阅读说明：本技术 油纸绝缘老化状态的识别方法、装置、设备和存储介质 (Method, device, equipment and storage medium for identifying insulation aging state of oil paper ) 是由 李艳 田杰 梁兆杰 张大宁 于 2021-07-14 设计创作，主要内容包括：本申请涉及一种油纸绝缘老化状态的识别方法、装置、设备和存储介质。所述油纸绝缘老化状态的识别方法包括：获取待测油纸绝缘的平均聚合度,并获取所述待测油纸绝缘的频域介电谱特征值；调用预先训练完成的含水量识别模型对所述频域介电谱特征值处理得到所述待测油纸绝缘的含水量,所述含水量识别模型通过若干油纸绝缘样品对应的频域介电谱特征值和含水量训练得到；根据所述平均聚合度和所述含水量识别所述待测油纸绝缘的老化状态。该油纸绝缘老化状态的识别方法能够提高油纸绝缘老化状态的识别精度。(The application relates to a method, a device, equipment and a storage medium for identifying the insulation aging state of oiled paper. The method for identifying the insulation aging state of the oiled paper comprises the following steps: acquiring the average polymerization degree of the oil paper insulation to be detected, and acquiring a frequency domain dielectric spectrum characteristic value of the oil paper insulation to be detected; calling a pre-trained water content identification model to process the frequency domain dielectric spectrum characteristic value to obtain the water content of the oilpaper insulation to be detected, wherein the water content identification model is obtained by training the frequency domain dielectric spectrum characteristic value and the water content corresponding to a plurality of oilpaper insulation samples; and identifying the insulation aging state of the oil paper to be detected according to the average polymerization degree and the water content. The method for identifying the oil paper insulation aging state can improve the identification precision of the oil paper insulation aging state.)

1. A method for identifying the insulation aging state of oiled paper is characterized by comprising the following steps:

acquiring the average polymerization degree of the oil paper insulation to be detected, and acquiring a frequency domain dielectric spectrum characteristic value of the oil paper insulation to be detected;

calling a pre-trained water content identification model to process the frequency domain dielectric spectrum characteristic value to obtain the water content of the oilpaper insulation to be detected, wherein the water content identification model is obtained by training the frequency domain dielectric spectrum characteristic value and the water content corresponding to a plurality of oilpaper insulation samples;

and identifying the insulation aging state of the oil paper to be detected according to the average polymerization degree and the water content.

2. The method for identifying the aging state of the oilpaper insulation according to claim 1, wherein the identifying the aging state of the oilpaper insulation to be detected according to the average polymerization degree and the water content comprises the following steps:

determining a first difference value between the average polymerization degree and a reference polymerization degree, and if the absolute value of the first difference value is greater than a first preset threshold value, determining that the oil paper insulation to be tested is in a slight aging state, wherein the reference polymerization degree is the polymerization degree of the oil paper insulation in an unaged state;

and determining a second difference value between the water content and the reference water content, and if the absolute value of the second difference value is greater than a second preset threshold value, determining that the oil paper insulation to be tested is in a severe aging state, wherein the reference water content is the water content of the oil paper insulation in an unaged state.

3. The method for identifying the aging state of the oiled paper insulation according to claim 1, wherein the obtaining of the frequency domain dielectric spectrum characteristic value of the oiled paper insulation to be tested comprises:

acquiring a frequency domain dielectric spectrum test result of the oil paper insulation to be tested;

and determining a frequency domain dielectric spectrum characteristic value of the oil paper insulation to be tested according to the frequency domain dielectric spectrum test result, wherein the frequency domain dielectric spectrum characteristic value comprises at least one of a complex capacitance real part and a complex capacitance virtual part.

4. The method for identifying the aging state of the oiled paper insulation according to claim 1, wherein the step of obtaining the average polymerization degree of the oiled paper insulation to be detected comprises the following steps:

determining a first frequency characteristic curve of the oil paper insulation to be detected;

performing spectrum resolution on the first frequency characteristic curve to obtain a second frequency characteristic curve;

and determining the average polymerization degree of the insulation of the oil paper to be tested according to the second frequency characteristic curve.

5. The method for identifying the aging state of the oiled paper insulation, according to claim 4, wherein the step of performing de-spectroscopy on the first frequency characteristic curve to obtain a second frequency characteristic curve comprises the following steps:

determining a third difference value between a real complex capacitance part corresponding to each frequency and a stable value in the first frequency characteristic curve, wherein the stable value is a constant value of influence of polarization dominated by the dipole of the oilpaper insulation on polarization dominated by carriers;

and determining a second frequency characteristic curve of the oiled paper insulation according to the third difference.

6. The method for identifying the aging state of the oiled paper insulation according to any one of claims 1 to 5, wherein the training mode of the water content identification model comprises the following steps:

testing the frequency domain dielectric spectrums of a plurality of oiled paper insulation samples, and determining the frequency domain dielectric spectrum characteristic values corresponding to the oiled paper insulation samples according to the frequency domain dielectric spectrums;

acquiring the water content of the plurality of oiled paper insulation samples;

taking the frequency domain dielectric spectrum characteristic values corresponding to the plurality of oiled paper insulation samples and the water content as sample data;

and inputting the sample data into a neural network model for training to obtain the water content identification model.

