A method for evaluating the compatibility of bopp film with impregnant

Abstract

The present application relates to the technical field of capacitor insulation material evaluation, and particularly relates to a BOPP film and impregnant compatibility evaluation method, comprising the following steps: performing compatibility experiments on the BOPP film and the impregnant, collecting comprehensive performance data samples of the BOPP film and the impregnant before and after the experiments; calculating a deep layer impregnation efficiency index DIEE; calculating an affinity index and a relative energy density of the BOPP film and the impregnant by means of the Hansen solubility parameter method; normalizing the comprehensive performance data samples and the three calculated indexes, and taking them as a data set; constructing a gradient boosting tree regression prediction model and training it to obtain a trained compatibility evaluation model; evaluating the compatibility of the BOPP film and the impregnant to obtain a compatibility evaluation score, and determining a compatibility grade according to the compatibility evaluation score. The present application remedies the defects of traditional single-index evaluation and pure theoretical prediction, and makes the compatibility evaluation process more efficient, the prediction more accurate, and the mechanism more clear.

Description

Technical Field

[0001] This invention relates to the field of capacitor insulation material evaluation technology, and in particular to a method for evaluating the compatibility of BOPP film with impregnating agent. Background Technology

[0002] Bis-directional polypropylene (BOPP) film, due to its excellent dielectric properties and mechanical strength, has become the core solid dielectric of power capacitors. To ensure the long-term reliability of capacitor operation, impregnating agents are used to fill the gaps between the films to eliminate partial discharge and improve heat dissipation. The compatibility between the BOPP film and the impregnating agent directly determines the dielectric strength, aging life, and long-term stability of the composite insulation system. Poor compatibility can lead to excessive swelling of the film by the impregnating agent, extraction of additives from the film, or accelerated aging of the impregnating agent itself, ultimately resulting in a sharp decline in the electrical performance of the capacitor.

[0003] Currently, the methods used in industry and academia to assess compatibility have several limitations. Traditional methods often rely on testing a single or a few endpoint performance indicators, such as comparing only the change in the breakdown field strength of the film before and after impregnation. This assessment has a single dimension, lacks systematicity and mechanistic correlation, and is difficult to reveal the underlying causes of performance degradation. More importantly, these methods heavily depend on accelerated thermal aging experiments that last for hundreds or even thousands of hours, which are inefficient and costly, significantly delaying the research and selection of new materials. Furthermore, existing methods are essentially "post-hoc verification," lacking forward-looking theoretical prediction tools that can quickly screen the compatibility potential of material combinations and provide risk warnings in the early stages of experiments.

[0004] Hansen's solubility parameter theory provides a valuable theoretical framework for predicting material compatibility at the molecular level, and the calculated relative energy density parameter is an important indicator for judging thermodynamic compatibility. However, this theory only makes qualitative or semi-quantitative predictions based on chemical affinity, and cannot quantitatively assess the comprehensive performance of materials under actual multi-field coupling conditions such as electricity, heat, and force, nor can it provide accurate quantitative values ​​for performance degradation. Therefore, it is difficult to apply directly in engineering practice.

[0005] Therefore, there is an urgent need for a compatibility assessment method that can integrate mechanism prediction and performance verification, while balancing efficiency and accuracy. An ideal method should be able to quickly screen risks through theoretical calculations in the early stages, comprehensively characterize them using multi-dimensional indicators in the experimental phase, and ultimately establish a precise and quantitative mapping relationship from the intrinsic properties of materials to the overall performance of the system through intelligent algorithms. This would overcome the bottlenecks of existing technologies and achieve rapid, reliable, and mechanistically clear compatibility assessments. Summary of the Invention

