum value of the sample data;

[0138] is the maximum value of the sample data.

[0139] Further, S4 can further include:

[0140] determining a target function corresponding to the sleeve type, wherein the target function includes a structure feature item, a life feature item, and a defect distribution item, the structure feature item is an item corresponding to a structure feature of the sleeve, the life feature item is an item corresponding to a life feature of the sleeve, and the defect distribution item is an item corresponding to a defect distribution of the sleeve; and determining a second defect feature corresponding to the sleeve to be detected according to the reference defect set and the target function.

[0141] S5, determining a target defect index corresponding to the sleeve to be detected according to the first defect feature and the second defect feature.

[0142] Further, S5 can further include:

[0143] calling a target model, wherein the target model is obtained according to a target matrix, the target matrix is used for dimensionality reduction processing on a non-key feature and dimensionality reduction processing on a key feature, the key feature is a defect feature with a key index greater than a key threshold, and the non-key feature is a defect feature with a key index less than or equal to the key threshold; and inputting the first defect feature and the second defect feature into the target model to obtain the target defect index corresponding to the sleeve to be detected.

[0144] Specifically, the target model can be built by using an extreme learning machine, including feature extraction, digitization, normalization, and other preprocessing of indexes affecting the defect index evaluation of the transformer sleeve, and forming a training set together with the defect evaluation score of the sample transformer sleeve (the score is comprehensively scored according to historical experience, as the basis for model training). After selecting the sample transformer sleeve index value as the training input data training, selecting the defect score of the sample transformer sleeve as the training output data training.

[0145] Wherein, the sample data is trained by using an extreme learning machine (ELM), and an optimal model is established by adjusting algorithm parameters. The global optimal output weight of ELM can be written as:

[0146]

[0147] Wherein:

[0148] is the output weight matrix;

[0149] is the output matrix of the network;

[0150] , represents the generalized inverse matrix (Moore-Penrose) of the hidden layer output matrix , is represented as:

[0151]

[0152] , wherein:

[0153] is the activation function output of the Rth hidden layer neuron to the input

[0154] is the nth input sample;

[0155] is the connection weight of the Rth input layer and the hidden layer neuron, and is an n 1-dimensional vector;

[0156] is the threshold of the hidden layer neuron.

[0157] Linear regression in high-dimensional space adopts the principle of structural risk minimization to reduce the complexity of the model, and the calculation method is:

[0158]

[0159] , wherein:

[0160] is the adjusted hidden layer output matrix;

[0161] is the weight coefficient.

[0162] Continuously read the data to obtain the output result , is represented as:

[0163]

[0164] Based on this, the data of the casing to be detected is input into the input layer of the neural network, and the output of the output layer of the neural network is the defect index of the casing to be detected, so as to determine whether to select the casing to be detected. In addition, different extreme learning machine models can be established for different devices to which the casing to be detected is applied to evaluate and obtain the final score of the casing to be detected. The casing to be detected is selected according to the final score.

[0165] Through the above optional implementation manner, at least the following beneficial effects can be achieved:

[0166] ​​(1) Compared with the related art, the present application realizes two-dimensional analysis of individual characteristics + population reference by combining individual operation data of the to-be-detected bushing with population defect rules of the same type of bushing, comprehensively analyzes and quantifies the defect state of the to-be-detected bushing, avoids one-sidedness of defect judgment, effectively improves the accuracy of defect evaluation, and thus solves the technical problem of inaccurate defect index determination in the related art when determining the defect index of the transformer bushing.

[0167] (2) Compared with the related art, the present application can accurately capture the influence of different production stages on the quality of the bushing by subdividing the bushing under different production adjustment parameters, provides a more detailed population reference benchmark for defect evaluation, enables the evaluation to fully consider factors such as production process changes, and thus more accurately reflects the defect state of the to-be-detected bushing.

[0168] (3) Compared with the related art, the present application can fully consider the influence of changes in the production process (such as process adjustment, material replacement, etc.) on the defects of the bushing by determining the first target bushing corresponding to each sub-time period from multiple reference bushings and sequentially analyzing the population defect characteristics in each sub-time period to ultimately determine the second defect characteristics, thereby enabling the defect evaluation to provide a more accurate and more representative population reference benchmark by analyzing the defect rules of the bushing population in different production stages in stages and in detail.