7. The method for identifying the aging state of oiled paper insulation according to claim 6, further comprising:

dividing frequency domain dielectric spectrum characteristic values corresponding to the plurality of oilpaper insulation samples according to frequency intervals corresponding to the frequency domain dielectric spectrums of the plurality of oilpaper insulation samples;

normalizing the divided frequency domain dielectric spectrum characteristic values;

and constructing the sample data through the frequency domain dielectric spectrum characteristic value after normalization processing.

8. The utility model provides an identification means of insulating ageing state of oiled paper which characterized in that includes:

the acquisition module is used for acquiring the average polymerization degree of the oil paper insulation to be detected and acquiring the frequency domain dielectric spectrum characteristic value of the oil paper insulation to be detected;

the calling module is used for calling a pre-trained water content identification model to process the frequency domain dielectric spectrum characteristic value to obtain the water content of the oil paper insulation to be detected, and the water content identification model is obtained by training the frequency domain dielectric spectrum characteristic value and the water content corresponding to a plurality of oil paper insulation samples;

and the identification module is used for identifying the aging state of the insulation of the oil paper to be detected according to the average polymerization degree and the water content.

9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 7.

10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.

Technical Field

The application relates to the technical field of aging identification, in particular to a method, a device, equipment and a storage medium for identifying an aging state of oil paper insulation.

Background

Oil paper insulation is an important insulation mode and is always applied to power equipment such as large transformers. In the oil paper insulation operation process of the transformer, the oil paper insulation of the transformer is subjected to various external stress effects such as heat, electricity, machinery, chemistry and the like for a long time, so that the transformer breaks down.

At present, the direct and effective detection method for judging the aging state of the oil paper insulation equipment is to measure the average polymerization degree of the oil paper insulation, but in actual situations, the aging of the oil paper insulation is influenced by other factors. Therefore, if the average polymerization degree is used as a criterion, the degree of the aging state of the oil paper insulation cannot be accurately judged.

Disclosure of Invention

In view of the above, it is desirable to provide a method, an apparatus, a device, and a storage medium for identifying an aged state of oil-impregnated paper insulation, which can improve the accuracy of identifying the aged state of oil-impregnated paper insulation.

A method for identifying the insulation aging state of oiled paper comprises the following steps:

acquiring the average polymerization degree of the oil paper insulation to be detected, and acquiring a frequency domain dielectric spectrum characteristic value of the oil paper insulation to be detected;

calling a pre-trained water content identification model to process the frequency domain dielectric spectrum characteristic value to obtain the water content of the oilpaper insulation to be detected, wherein the water content identification model is obtained by training the frequency domain dielectric spectrum characteristic value and the water content corresponding to a plurality of oilpaper insulation samples;

and identifying the insulation aging state of the oil paper to be detected according to the average polymerization degree and the water content.

In one embodiment, the identifying the aging state of the to-be-tested oilpaper insulation according to the average polymerization degree and the water content includes:

determining a first difference value between the average polymerization degree and a reference polymerization degree, and if the absolute value of the first difference value is greater than a first preset threshold value, determining that the oil paper insulation to be tested is in a slight aging state, wherein the reference polymerization degree is the polymerization degree of the oil paper insulation in an unaged state;

and determining a second difference value between the water content and the reference water content, and if the absolute value of the second difference value is greater than a second preset threshold value, determining that the oil paper insulation to be tested is in a severe aging state, wherein the reference water content is the water content of the oil paper insulation in an unaged state.

In one embodiment, the obtaining of the frequency domain dielectric spectrum characteristic value of the insulation of the oil paper to be tested includes:

acquiring a frequency domain dielectric spectrum test result of the oil paper insulation to be tested;

and determining a frequency domain dielectric spectrum characteristic value of the oil paper insulation to be tested according to the frequency domain dielectric spectrum test result, wherein the frequency domain dielectric spectrum characteristic value comprises at least one of a complex capacitance real part and a complex capacitance virtual part.

In one embodiment, the obtaining the average degree of polymerization of the paper oil insulation to be measured includes:

determining a first frequency characteristic curve of the oil paper insulation to be detected;

performing spectrum resolution on the first frequency characteristic curve to obtain a second frequency characteristic curve;

and determining the average polymerization degree of the insulation of the oil paper to be tested according to the second frequency characteristic curve.

In one embodiment, the spectrally resolving the first frequency characteristic to obtain a second frequency characteristic includes:

determining a third difference value between a real complex capacitance part corresponding to each frequency and a stable value in the first frequency characteristic curve, wherein the stable value is a constant value of influence of polarization dominated by the dipole of the oilpaper insulation on polarization dominated by carriers;

and determining a second frequency characteristic curve of the oiled paper insulation according to the third difference.

In one embodiment, the training mode of the water content recognition model includes:

testing the frequency domain dielectric spectrums of a plurality of oiled paper insulation samples, and determining the frequency domain dielectric spectrum characteristic values corresponding to the oiled paper insulation samples according to the frequency domain dielectric spectrums;

acquiring the water content of the plurality of oiled paper insulation samples;

taking the frequency domain dielectric spectrum characteristic values corresponding to the plurality of oiled paper insulation samples and the water content as sample data;

and inputting the sample data into a neural network model for training to obtain the water content identification model.