[0006] To address the shortcomings of existing technologies in evaluating the compatibility of BOPP films with impregnating agents, such as a single evaluation dimension, heavy reliance on long-term aging experiments, and a lack of quantitative prediction tools that correlate intrinsic material properties with macroscopic performance, this invention provides a method for evaluating the compatibility of BOPP films with impregnating agents. This method is based on the coupling of the comprehensive performance of the BOPP film and the impregnating agent with the Hansen solubility parameter. The specific technical solution is as follows: A method for evaluating the compatibility of BOPP film with impregnating agent, comprising the following steps: Step S1: Conduct a compatibility test on the BOPP film and the impregnating agent, and collect comprehensive performance data samples of the BOPP film and the impregnating agent before and after the experiment; the comprehensive performance of the BOPP film includes electrical properties, thermal properties, mechanical properties and physicochemical properties, and the comprehensive performance of the impregnating agent includes the breakdown field strength and dielectric loss of the impregnating agent; among which, the physicochemical properties of the BOPP film include the density and thickness of the BOPP film. Step S2: Calculate the depth impregnation efficiency index (DIEE) based on the density and thickness of the BOPP film before and after the experiment; calculate the affinity index between the BOPP film and the impregnating agent using the Hansen solubility parameter method. and relative energy density ; Step S3: Combine the comprehensive performance data sample collected in step S1 with the deep impregnation efficiency index (DIEE) and affinity index calculated in step S2. and relative energy density Normalize the dataset and use it as a dataset; Step S4: Construct a gradient boosting tree regression prediction model, train the model using input features, and obtain a trained compatibility evaluation model. Step S5: Use the trained compatibility assessment model to assess the compatibility between the BOPP film and the impregnating agent, obtain a compatibility assessment score, and determine the compatibility level based on the compatibility assessment score.

[0007] Preferably, the electrical properties of the BOPP film in step S1 include breakdown field strength and dielectric loss; the thermal properties include melting temperature and crystallinity; and the mechanical properties include tensile strength and elongation at break.

[0008] Preferably, the Deep Impregnation Efficiency Index (DIEE) in step S2 is calculated as follows: ; in, ρ 0 indicates the density of the BOPP film before impregnation; ρi t0 represents the density of the BOPP film after impregnation; t0 represents the thickness of the BOPP film before impregnation; ti represents the thickness of the BOPP film after impregnation.

[0009] Preferably, the affinity index between the BOPP film and the impregnating agent in step S2 is... The calculation method is as follows: ; in, , , These are the dispersion force, polar force, and hydrogen bonding force components, respectively; subscript 1 represents BOPP film, and subscript 2 represents impregnating agent.

[0010] Preferably, the relative energy density of the BOPP film and the impregnating agent in step S2 is... The calculation method is as follows: ; Where R0 is the radius of the dissolution sphere of the BOPP film in the Hansen solubility parameter space.

[0011] Preferably, the normalization process in step S3 is specifically performed using the minimum-maximum normalization method, and the formula for calculating the standardized value is: ; in, X max and X min represents the maximum and minimum values ​​of the data sample on this indicator, x represents the corresponding indicator, and xnorm represents the normalized value of the corresponding indicator.

[0012] Preferably, step S4 specifically includes the following steps: The normalized dataset is divided into training, validation, and test sets; the gradient boosting tree regression prediction model is initialized; based on the training and validation sets, iterative training is performed with the goal of reducing the preset loss function to build a compatibility evaluation model; Input a test set, compare the predicted compatibility evaluation score output by the compatibility evaluation model with the benchmark compatibility evaluation score obtained based on the test set, and calculate the accuracy evaluation index of the compatibility evaluation model based on the comparison result. If all accuracy evaluation indicators reach or exceed the preset accuracy threshold, the compatibility evaluation model is considered to have completed training, and a trained compatibility evaluation model is obtained. If any key accuracy evaluation indicator fails to reach the preset accuracy threshold, the model training phase is returned, the model parameters are adjusted, and the model is retrained until the accuracy of the compatibility evaluation model meets the requirements.

[0013] Preferably, the accuracy evaluation index includes at least the coefficient of determination R² and the mean absolute error (MAE) or mean absolute percentage error (MAPE) between the predicted and actual values.