[0169] It should be noted that, for the foregoing method embodiments, in order to simply describe, they are all expressed as a series of action combinations, but those skilled in the art should know that the present application is not limited by the action sequence described, because according to the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions and modules involved are not necessarily essential to the present application.

[0170] From the above description of the embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be realized by means of software and a general hardware platform as necessary, and of course, it can also be realized by hardware, but in many cases, the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as a ROM / RAM, a magnetic disk, or an optical disk), and includes a plurality of instructions for causing an end device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the method of each embodiment of the present application.

[0171] Example 2

[0172] According to the embodiment of the present application, a device for implementing the transformer bushing defect index determination method is also provided, Figure 2 is a structural block diagram of the transformer bushing defect index determination device according to the embodiment of the present application, as shown in the figure, the device comprises: an acquisition module 202, a first determination module 204, a second determination module 206, a third determination module 208 and a fourth determination module 210, which will be described in detail below. Figure 2

[0173] The acquisition module 202 is configured to acquire bushing data of a to-be-detected bushing, wherein the bushing data comprises a bushing type and bushing operation data;

[0174] The first determination module 204 is connected to the acquisition module 202 and configured to determine a reference defect set corresponding to the bushing type, wherein the reference defect set comprises reference defect features corresponding to a plurality of reference bushings respectively;

[0175] The second determination module 206 is connected to the first determination module 204 and configured to determine a first defect feature according to the bushing operation data, wherein the first defect feature represents a defect feature of the to-be-detected bushing;

[0176] The third determination module 208 is connected to the second determination module 206 and configured to determine a second defect feature according to the reference defect features corresponding to the plurality of reference bushings respectively, wherein the second defect feature is a defect feature corresponding to the reference defect set;

[0177] The fourth determination module 210 is connected to the third determination module 208 and configured to determine a target defect index corresponding to the to-be-detected bushing according to the first defect feature and the second defect feature.

[0178] It should be noted that the acquisition module 202, the first determination module 204, the second determination module 206, the third determination module 208 and the fourth determination module 210 correspond to steps S102 to S110 in the implementation of the transformer bushing defect index determination method, and the plurality of modules have the same instances and application scenarios as the corresponding steps, but are not limited to the content disclosed in the above embodiment 1.

[0179] Embodiment 3

[0180] According to another aspect of the embodiment of the present application, an electronic device is also provided, comprising: a processor; a memory for storing processor-executable instructions, wherein the processor is configured to execute the instructions to implement the transformer bushing defect index determination method of any one of the above.

[0181] Embodiment 4

[0182] ​According to another aspect of the embodiments of the present application, a computer readable storage medium is also provided, which, when instructions in the computer readable storage medium are executed by a processor of an electronic device, enables the electronic device to perform the transformer bushing defect index determination method of any one of the above.

[0183] The above-mentioned embodiment numbers of the present application are only for description, and do not represent the advantages or disadvantages of the embodiments.

[0184] In the above-mentioned embodiments of the present application, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0185] In several embodiments provided in the present application, it should be understood that the disclosed technical contents can be implemented by other manners. Among them, the above-mentioned device embodiments are only schematic, for example, the division of the units can be a logical function division, and in actual implementation, there can be another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units or modules shown or discussed can be indirect coupling or communication connection through some interfaces, units or modules, which can be electrical or other forms.

[0186] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed to multiple units. According to actual needs, part or all of the units can be selected to achieve the purpose of the present embodiment scheme.

[0187] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be realized in the form of hardware or in the form of software functional unit.

[0188] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application, essentially or in other words, the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions to make a computer device (which can be a personal computer, a server or a network device, etc.) execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: a U disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0189] The above is only the preferred embodiment of the present application, it should be pointed out that, for those skilled in the art, without departing from the principles of the present application, can make a number of improvements and refinements, these improvements and refinements should also be considered as the protection scope of the present application.

Claims

1. A method of determining a defect index of a transformer bushing, characterized by, The method comprises: obtaining casing data of a casing to be detected, wherein the casing data comprises a casing type and casing operation data; determining a reference defect set corresponding to the casing type, wherein the reference defect set comprises reference defect features corresponding to a plurality of reference casings respectively; determining a first defect feature according to the casing operation data, wherein the first defect feature represents a defect feature of the casing to be detected; determining a second defect feature according to reference defect features corresponding to the plurality of reference casings respectively, wherein the second defect feature is a defect feature corresponding to the reference defect set; determining a target defect index corresponding to the casing to be detected according to the first defect feature and the second defect feature.