In one embodiment, the method further comprises:

dividing frequency domain dielectric spectrum characteristic values corresponding to the plurality of oilpaper insulation samples according to frequency intervals corresponding to the frequency domain dielectric spectrums of the plurality of oilpaper insulation samples;

normalizing the divided frequency domain dielectric spectrum characteristic values;

and constructing the sample data through the frequency domain dielectric spectrum characteristic value after normalization processing.

An apparatus for identifying the state of aging of oiled paper insulation, comprising:

the acquisition module is used for acquiring the average polymerization degree of the oil paper insulation to be detected and acquiring the frequency domain dielectric spectrum characteristic value of the oil paper insulation to be detected;

the calling module is used for calling a pre-trained water content identification model to process the frequency domain dielectric spectrum characteristic value to obtain the water content of the oil paper insulation to be detected, and the water content identification model is obtained by training the frequency domain dielectric spectrum characteristic value and the water content corresponding to a plurality of oil paper insulation samples;

and the identification module is used for identifying the aging state of the insulation of the oil paper to be detected according to the average polymerization degree and the water content.

A computer device comprising a memory storing a computer program and a processor implementing the steps of the method described above when executing the computer program.

A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the above-mentioned method.

The method for identifying the aging state of the oil paper insulation comprises the steps of obtaining the average polymerization degree of the oil paper insulation to be detected, obtaining the frequency domain dielectric spectrum characteristic value of the oil paper insulation to be detected, calling a water content identification model which is trained in advance to process the frequency domain dielectric spectrum characteristic value to obtain the water content of the oil paper insulation to be detected, obtaining the water content identification model through the frequency domain dielectric spectrum characteristic value and the water content training corresponding to a plurality of oil paper insulation samples, and identifying the aging state of the oil paper insulation to be detected according to the average polymerization degree and the water content, wherein the aging state of the oil paper insulation to be detected is identified by a composite distinguishing mechanism consisting of the water content and the average polymerization degree, so that compared with the method which only depends on the average polymerization degree as a criterion, the problem that the identification is not accurate due to the fact that the aging state is identified only depending on the average polymerization degree is solved, the recognition accuracy of the insulation aging state of the oiled paper is improved.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments or the conventional technologies of the present application, the drawings used in the descriptions of the embodiments or the conventional technologies will be briefly introduced below, it is obvious that the drawings in the following descriptions are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic flow chart of a method for identifying an aging state of an oil paper insulation according to an embodiment;

FIG. 2 is a flowchart illustrating a refinement of step 110 of FIG. 1, according to an embodiment;

fig. 3 is a schematic structural diagram of an apparatus for identifying an aging state of an oiled paper insulation according to an embodiment.

Detailed Description

To facilitate an understanding of the present application, the present application will now be described more fully with reference to the accompanying drawings. Embodiments of the present application are set forth in the accompanying drawings. This application may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the description of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

It will be understood that, as used herein, the terms "first," "second," and the like may be used herein to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish one element from another.

Spatial relational terms, such as "under," "below," "under," "over," and the like may be used herein to describe one element or feature's relationship to another element or feature as illustrated in the figures. It will be understood that the spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures. For example, if the device in the figures is turned over, elements or features described as "below" or "beneath" other elements or features would then be oriented "above" the other elements or features. Thus, the exemplary terms "under" and "under" can encompass both an orientation of above and below. In addition, the device may also include additional orientations (e.g., rotated 90 degrees or other orientations) and the spatial descriptors used herein interpreted accordingly.

It will be understood that when an element is referred to as being "connected" to another element, it can be directly connected to the other element or be connected to the other element through intervening elements. Further, "connection" in the following embodiments is understood to mean "electrical connection", "communication connection", or the like, if there is a transfer of electrical signals or data between the connected objects.

As used herein, the singular forms "a", "an" and "the" may include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises/comprising," "includes" or "including," etc., specify the presence of stated features, integers, steps, operations, components, parts, or combinations thereof, but do not preclude the presence or addition of one or more other features, integers, steps, operations, components, parts, or combinations thereof. Several means one or more, and are not limited herein. Plural means two or more. Also, as used in this specification, the term "and/or" includes any and all combinations of the associated listed items.

In recent years, the power industry in China develops rapidly, the scale of a power grid exceeds the first world of the United states, and the installed capacity of power generation is the second world. The ever-expanding grid size also puts higher demands on the safe and reliable power supply for the operation of the power system. The transformer, especially a large power transformer, is used as the core of energy conversion and transmission in power transformation and distribution, is often the most important, critical and expensive electrical equipment in power transmission and transformation equipment, and the safe operation of the transformer is significant for guaranteeing the safety of a power grid. When a large transformer fails during operation, large-area power failure can be caused, and huge economic loss and negative effects can be caused due to the fact that the overhaul period of the large transformer is as long as more than half a year.

Practice proves that as most of insulating materials of the power transformer are organic materials, such as mineral oil, insulating paper, various organic synthetic materials and the like, compared with metal materials, the insulating materials are more easily damaged, and most of damage and faults of the power transformer are caused by damage of insulation. The research finds that the aging process of the insulating paper is the process of cellulose degradation, 1-4 glycosidic bonds (oxygen bridges) among glucose monomer units are broken, molecules are changed from long chains to short chains, and the average polymerization degree value is reduced. At present, the direct and effective detection method for judging the aging state of the oil paper insulation equipment is to measure the average polymerization degree of the oil paper insulation, but in actual situations, the aging of the oil paper insulation is influenced by other factors. Therefore, if the average polymerization degree is used as a criterion, the degree of the aging state of the oil paper insulation cannot be accurately judged.