[0014] Preferably, the loss function is the least squares loss function, as follows: ; in, L ( y i , F ( X i The loss function is used to measure the error between the predicted value and the true value of a single sample. y i For the first i The true compatibility evaluation value of each sample; F ( x i ) for the current compatibility assessment model for the first i Predicted compatibility assessment scores for each sample; x i Let be the input feature vector of the i-th sample.

[0015] Preferably, the evaluation function of the compatibility evaluation model is: ; in, y The compatibility assessment score is the output of the compatibility assessment model. F 0( X () represents the initial predicted value of the gradient boosting tree regression prediction model; η To increase the learning rate of the tree regression prediction model for gradient gradation; f m ( X ) is the first m A regression decision tree for input features X The predicted output value.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: The evaluation of this invention uses multi-dimensional performance data before and after impregnation and Hansen solubility parameters as fusion input features, and adopts a gradient boosting tree regression model to establish a nonlinear mapping from features to comprehensive score. At the same time, it proposes innovative characterization indicators such as deep impregnation efficiency index, which makes up for the shortcomings of traditional single indicator evaluation and pure theoretical prediction, making the compatibility evaluation process more efficient, the prediction more accurate, and the mechanism clearer. Attached Figure Description

[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0018] Figure 1 This is a flowchart of the evaluation method of the present invention.

[0019] Figure 2 This is a comparison chart of the predicted and actual final evaluation scores in an embodiment of the present invention.

[0020] Figure 3 This is a schematic diagram illustrating the error rate of the predicted score in an embodiment of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0023] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0024] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0025] like Figure 1As shown, this embodiment provides a method for evaluating the compatibility of BOPP film with impregnating agents. The method first collects multi-dimensional performance data before and after impregnation through experiments, and calculates the deep impregnation efficiency index and Hansen parameter as theoretical indicators. Then, the above data is fused to construct a feature vector, which, after normalization, is trained using a gradient boosting tree regression algorithm to obtain a high-precision quantitative compatibility evaluation model. Finally, the model outputs a compatibility evaluation score to achieve automatic grading. This invention innovatively combines the intrinsic thermodynamic parameters of the material with changes in macroscopic performance, and establishes a mapping from features to a comprehensive score through machine learning. This overcomes the shortcomings of traditional methods, such as single evaluation dimensions and reliance on long-term aging experiments, achieving rapid, accurate, and mechanistic compatibility evaluation. Specifically, it includes the following steps: Step S1: Conduct a compatibility test on the BOPP film and the impregnating agent, and collect comprehensive performance data samples of the BOPP film and the impregnating agent before and after the experiment. The comprehensive performance of the BOPP film includes electrical, thermal, mechanical, and physicochemical properties, while the comprehensive performance of the impregnating agent includes its breakdown field strength and dielectric loss. The physicochemical properties of the BOPP film include its density and thickness. The electrical properties of the BOPP film include its breakdown field strength and dielectric loss; the thermal properties include its melting temperature and crystallinity; and the mechanical properties include its tensile strength and elongation at break.

[0026] In this embodiment, the compatibility experiment between the BOPP film and the impregnating agent was conducted under eight different experimental conditions: immersion at 80℃, 90℃, 100℃, and 110℃ for 24 hours and 48 hours respectively. The impregnating agents included four types: benzyltoluene, diarylethane (PXE), FR3 natural ester vegetable insulating oil, and phenylethylphenyl ethane (PEPE). In summary, in this embodiment, a total of 32 sets of data samples were used for training the Gradient Boosting Tree Regression (GBDT) prediction model.

[0027] Step S2: Calculate the depth impregnation efficiency index (DIEE) based on the density and thickness of the BOPP film before and after the experiment; calculate the affinity index between the BOPP film and the impregnating agent using the Hansen solubility parameter method. and relative energy density .

[0028] The Deep Impregnation Efficiency Index (DIEE), which is defined as the rate of change of density divided by the rate of change of thickness, is calculated as follows: ; in, ρ 0 indicates the density of the BOPP film before impregnation; ρit0 represents the density of the BOPP film after impregnation; t0 represents the thickness of the BOPP film before impregnation; ti represents the thickness of the BOPP film after impregnation. DIEE A value >1 indicates that the deep impregnation effect is good.

