CN117554762B - Transformer insulation part aging model building method, medium and system - Google Patents

Abstract

The present invention provides a method, medium and system for establishing an aging model of transformer insulation parts, belonging to the technical field of transformer insulation part aging detection, including: obtaining accelerated aging experimental data of transformer insulation parts under M groups of different experimental parameters; establishing an insulation part aging index calculation model and an insulation part aging index change model according to the experimental data; obtaining initial data when the insulation part to be tested is installed; obtaining an operating parameter set collected during the operation of the transformer with the insulation part to be tested installed; calculating the initial aging index of the insulation part to be tested according to the initial data; performing similarity analysis on each operating parameter in the operating parameter set and the M groups of experimental parameters, obtaining the experimental parameter with the highest similarity as the matching experimental parameter of each operating parameter; calculating the aging index change corresponding to each operating parameter; and calculating the current aging index of the insulation part to be tested. The present invention solves the technical problem of quantitative aging detection of transformer insulation parts during operation.

Description

A method, medium and system for establishing an aging model of transformer insulation components

Technical Field

The present invention belongs to the technical field of transformer insulation component aging detection, and in particular, relates to a transformer insulation component aging model establishment method, medium and system.

Background technique

Transformers are important equipment in power systems, and the aging of their insulators will directly affect the service life and operational reliability of transformers. At present, transformer insulation aging assessment mainly relies on the traditional excuse analysis method, that is, through the chemical analysis of the product of insulation paper, to evaluate the degree of thermal aging of insulation. However, this method has a long sampling and testing cycle and cannot achieve online monitoring. At the same time, domestic and foreign scholars have conducted a lot of research and exploration on insulation aging assessment. For different insulation materials, an aging model based on partial discharge has been established. This method can monitor the partial discharge of insulation online and reflect the insulation aging. However, partial discharge is not indicative of global aging, and quantitative evaluation is difficult. By detecting changes in the density, hardness, tensile strength and other changes of insulation materials, an aging model based on mechanical property attenuation has been established. However, this requires shutdown detection and cannot be achieved online. In addition, the aging of different parts is uneven, and the reliability of the results is poor. Advanced equipment such as infrared thermal imaging and X-ray tomography can directly observe the structural damage of insulation sheets. However, the equipment is expensive and difficult to promote. The results obtained are relatively qualitative, and quantitative evaluation is still difficult. In other words, it is difficult for existing technologies to solve the technical problem of quantitative detection of aging in transformer insulation during operation.

Summary of the invention

In view of this, the present invention provides a method, medium and system for establishing an aging model of transformer insulation components, which can solve the technical problem of quantitatively detecting the aging of transformer insulation components during operation.

The present invention is achieved in that:

A first aspect of the present invention provides a method for establishing an aging model of an insulating component of a transformer, wherein the method is used to calculate a current aging index of an insulating component according to operating parameters of the transformer, and comprises the following steps:

S10, obtaining M groups of accelerated aging test data of transformer insulation under different experimental parameters, wherein the experimental parameters include temperature and load, and the experimental data include elastic modulus, yield strength, Young's modulus, displacement, surface image and the number of fractures occurring during the experiment of the transformer insulation before and after the experiment, wherein each group of accelerated aging experiments lasts N hours, and the experimental parameters include constant parameters and variable parameters, wherein the constant parameters are experimental parameters that are stable within a constant threshold range during an aging experiment; and the variable parameters are experimental parameters that vary within a constant threshold range during an aging experiment;

S20. Establishing an insulation component aging index calculation model based on the experimental data, for calculating the insulation component aging index when the experiment is completed based on the experimental data;

S30, establishing an insulation component aging index variation model based on the experimental data, which is used to represent the aging index variation caused by different experimental parameters;

S40, obtaining initial data of the insulating component to be tested when it is installed, including elastic modulus, yield strength, and Young's modulus of the insulating component;