The embodiment of the application provides a method for judging the insulation aging state of the oil paper, solves the problem of inaccurate identification caused by the fact that the aging state is identified only by means of average polymerization degree, and improves the identification precision of the insulation aging state of the oil paper.

The technical solutions proposed in the embodiments of the present application are described in detail below with reference to the accompanying drawings.

Referring to fig. 1, fig. 1 is a schematic flow chart of a method for identifying an aging state of an oiled paper insulation according to an embodiment. In one embodiment, as shown in fig. 1, the method for identifying the aging state of the oiled paper insulation includes steps 110 to 130.

And 110, acquiring the average polymerization degree of the oil paper insulation to be detected, and acquiring a frequency domain dielectric spectrum characteristic value of the oil paper insulation to be detected.

The oil paper insulation to be tested in the embodiment can be transformer oil paper insulation. The degree of polymerization is an index for measuring the molecular size of a polymer. In this example, the average degree of polymerization measures the molecular size of the polymer of the paper-oil insulation to be measured. Optionally, the frequency domain dielectric spectrum eigenvalue includes at least one of a real complex capacitance part and an imaginary complex capacitance part.

In one embodiment, the frequency domain dielectric spectrum test result of the oilpaper insulation is obtained by measuring through a frequency domain dielectric spectrum test instrument. And the computer equipment acquires a frequency domain dielectric spectrum test result of the oil paper insulation to be tested and determines a frequency domain dielectric spectrum characteristic value of the oil paper insulation to be tested according to the frequency domain dielectric spectrum test result. Specifically, the computer device extracts the real part and the imaginary part of the complex capacitance in the frequency domain dielectric spectrum test result, and stores the real part and the imaginary part as the frequency domain dielectric spectrum characteristic value in the frequency domain dielectric spectrum characteristic value storage unit.

And step 120, calling a water content identification model which is trained in advance to process the frequency domain dielectric spectrum characteristic value to obtain the water content of the oil paper insulation to be detected.

The water content identification model of the embodiment is obtained by training frequency domain dielectric spectrum characteristic values and water contents corresponding to a plurality of oiled paper insulation samples. Specifically, the frequency domain dielectric spectrum characteristic values and the water content corresponding to a plurality of oiled paper insulation samples are used as sample data and input into the neural network model for training, so that a water content identification model is obtained.

In the step, the frequency domain dielectric spectrum characteristic value obtained by the insulation test of the oil paper to be tested is input into the water content identification model, so that the water content of the oil paper to be tested can be directly predicted, and the water content does not need to be tested through a complicated materialization experiment.

And step 130, identifying the aging state of the to-be-tested oilpaper insulation according to the average polymerization degree and the water content.

In this step, the aging state of the oilpaper insulation is identified from the average degree of polymerization and the water content together.

In this embodiment, because the ageing state of this application is through the compound mechanism of distinguishing that comprises water content and average polymerization degree, compare in relying on average polymerization degree as the criterion alone, solved and relied on average polymerization degree alone to discern the ageing state and lead to discerning inaccurate problem, realized improving the discernment precision of oil paper insulation ageing state.

In one embodiment, identifying the aging state of the to-be-tested oilpaper insulation according to the average polymerization degree and the water content comprises:

determining a first difference value between the average polymerization degree and a reference polymerization degree, and if the absolute value of the first difference value is greater than a first preset threshold value, determining that the oil paper insulation to be tested is in a slight aging state, wherein the reference polymerization degree is the polymerization degree of the oil paper insulation in an unaged state;

and determining a second difference value between the water content and the reference water content, and if the absolute value of the second difference value is greater than a second preset threshold value, determining that the oil paper insulation to be tested is in a severe aging state, wherein the reference water content is the water content of the oil paper insulation in an unaged state.

In this embodiment, first, the model information of the oil paper insulation of the transformer to be tested is used to search the reference polymerization degree of the oil paper insulation of the transformer with the same model in the power grid database in the unaged state. And then, performing difference calculation on the average polymerization degree of the oil paper insulation of the transformer to be detected obtained by calculation in the embodiment of the application and the reference polymerization degree of the oil paper insulation of the transformer in the searched unaged state to obtain a first difference value, and calculating the absolute value of the first difference value. Judging whether the absolute value of the first difference is larger than a first preset threshold value or not, and if so, proving that the oil paper insulation of the transformer to be tested is in a slight aging state; if the insulation is smaller than the first preset threshold value, the transformer oil paper insulation to be tested is proved to have no aging state. At which point a further determination of the specific degree of aging is required.

Further, according to the model information of the transformer oil paper insulation to be detected again, the water content of the transformer oil paper insulation in the same model under the unaged state is searched in the database, then the water content obtained by prediction in the estimation model of the water content of the transformer oil paper insulation in the embodiment of the application is subjected to difference calculation with the reference water content of the transformer oil paper insulation in the unaged state, so that a second difference is obtained. And then judging whether the second difference is greater than a second preset threshold value, if so, verifying that the transformer is in a severe aging state, and timely overhauling the oil paper insulation of the transformer to be tested so as to avoid influencing the normal operation of the power grid.