2. The method of claim 1, wherein, The method comprises: determining a production time period of the casing to be detected; determining a production adjustment parameter corresponding to the casing type in the production time period; determining an adjustment time point corresponding to the production adjustment parameter; dividing the production time period according to the adjustment time point to obtain a plurality of sub-time periods; determining the second defect feature according to the plurality of sub-time periods and the reference defect set.

3. The method of claim 2, wherein, The method comprises: determining a first target casing corresponding to the plurality of sub-time periods from the plurality of reference casings respectively; determining a first sub-population feature corresponding to a first sub-time period according to reference defect features corresponding to the first target casing in the first sub-time period according to an execution order of the plurality of sub-time periods; determining a next sub-population feature corresponding to a next sub-time period according to the first sub-population feature and reference defect features corresponding to the first target casing in the next sub-time period, until the plurality of sub-time periods are processed to obtain sub-population features corresponding to the plurality of sub-time periods respectively; determining the second defect feature according to the sub-population features corresponding to the plurality of sub-time periods respectively.

4. The method of claim 1, wherein, The method comprises: determining a target production batch of the casing to be detected; determining a plurality of second target casings from the plurality of reference casings according to the target production batch, wherein the plurality of second target casings are reference casings in the plurality of reference casings whose reference production batch and the target production batch have a proximity index greater than a proximity threshold, and the reference production batch is a production batch corresponding to the reference casing; determining the second defect feature according to reference defect features corresponding to the plurality of second target casings respectively.

5. The method of claim 1, wherein, The method comprises: determining a target function corresponding to the casing type, wherein the target function comprises a structure feature item, a life feature item, and a defect distribution item, the structure feature item is an item corresponding to a casing structure feature, the life feature item is an item corresponding to a casing life feature, and the defect distribution item is an item corresponding to a casing defect distribution; Determine the second defect feature corresponding to the to-be-detected bushing according to the reference defect set and the target function.

6. The method of claim 1, wherein, The determining of the target defect index corresponding to the to-be-detected bushing according to the first defect feature and the second defect feature comprises: Accessing a target model, wherein the target model is trained according to a target matrix and sample data, the target matrix is used for dimensionality reduction processing of a non-key feature and dimensionality increase processing of a key feature, the key feature is a defect feature with a key index greater than a key threshold, and the non-key feature is a defect feature with a key index less than or equal to the key threshold; Inputting the first defect feature and the second defect feature into the target model to obtain the target defect index corresponding to the to-be-detected bushing.

7. The method according to any one of claims 1 to 6, characterized in that, The determining of the reference defect set corresponding to the bushing type comprises: Determining a plurality of reference bushings corresponding to the bushing type; Determining reference data corresponding to the plurality of reference bushings respectively, wherein the corresponding reference data comprises text data and image data; According to the reference data corresponding to the plurality of reference bushings respectively, determining reference defect features corresponding to the plurality of reference bushings respectively to construct a reference defect set.

8. An apparatus for determining a defect index of a transformer bushing, characterized by Comprise: An acquisition module configured to acquire bushing data of a to-be-detected bushing, wherein the bushing data comprises a bushing type and bushing operation data; A first determination module configured to determine a reference defect set corresponding to the bushing type, wherein the reference defect set comprises reference defect features corresponding to a plurality of reference bushings respectively; A second determination module configured to determine a first defect feature according to the bushing operation data, wherein the first defect feature represents a defect feature of the to-be-detected bushing; A third determination module configured to determine a second defect feature according to the reference defect features corresponding to the plurality of reference bushings respectively, wherein the second defect feature is a defect feature corresponding to the reference defect set; A fourth determination module configured to determine a target defect index corresponding to the to-be-detected bushing according to the first defect feature and the second defect feature.

9. An electronic device, comprising: Comprise: A processor; A memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the transformer bushing defect index determination method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, When the instructions in the computer readable storage medium are executed by the processor of the electronic device, the electronic device can execute the transformer bushing defect index determination method according to any one of claims 1 to 7.

Images