S50, obtaining an operating parameter set collected during the operation of the transformer equipped with the insulation component to be tested, wherein the operating parameter set includes multiple operating parameters, and the operating parameters include temperature and load data, wherein the collection time interval is 1s;

S60, calculating the initial aging index of the insulating component to be tested using an insulating component aging index calculation model according to the initial data;

S70, performing similarity analysis on each operating parameter in the operating parameter set and the M groups of experimental parameters, and obtaining the experimental parameter with the highest similarity as the matching experimental parameter for each operating parameter;

S80, calculating the aging index change amount corresponding to each operating parameter using the insulation component aging index change model according to the matching experimental parameters;

S90, calculating the current aging index of the insulating component to be tested by using the aging index change and the initial aging index.

On the basis of the above technical solution, the method for establishing a transformer insulation component aging model of the present invention can also be improved as follows:

Among them, the steps of establishing the insulation component aging index calculation model are specifically: using a machine learning algorithm, taking experimental parameters as input, and taking the elastic modulus, yield strength, Young's modulus, displacement, surface image before and after the experiment, and the number of fractures occurring during the experiment as target outputs, to establish a first regression model; linearly combining the first regression model to obtain a comprehensive aging index calculation model.

Among them, the steps of establishing the insulation component aging index change model are specifically: based on the experimental data, establishing a curve including the change of experimental parameters over time; calculating the change of aging index in each time period; establishing a second regression model, with the experimental parameters as input and the change of aging index in each time period as output; using the tuning algorithm to fit the second regression model to obtain the insulation component aging index change model.

The step of calculating the initial aging index of the insulating component to be tested using the insulating component aging index calculation model according to the initial data is specifically: bringing the initial data of the insulating component to be tested into the insulating component aging index calculation model; and the model calculation result is the initial aging index.

Wherein, in the step of performing similarity analysis between each operating parameter in the operating parameter set and the M groups of experimental parameters, the similarity analysis method adopted is cosine similarity.

Among them, the step of calculating the aging index change corresponding to each operating parameter using the insulating component aging index change model based on the matching experimental parameters is specifically: substituting the matching experimental parameters into the insulating component aging index change model; the model calculation result is the corresponding aging index change.

Among them, the step of calculating the current aging index of the insulating component to be tested using the aging index change and the initial aging index is specifically: accumulating the index changes corresponding to all operating parameters to obtain the total change; adding the total change to the initial aging index to obtain the current aging index.

Furthermore, M is an integer greater than 100; and N is a floating point number between 0.5 and 96.0.

A second aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores program instructions, and when the program instructions are executed, they are used to execute the above-mentioned method for establishing an aging model of transformer insulation components.

A third aspect of the present invention provides a system for establishing an aging model of transformer insulation components, which is used to calculate the current aging index of insulation components based on operating parameters of the transformer, and includes the above-mentioned computer-readable storage medium.

Compared with the prior art, the beneficial effects of the method, medium and system for establishing an aging model of transformer insulation parts provided by the present invention are as follows: the present invention comprehensively and directly observes the insulation performance attenuation law under various aging conditions by building an accelerated aging test platform. The machine learning algorithm is applied to train the model with rich aging test data. With the help of the real-time operating parameters such as temperature and load of the transformer, the aging model provided by the present invention is used to realize the online quantitative evaluation of the aging degree of transformer insulation parts. Compared with the traditional periodic detection method, it can not only quantitatively evaluate the current aging status, but also predict future change trends, provide a basis for operation and maintenance decisions, and solve the technical problem of quantitative detection of aging of transformer insulation parts during operation.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

FIG1 is a flow chart of the method provided by the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not require further definition and explanation in the subsequent drawings.