In this embodiment, if it is detected that the absolute value of the first difference between the average degree of polymerization of the paper-oil insulation of the transformer to be measured and the reference degree of polymerization of the paper-oil insulation of the transformer in the unaged state is smaller than the first preset threshold, the prediction of the water content is not required. Because the aging of the oil paper insulation of the transformer to be tested can be reflected only from the average polymerization degree value, the maintenance treatment is not needed. A transformer oil paper insulation water content estimation model is introduced, and the problem that the specific aging degree cannot be accurately reflected is solved by the calculated average polymerization degree of the transformer oil paper insulation to be detected.

It should be noted that, if the second difference between the predicted moisture content of the transformer oil paper insulation and the reference moisture content of the transformer oil paper insulation in the unaged state is smaller than the second preset threshold, it cannot be determined what aging degree the transformer oil paper insulation to be tested is specifically in, but it can be determined that the transformer oil paper insulation to be tested is actually aged. Therefore, in this case, alarm processing is also required to prevent the accelerated aging of the oil paper insulation of the transformer to be detected in a short time in the future, which results in the failure of the maintenance work and great loss.

In the embodiment, different variation factors of the transformer oil paper insulation equipment in the aging process are comprehensively considered, and the single criterion of average polymerization degree is not used as an influencing factor for determining the aging state degree of the transformer oil paper insulation. But also takes the water content of the transformer oil paper insulation into consideration, establishes a novel average polymerization degree-water content comprehensive distinguishing mode, and obtains a more accurate distinguishing method of the transformer oil paper insulation aging state.

It should be noted that, because the oil paper insulation of the transformer is lossy, polarization loss occurs under the action of the alternating electric field. Therefore, the dielectric loss factor is one of the basic characteristics of the oil paper insulation material, and the influence of the dielectric loss factor in the oil paper insulation of the transformer cannot be ignored. In order to obtain a more accurate result, the embodiment of the application also takes the dielectric loss factor into consideration. In the embodiment of the application, the frequency domain dielectric spectrum tester comprises a testing unit of the dielectric loss factor, and the dielectric loss factor of the transformer oilpaper insulation under different frequencies can be tested. And the measured dielectric loss factor is also used as a frequency domain dielectric spectrum characteristic value and is stored in the frequency domain dielectric spectrum characteristic value storage unit.

It can be understood that the obtained frequency domain dielectric spectrum characteristic value refers to the loss of the oil paper insulation of the transformer, the obtained frequency domain dielectric spectrum characteristic value is more accurate, the correspondingly obtained water content is more accurate, and the identification of the aging state of the oil paper insulation is more accurate.

In one embodiment, the training of the moisture content recognition model includes:

testing the frequency domain dielectric spectrums of a plurality of oiled paper insulation samples, and determining the frequency domain dielectric spectrum characteristic values corresponding to the oiled paper insulation samples according to the frequency domain dielectric spectrums; acquiring the water content of the plurality of oiled paper insulation samples; taking the frequency domain dielectric spectrum characteristic values corresponding to the plurality of oiled paper insulation samples and the water content as sample data; and inputting the sample data into a neural network model for training to obtain the water content identification model.

Optionally, a part of the sample data is used as a test set, and another part is used as a verification set. The model is trained through a testing machine, and then the training result of the model is verified through a verification set.

In the embodiment of the application, at first, a plurality of transformer oil paper insulation samples are taken, and frequency domain dielectric spectrum is tested to obtain frequency domain dielectric spectrum characteristic values corresponding to the plurality of transformer oil paper insulation samples, wherein the frequency domain dielectric spectrum characteristic values of the plurality of transformer oil paper insulation samples include: and the real part of the complex capacitance, the imaginary part of the complex capacitance and the dielectric loss factor are stored in a data processing center.

Furthermore, the water content of the transformer oil paper insulation samples is measured, the water content of the insulation paper can be obtained by measuring the content of micro water in the insulation oil where the transformer oil paper insulation samples are located, and the water content of the transformer oil paper insulation samples can also be measured through other physical and chemical experiments.

In this embodiment, in order to ensure that sample data input to the neural network model is accurately learned and trained, it is necessary to perform multiple tests on frequency domain dielectric spectrum characteristic values and moisture contents corresponding to a plurality of transformer oil paper insulation samples, and screen out data with erroneous tests. And the computer equipment takes the frequency domain dielectric spectrum characteristic values corresponding to the plurality of transformer oil paper insulation samples after the plurality of tests and the water contents corresponding to the plurality of transformer oil paper insulation samples as sample data of the neural network model training. Optionally, the neural network model of the present embodiment includes, but is not limited to, a bp (back propagation) neural network, a Hopfield network (hassfield network), an ART network (ART network), and a Kohonen network (ad hoc feature mapping model). In one embodiment, the neural network model uses a BP neural network. The accuracy of recognition can be improved by training through the BP neural network.