[0029] Affinity index of BOPP film to impregnating agent The calculation method is as follows: ; in, , , These are the dispersion force, polar force, and hydrogen bonding force components, respectively; subscript 1 represents BOPP film, and subscript 2 represents impregnating agent.

[0030] Relative energy density of BOPP film and impregnating agent The calculation method is as follows: ; Where R0 is the radius of the dissolution sphere of the BOPP film in the Hansen solubility parameter space. The affinity index between BOPP and each impregnating agent is calculated using the Hansen solubility parameter method. R a With relative energy density RED . RED< A value of 1 indicates a good thermodynamic compatibility basis.

[0031] Step S3: Combine the comprehensive performance data sample collected in step S1 with the deep impregnation efficiency index (DIEE) and affinity index calculated in step S2. and relative energy density Normalize the data and use it as a dataset.

[0032] Specifically, the normalization process uses the minimum-maximum normalization method for standardization, and the formula for calculating the standardized value is: ; in, X max and X min represents the maximum and minimum values ​​of the data sample on this indicator, x represents the corresponding indicator, and xnorm represents the normalized value of the corresponding indicator.

[0033] The collected chemical, thermal, mechanical, and physicochemical properties of each component will be analyzed. RED and DIEE Data samples, used as input feature parameters of the GBDT model, are standardized using the min-maximum normalization method to map them to the [0,1] interval. After standardization, they are uniformly forward-oriented, that is, the inverse index parameters are transformed into 1- XnormThis ensures that all indicators conform to the principle that larger values ​​indicate better compatibility.

[0034] Step S4: Construct a gradient boosting tree regression prediction model, train the model using the input features, and obtain a trained compatibility evaluation model. This specifically includes the following steps: Step S41: Randomly divide the normalized dataset into training, validation, and test sets in a 7:2:1 ratio; the total number of regression decision trees... M The learning rate is 100. η The gradient boosting tree regression prediction model is initialized with a value of 0.1050. Iterative training is then performed based on the training and validation sets, aiming to reduce the preset loss function, to construct a compatibility evaluation model. The least squares loss function is selected, as detailed below: ; in, L ( y i , F ( X i The loss function is used to measure the error between the predicted value and the true value of a single sample. y i For the first i The true compatibility evaluation value of each sample; F ( x i ) for the current compatibility assessment model for the first i Predicted compatibility assessment scores for each sample; x i Let be the input feature vector of the i-th sample.

[0035] Negative gradient iterative residual calculation is used, through m The addition model is constructed through rounds of iterations, and the formula is expressed as: ; in: r mi For the first m The residual of the i-th sample in the next iteration F m-1 ( x i ) is the first m-1 The compatibility evaluation model in the second iteration is for the first... i Predicted compatibility assessment scores for each sample.

[0036] Input features of the training set X The independent variable is the residual. r mi Generate a regression decision tree for the target value. fm ( X By calculating the optimal output value of the leaf nodes of the decision tree, the predicted value of the GBDT model is corrected, and a compatibility evaluation model based on GBDT is obtained.

[0037] The evaluation function of the compatibility assessment model is: ; in, y The compatibility assessment score is the output of the compatibility assessment model. F 0( X () represents the initial predicted value of the gradient boosting tree regression prediction model; η To increase the learning rate of the tree regression prediction model for gradient gradation; f m ( X ) is the first m A regression decision tree for input features X The predicted output value.

[0038] Step S42: Input the test set, compare the predicted compatibility evaluation score output by the compatibility evaluation model with the benchmark compatibility evaluation score derived from the test set, and calculate the accuracy evaluation index of the compatibility evaluation model based on the comparison result. The accuracy evaluation index includes at least the coefficient of determination R² and the mean absolute error (MAE) or mean absolute percentage error (MAPE) between the predicted and actual values.