In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicating orientations or positional relationships are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

As shown in FIG. 1 , an embodiment of a method for establishing an aging model of a transformer insulation component is provided according to the first aspect of the present invention. In this embodiment, the method for calculating the current aging index of the insulation component according to the operating parameters of the transformer includes the following steps:

S10, obtaining M groups of accelerated aging test data of transformer insulation under different experimental parameters, wherein the experimental parameters include temperature and load, and the experimental data include elastic modulus, yield strength, Young's modulus, displacement, surface image and the number of fractures occurring during the experiment of the transformer insulation before and after the experiment, each group of accelerated aging test lasts N hours, and the experimental parameters include constant parameters and variable parameters, wherein the constant parameters refer to the experimental parameters being stable within a constant threshold range during an aging test; the variable parameters refer to the experimental parameters changing within a constant threshold range during an aging test;

S20. Establishing an insulation component aging index calculation model based on the experimental data, for calculating the insulation component aging index when the experiment is completed based on the experimental data;

S30, establishing an insulation component aging index variation model based on the experimental data, which is used to represent the aging index variation caused by different experimental parameters;

S40, obtaining initial data of the insulating component to be tested when it is installed, including elastic modulus, yield strength, and Young's modulus of the insulating component;

S50, obtaining an operating parameter set collected during the operation of the transformer equipped with the insulation component to be tested, wherein the operating parameter set includes multiple operating parameters, and the operating parameters include temperature and load data, wherein the collection time interval is 1s;

S60, calculating the initial aging index of the insulating component to be tested using an insulating component aging index calculation model according to the initial data;

S70, performing similarity analysis on each operating parameter in the operating parameter set and the M groups of experimental parameters, and obtaining the experimental parameter with the highest similarity as the matching experimental parameter for each operating parameter;

S80, calculating the aging index change amount corresponding to each operating parameter using the insulation component aging index change model according to the matching experimental parameters;

S90, calculating the current aging index of the insulating component to be tested by using the aging index change and the initial aging index, wherein M is an integer greater than 100; and N is a floating point number between 0.5 and 96.0.

The specific implementation methods of the above steps are described in detail below:

The specific implementation of step S10 is described as follows;

1) Prepare M groups of transformer insulation samples, with each group containing no less than 10 samples, to ensure that the results are representative.

2) Under exactly the same environmental conditions, measure the initial elastic modulus, yield strength, and Young's modulus of each group of samples and record them as E0, σ0, and μ0.

3) Configure M groups of different experimental parameter combinations, including temperature (T) and load (F). For each group, select the values of temperature and load, at least one of which is a variable parameter and the other is a constant parameter.

4) Place the sample in the test device and run M groups of accelerated aging tests in sequence. Each test lasts for N hours (N ≥ 100 hours), and the temperature T and load F are continuously recorded during the process.

5) During operation, various mechanical data such as displacement, surface image, etc. are collected through sensors. The fracture conditions are manually observed and recorded.

6) After the experiment, measure the modulus E1, σ1, and μ1 of each group of samples. Calculate the modulus change ΔE = E1-E0, Δσ, and Δμ.

7) Summarize and record the parameter combination, process data and result data of each set of experiments, where the variable parameters form a curve and the constant parameters form a single value.

The specific implementation of step S20 is described as follows;

1) Collect and organize all experimental data in S10.

2) Use machine learning algorithm, take experimental parameters as input and each change ΔE, Δσ, Δμ as target output, and establish a regression model.

3) Linearly combine the regression models to obtain a comprehensive aging index calculation model, which is used to indicate the aging degree of insulation parts at the end of the experiment.

4) Optionally, tune the algorithm model to improve the prediction accuracy and obtain the final version of the model for "calculating the insulation component aging index at the end of the experiment based on the experimental data".

The specific implementation of step S30 is described as follows;

1) Collect the process data in S10, including the curves of T and F changing with time.

2) Calculate the change in aging index in each period of time: ΔI1, ΔI2, etc. The specific method is:

Substitute T and F of each time period into the model established by S20 to calculate the aging index at both ends of the interval. The difference between the two indexes is the change ΔI of the aging index in the time period.