In this embodiment, the frequency domain dielectric spectrum characteristic values of a plurality of transformer oilpaper insulation samples in a training set are assigned with water content labels, the frequency domain dielectric spectrum characteristic values of the plurality of transformer oilpaper insulation samples assigned with the labels are input into a neural network model, the neural network model identifies the labels on the frequency domain dielectric spectrum characteristic values of the plurality of transformer oilpaper insulation samples, and then the water content corresponding to each water content estimation label is simultaneously input into the neural network model for training, so as to complete the training process of the neural network model.

Optionally, the neural network model includes an encoding module and a decoding module, where the encoding module uses a hole separation convolution operation, that is, the hole convolution is applied to a depth separation convolution, thereby solving the problems that the operation is complex and the decoding operation needs to be completed in a long time due to the use of only the depth separation convolution in the past. The design of the cavity separation convolution can ensure that the neural network model can not only maintain the original performance unchanged, but also greatly reduce the operation complexity and optimize the training of the neural network model.

In this embodiment, the frequency domain dielectric spectrum characteristic values corresponding to the plurality of concentrated transformer oilpaper insulation samples are input into the transformer oilpaper insulation water content estimation model to obtain the predicted water contents corresponding to the plurality of transformer oilpaper insulation samples. And then, calculating the matching degree of the real water content and the predicted water content, and continuously adjusting the model according to the matching degree result.

Specifically, in the embodiment of the application, the transformer paper oil insulation moisture content estimation model has a large difference between the output predicted moisture content and the real moisture content in the previous test processes. Taking the ratio of the water content predicted each time to the real water content as the matching degree, and continuously adjusting the model parameters along with the increasing of the test times to enable the value of the matching degree to gradually approach to 1; i.e. the predicted moisture content is closer to the actual moisture content.

In one embodiment, after the neural network model is trained once to obtain the water content identification model, the matching degree result of the water content identification model to the water content identification needs to be tested. Stopping training when the matching degree result meets the requirement of the matching degree threshold, increasing the test times if the matching degree result does not meet the matching degree threshold, and adjusting the model parameters.

It can be understood that the neural network model is trained, and the training is stopped when the matching degree result of the neural network model reaches the matching threshold, so that the obtained water content recognition model is more accurate to the recognition result of the water content. The model can be optimized by testing the accuracy of the model for estimating the water content of the transformer oil paper insulation, so that the trained model can complete the work of predicting the water content of the transformer oil paper insulation.

In one embodiment, the sample data needs to be pre-processed before training with the sample data. Optionally, the step of preprocessing the sample data includes:

dividing frequency domain dielectric spectrum characteristic values corresponding to the plurality of oilpaper insulation samples according to frequency intervals corresponding to the frequency domain dielectric spectrums of the plurality of oilpaper insulation samples;

normalizing the divided frequency domain dielectric spectrum characteristic values;

and constructing the sample data through the frequency domain dielectric spectrum characteristic value after normalization processing.

In this embodiment, the frequency domain dielectric spectrum characteristic values corresponding to the plurality of transformer oilpaper insulation samples are divided according to the frequency intervals corresponding to the frequency domain dielectric spectra of the plurality of transformer oilpaper insulation samples, and this process may be regarded as grouping all the test frequencies into a plurality of frequency interval groups, for example: … … of 1Hz-10Hz, 10Hz-20Hz, 20Hz-30 Hz; and then, normalizing the divided frequency domain dielectric spectrum characteristic values, and constructing or updating sample data through the normalized frequency domain dielectric spectrum characteristic values. Therefore, sample data input into the neural network model can be more suitable for training and learning, and a better estimation model of the water content of the transformer oil paper insulation can be obtained.

Specifically, the normalization process is as follows: and carrying out data standardization processing on the mean value and the standard deviation of the data in the frequency domain dielectric spectrum characteristic values of the oil paper insulation samples of the transformer. So that the frequency domain dielectric spectrum characteristic values of the plurality of processed transformer oil paper insulation samples accord with standard normal distribution, namely the mean value is 0 and the standard deviation is 1. The frequency domain dielectric spectrum characteristic value data of the transformer oil paper insulation samples after normalization processing are easier to be utilized by a neural network, and the learning accuracy of the model can be improved.

In the embodiment of the application, frequency domain dielectric spectrum characteristic values corresponding to a plurality of transformer oilpaper insulation samples after normalization processing and water contents corresponding to the frequency domain dielectric spectrum characteristic values and the water contents are used as sample data, and the sample data are divided into a training set and a testing set according to the ratio of 8: 2.

Referring to fig. 2, fig. 2 is a flowchart illustrating a refinement of step 110 in fig. 1 according to an embodiment. In one embodiment, as shown in fig. 2, the step of obtaining the average degree of polymerization of the paper-oil insulation to be measured includes steps 210 to 230.

And step 210, determining a first frequency characteristic curve of the oil paper insulation to be tested.

In the embodiment of the application, the oil paper insulation of the transformer to be tested is disconnected with the power line, the frequency domain dielectric spectrum tester is reliably grounded, and the frequency domain dielectric spectrum tester is connected with the oil paper insulation of the transformer to be tested through the conducting wire. And after the instrument connection is completed, measuring the dielectric response parameters of the transformer oil paper insulation.