3) Establish a regression model with T and F as input and the exponential change ΔI in each time period as output.

4) Optionally, the algorithm is tuned to obtain a final version of the "aging index variation model corresponding to different experimental parameters".

The specific implementation of step S40 is described as follows;

1) Extract the factory test report of the insulation component from the database and obtain the initial elastic modulus E0, yield strength σ0, and Young's modulus μ0.

2) If there is no test report, the insulation must be tested before installation to measure E0, σ0, and μ0.

3) Record and retain the initial data of the insulation component.

The specific implementation of step S50 is described as follows;

1) During the operation of the transformer, sensors are set to collect temperature T and load F in real time.

2) The sensor acquisition frequency is set to 1 Hz, that is, data is collected once every 1 second.

3) The sensor is connected to the data acquisition system and transmits the collected data to the system in real time.

4) The data acquisition system continuously receives and stores the sensor data.

5) Finally, the time series data {T(t), F(t)} during the operation of the transformer is obtained, including the temperature T and load F at multiple time points t.

The specific implementation of step S60 is described as follows;

1) Take out the recorded initial data of the insulating component, namely, elastic modulus E0, yield strength σ0, and Young's modulus μ0.

2) Substitute E0, σ0, and μ0 into the “calculate aging index based on experimental data” model established by S20.

3) The calculation result of the insulation component aging index calculation model is the initial aging index I0 of the insulation component.

4) Record I0 for subsequent calculations.

The specific implementation of step S70 is described as follows;

1) Take out the operating parameter time series {T(t), F(t)} obtained in S50.

2) Take out the M groups of experimental parameter combinations of S10.

3) For each set of experimental parameters, calculate its similarity with the operating parameter segment {T(t), F(t)}, using a specific method such as Euclidean distance algorithm; where the operating parameter segment is the operating parameters within a continuous collection time period.

4) Repeat step 3 to calculate the similarity for all fragments {T(t), F(t)} and all experimental parameter groups.

5) For each running parameter fragment, find the most similar (highest similarity) experimental parameter combination.

6) This experimental parameter combination is the matching experimental parameter of the running parameter segment.

The specific implementation of step S80 is described as follows;

1) Take out the matching experimental parameters corresponding to each operating parameter segment {T(t), F(t)} found in step S70.

2) Substitute each matching experimental parameter into the “aging index change model” established in S30.

3) The result of the modular calculation of the aging index change of the insulation component is the corresponding aging index change ΔI.

4) Repeat steps 1-3 to obtain the aging index change ΔI corresponding to each operating parameter segment {T(t), F(t)}.

The specific implementation of step S90 is described as follows;

1) Take out the aging index change ΔI corresponding to all operating parameter segments {T(t), F(t)}.

2) Take out the initial aging index I0 calculated by S60.

3) By adding up all ΔI, we can get the total change ΔIΣ of the aging index from calibration to the current moment.

4) Add ΔIΣ to I0 to obtain the current aging index It of the insulation component.

5) By comparing the current aging index It with the initial aging index I0, the aging degree of the insulation can be evaluated.

In summary, this method has established an aging index calculation model and an aging index change model through a large number of accelerated aging experiments. In actual operation, by matching the experimental parameters, the increment of the aging index can be obtained, and then the current aging degree can be calculated.

The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (10)

1. The method for establishing the aging model of the transformer insulating part is characterized by comprising the following steps of:

S10, acquiring accelerated aging test data of a transformer insulating part under M groups of different test parameters, wherein the test parameters comprise temperature and load, the test data comprise elastic modulus, yield strength, young modulus, displacement, surface images and the number of breaks in the test process of the transformer insulating part, the time length of each group of accelerated aging test is N hours, the test parameters comprise constant parameters and variable parameters, and the constant parameters are that the test parameters are stable within a constant threshold range in one aging test process; the variation parameter is a constant threshold range of the experimental parameter in the aging experimental process;

s20, establishing an insulation part ageing index calculation model according to experimental data, wherein the insulation part ageing index calculation model is used for calculating an insulation part ageing index when an experiment is completed according to the experimental data;

S30, establishing an insulation part ageing index change model according to experimental data, wherein the model is used for representing ageing index change amounts caused by different experimental parameters;

s40, acquiring initial data of the insulating piece to be tested during installation, wherein the initial data comprise the elastic modulus, the yield strength and the Young modulus of the insulating piece;

S50, acquiring an operation parameter set acquired in the operation process of the transformer provided with the insulating piece to be detected, wherein the operation parameter set comprises a plurality of operation parameters, the operation parameters comprise temperature and load data, and the acquired time interval is 1S;

S60, calculating an initial ageing index of the insulating piece to be tested by using an insulating piece ageing index calculation model according to the initial data;

S70, carrying out similarity analysis on each operation parameter in the operation parameter set and the M groups of experiment parameters to obtain the experiment parameter with the highest similarity as a matched experiment parameter of each operation parameter;

S80, calculating the aging index variation corresponding to each operation parameter by using an insulation part aging index variation model according to the matching experiment parameters;

And S90, calculating the current ageing index of the to-be-measured insulating piece by using the ageing index variation and the initial ageing index.

2. The method for building an aging model of a transformer insulation according to claim 1, wherein the step of building an aging index calculation model of the insulation comprises the following steps: using a machine learning algorithm, taking experimental parameters as input, taking elastic modulus, yield strength, young modulus, displacement, surface images before and after an experiment and the number of times of fracture in the experimental process as target output, and establishing a first regression model; and linearly combining the first regression models to obtain the comprehensive ageing index calculation model.

3. The method for building an aging model of a transformer insulation part according to claim 1, wherein the step of building an aging index change model of the insulation part comprises the following steps: establishing a change curve comprising experimental parameters along with time according to experimental data; calculating the aging index variation in each period of time; establishing a second regression model, wherein experimental parameters are taken as input, and aging index variation in each time period is taken as output; and fitting the second regression model by using a tuning algorithm to obtain an insulation part aging index change model.

4. The method for building an aging model of a transformer insulation according to claim 1, wherein the step of calculating the initial aging index of the insulation to be tested by using the insulation aging index calculation model according to the initial data comprises the following steps: the initial data of the insulating piece to be tested is brought into an insulating piece ageing index calculation model; the model calculation result is the initial ageing index.

5. The method for building a transformer insulation aging model according to claim 1, wherein in the step of performing similarity analysis on each of the operation parameters in the operation parameter set and the M sets of experimental parameters, the similarity analysis method is cosine similarity.

6. The method for building an aging model of a transformer insulation part according to claim 1, wherein the step of calculating the aging index variation corresponding to each operation parameter by using the aging index variation model of the insulation part according to the matching experiment parameters comprises the following steps: substituting the matching experimental parameters into an insulation part aging index change model; the model operation result is the corresponding aging index variation.

7. The method for building an aging model of a transformer insulation according to claim 1, wherein the step of calculating the current aging index of the insulation to be tested by using the aging index variation and the initial aging index comprises the following steps: accumulating index variation corresponding to all the operation parameters to obtain total variation; the total change is added to the initial aging index, i.e., the current aging index.

8. The method for modeling aging of a transformer insulation according to any one of claims 1 to 7, wherein M is an integer greater than 100; n is a floating point number of 0.5 to 96.0.

9. A computer readable storage medium, wherein program instructions are stored in the computer readable storage medium, which program instructions, when executed, are adapted to carry out the method for modeling the aging of a transformer insulation according to any one of claims 1-8.

10. A transformer insulation aging model building system for calculating a current aging index of an insulation based on operating parameters of a transformer, comprising the computer readable storage medium of claim 9.