The frequency domain dielectric spectrum testing instrument adopted in the embodiment of the application is a DIRANA dielectric response analyzer developed by Omicron company, can simultaneously carry out a frequency domain dielectric spectrum test on the oil-paper insulation of the transformer to be tested through two channels, can acquire a frequency domain dielectric spectrum testing result more quickly, and needs to be matched with a computer to acquire tested data when in use.

In the embodiment of the application, computer equipment records the measured frequency domain dielectric spectrum test result of the oil paper insulation of the transformer, and then draws a first frequency characteristic curve according to the real part of the complex capacitance and the imaginary part of the complex capacitance corresponding to each frequency in the frequency domain dielectric spectrum test result, wherein the horizontal axis of the first frequency characteristic curve represents frequency, and the vertical axis represents the complex capacitance. And after the conversion of the first frequency characteristic curve is completed, displaying the first frequency characteristic curve of the transformer oilpaper insulated complex capacitor on a computer equipment screen.

It should be noted that, when testing, the frequency domain dielectric spectrum tester needs to be adjusted by preset parameters, including setting the peak value of the variable frequency voltage to 200V, setting the lowest test frequency to 0.1mHz, and setting the highest test frequency to 1 KHz. In the embodiment of the application, only the frequency domain dielectric spectrum in a certain frequency interval is tested, and the frequency domain dielectric spectrum under the full frequency is not required to be tested, so that the test time can be shortened. Wherein the preset parameters include any one or more of the following: a peak value of the variable frequency voltage, a highest test frequency, and a lowest test frequency.

Step 220, performing spectrum decomposition on the first frequency characteristic curve to obtain a second frequency characteristic curve.

In the embodiment of the application, the first frequency characteristic curve of the oil paper insulation of the transformer measured by the frequency domain dielectric spectrum testing instrument is not a complex capacitance frequency characteristic curve in a real state. This is because the dipole-dominated polarization is also affected by the carrier-dominated polarization.

Further, the polarization state of the dielectric in the oil paper insulation of the transformer is mainly divided into two cases, one is polarization dominated by dipoles and the other is polarization dominated by carriers. Since the dipole and the carrier are affected differently by frequency, the polarization of these two states usually shows a large difference. The change of the real part of the complex capacitance of the left frequency band of the polarization characteristic frequency dominated by the dipole is small relative to the polarization of a carrier, so that for a determined transformer oilpaper insulation, the influence of the real part of the complex capacitance of the interface polarization between the oilpaper on the real part of the complex capacitance of the dipole in different states can be considered as a constant value, and the polarization of the oilpaper interface of the oilpaper insulation can be separated from the polarization influence of the dipole by subtracting the constant value from the actually measured curve of the real part of the complex capacitance, so that the purpose of spectrum solution is achieved.

In the embodiment of the application, the first frequency characteristic curve of the complex capacitor of the oil paper insulation of the transformer to be tested is subjected to spectrum solution operation, specifically, in the first frequency characteristic curve, a stable value is subtracted from a real part of the complex capacitor corresponding to each frequency to obtain a third difference value, and then the frequency characteristic curve of the complex capacitor of the oil paper insulation of the transformer to be tested, namely, the second frequency characteristic curve, is determined again according to the third difference value. Wherein the stabilization value is a constant value of the influence of the oilpaper insulated dipole-dominated polarization on the carrier-dominated polarization. By performing spectrum resolution on the measured first frequency characteristic curve, errors of a final calculation result caused by the influence of a stable value can be effectively avoided.

It should be noted that the stable value of the influence of the polarization dominated by the dipole on the polarization dominated by the carrier can be obtained through field test, and can also be searched in a power grid database through the model of the oil paper insulation of the transformer to be tested, generally speaking, the stable value is kept to be recorded no matter the transformer is overhauled or leaves a factory every time, and if the value of the stable value can be directly obtained, the time cost can be saved.

It can be understood that the influence of the dipole polarization on the carrier polarization can be avoided by the spectrum resolving operation, so as to obtain a more accurate second frequency characteristic curve.

And step 230, determining the average polymerization degree of the to-be-tested oilpaper insulation according to the second frequency characteristic curve.

In this embodiment, the calculation formula of the average polymerization degree is:

Z＝(p1+p2lnx+p3ln2x+p4ln3x+p5y+p6y2)/(1+p7lnx+p8ln2x+p9y)，

the average polymerization degree of the oil paper insulation of the transformer to be measured can be calculated through the formula, wherein Z is the average polymerization degree; x is lgDP, and DP is the polymerization degree of the oil paper insulation paper board of the transformer to be measured; y is the temperature of the oil paper insulation of the transformer to be measured; p1, p2, p3, p4, p5, p6, p7, p8 and p9 are preset parameter values. In the embodiment of the present application, the preset parameter values are set to p1 ═ 7.54 × 103, p2 ═ 2.35 × 104, p3 ═ 2.47 × 104, p4 ═ 8.69 × 103, p5 ═ 6.57 × 10-2, p6 ═ 6.57 × 10-2, p7 ═ 1.04 × 10-2, p8 ═ 8.74 × 101, and p9 ═ 1.5 × 100.

It should be noted that, the average degree of polymerization of the oil paper insulation of the transformer to be measured is calculated, the parameters need to be changed for many times, the calculation needs to be performed for many times, and then the average degree of polymerization values obtained through the calculation for many times are calculated by taking the average value, so that accidental errors which easily occur in one calculation are reduced.

For example, the average polymerization degree may be calculated by adjusting temperature parameters of the transformer paper oil insulation and calculating the average polymerization degree at different temperature values, or by converting parameter values of p1, p2, p3, p4, p5, p6, p7, p8, and p9 and calculating the average polymerization degree under a plurality of different parameter sets.

It should be understood that although the various steps in the flowcharts of fig. 1-2 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 1-2 may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed in turn or alternately with other steps or at least some of the other steps or stages.

Referring to fig. 3, fig. 3 is a schematic structural diagram of an apparatus for identifying an aging state of an oiled paper insulation according to an embodiment. In one embodiment, as shown in fig. 3, an apparatus for identifying the aging state of the oiled paper insulation is provided, which includes an acquiring module 310, a calling module 320 and an identifying module 330. Wherein:

the obtaining module 310 is configured to obtain an average polymerization degree of an oil paper insulation to be detected, and obtain a frequency domain dielectric spectrum characteristic value of the oil paper insulation to be detected;

the calling module 320 is configured to call a pre-trained water content identification model to process the frequency domain dielectric spectrum characteristic value to obtain a water content of the oil paper insulation to be tested, where the water content identification model is obtained by training frequency domain dielectric spectrum characteristic values and water contents corresponding to a plurality of oil paper insulation samples;

the identification module 330 is configured to identify an aging state of the to-be-tested oilpaper insulation according to the average polymerization degree and the water content.

In one embodiment, the identification module 330 includes:

the first identification unit is used for determining a first difference value between the average polymerization degree and a reference polymerization degree, and if the absolute value of the first difference value is greater than a first preset threshold value, the oil paper insulation to be detected is determined to be in a slight aging state, and the reference polymerization degree is the polymerization degree of the oil paper insulation in an unaged state;

and the second identification unit is used for determining a second difference value between the water content and the reference water content, and if the absolute value of the second difference value is greater than a second preset threshold value, the oil paper insulation to be detected is determined to be in a severe aging state, and the reference water content is the water content of the oil paper insulation in an unaged state.

In one embodiment, the obtaining module 310 includes:

the first acquisition unit is used for acquiring a frequency domain dielectric spectrum test result of the insulation of the oil paper to be tested;

and the first determining unit is used for determining a frequency domain dielectric spectrum characteristic value of the to-be-tested oilpaper insulation according to the frequency domain dielectric spectrum test result, wherein the frequency domain dielectric spectrum characteristic value comprises at least one of a complex capacitance real part and a complex capacitance virtual part.

In one embodiment, the obtaining module 310 further comprises:

the second determining unit is used for determining a first frequency characteristic curve of the oil paper insulation to be detected;

the spectrum resolving unit is used for resolving the spectrum of the first frequency characteristic curve to obtain a second frequency characteristic curve;

and the third determining unit is used for determining the average polymerization degree of the to-be-measured oilpaper insulation according to the second frequency characteristic curve.

In one embodiment, the spectrum analysis unit is specifically configured to determine, in the first frequency characteristic curve, a third difference between a real complex capacitance part and a stable value corresponding to each frequency, where the stable value is a constant value of an influence of the polarization dominated by the dipole insulated by the oil paper on the polarization dominated by the carrier;

and determining a second frequency characteristic curve of the oiled paper insulation according to the third difference.

In one embodiment, the apparatus further includes a training module, configured to test frequency domain dielectric spectrums of the plurality of oilpaper insulation samples, and determine frequency domain dielectric spectrum characteristic values corresponding to the plurality of oilpaper insulation samples according to the frequency domain dielectric spectrums;

acquiring the water content of the plurality of oiled paper insulation samples;

taking the frequency domain dielectric spectrum characteristic values corresponding to the plurality of oiled paper insulation samples and the water content as sample data;

and inputting the sample data into a neural network model for training to obtain the water content identification model.

In one embodiment, the training module is further configured to divide the frequency domain dielectric spectrum characteristic values corresponding to the plurality of oilpaper insulation samples according to the frequency intervals corresponding to the frequency domain dielectric spectra of the plurality of oilpaper insulation samples;

normalizing the divided frequency domain dielectric spectrum characteristic values;

and constructing the sample data through the frequency domain dielectric spectrum characteristic value after normalization processing.

For specific definition of the device for identifying the aging state of the oiled paper insulation, reference may be made to the above definition of the method for identifying the aging state of the oiled paper insulation, and details are not described herein again. All or part of the modules in the device for identifying the aging state of the oiled paper insulation can be realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules. It should be noted that, in the embodiment of the present application, the division of the module is schematic, and is only one logic function division, and there may be another division manner in actual implementation.

In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having stored therein a computer program, the processor implementing the steps of the above-described method embodiments when executing the computer program.

In an embodiment, a computer-readable storage medium is provided, on which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the above-mentioned method embodiments.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database or other medium used in the embodiments provided herein can include at least one of non-volatile and volatile memory. Non-volatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical storage, or the like. Volatile Memory can include Random Access Memory (RAM) or external cache Memory. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others.

In the description herein, references to the description of "some embodiments," "other embodiments," "desired embodiments," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, a schematic description of the above terminology may not necessarily refer to the same embodiment or example.

The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.

The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.