[0039] Step S43: If all accuracy evaluation indicators reach or exceed the preset accuracy threshold, the compatibility evaluation model is determined to be trained and a trained compatibility evaluation model is obtained; if any key accuracy evaluation indicator fails to reach the preset accuracy threshold, the model training stage is returned, the model parameters are adjusted and retrained until the accuracy of the compatibility evaluation model meets the requirements.

[0040] Step S5: The compatibility between the BOPP film and the impregnating agent is evaluated using the trained compatibility assessment model to obtain a compatibility assessment score, and the compatibility level is determined based on the score. Compatibility is divided into different levels according to a preset scoring threshold. For excellence, For good, To pass, This is not up to standard.

[0041] In this embodiment, 10 independent immersion test data samples were collected and input into the trained compatibility evaluation model. The sample data are shown in Table 1. Table 1. Data from the Immersion Test Set The predicted values ​​output by the model are compared with the benchmark evaluation values ​​derived from the experimental data. The comparison results are as follows: Figure 2 As shown, the coefficient of determination R² of the calculated model is 0.9956, the mean absolute error (MAE) between the predicted and actual values ​​is 0.923, and the accuracy threshold is set to R. 2 A value >0.95 and a MAE <2.0 indicate that the compatibility evaluation model obtained in this invention has high prediction accuracy, and the error rate of the predicted value is as follows: Figure 3 As shown. By Figure 2 , Figure 3 It can be seen that the predicted values ​​output by the compatibility evaluation model based on the present invention have high accuracy, small difference from the actual values, and strong consistency, indicating that the compatibility evaluation method of the present invention is comprehensive, efficient and accurate.

[0042] The method of this invention can be applied in the evaluation of BOPP film impregnation systems. Based on the characteristics of capacitor composite insulation systems, this invention effectively evaluates the compatibility of BOPP film and impregnating agent, providing quantitative guidance for the selection and formulation design of capacitor insulation materials. It fills the gap in the existing technology in terms of multi-dimensional mechanism correlation and rapid quantitative prediction, which is of great significance for improving the long-term reliability and durability design of capacitor insulation systems. It makes the compatibility evaluation process more efficient and the evaluation results more accurate and reliable, which is more conducive to the construction of high-performance, long-life power capacitors.

[0043] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A method for evaluating the compatibility of BOPP film with impregnating agent, characterized in that, Includes the following steps: Step S1: Conduct a compatibility test on the BOPP film and the impregnating agent, and collect comprehensive performance data samples of the BOPP film and the impregnating agent before and after the test. The comprehensive properties of BOPP film include electrical, thermal, mechanical, and physicochemical properties, while the comprehensive properties of impregnating agent include the breakdown field strength and dielectric loss of the impregnating agent; the physicochemical properties of BOPP film include its density and thickness. Step S2: Calculate the Deep Impregnation Efficiency Index (DIEE) based on the density and thickness of the BOPP film before and after the experiment; calculate the affinity index between the BOPP film and the impregnating agent using the Hansen solubility parameter method. and relative energy density ; Step S3: Combine the comprehensive performance data sample collected in step S1 with the deep impregnation efficiency index (DIEE) and affinity index calculated in step S2. and relative energy density Normalize the data and use it as a dataset; Step S4: Construct a gradient boosting tree regression prediction model, train the model using input features, and obtain a trained compatibility evaluation model. Step S5: Use the trained compatibility assessment model to assess the compatibility between the BOPP film and the impregnating agent, obtain a compatibility assessment score, and determine the compatibility level based on the compatibility assessment score.

2. The method for evaluating the compatibility of BOPP film and impregnating agent according to claim 1, characterized in that, The electrical properties of the BOPP film in step S1 include breakdown field strength and dielectric loss; the thermal properties include melting temperature and crystallinity; and the mechanical properties include tensile strength and elongation at break.

3. The method for evaluating the compatibility of BOPP film and impregnating agent according to claim 1, characterized in that, The calculation method for the Deep Impregnation Efficiency Index (DIEE) in step S2 is as follows: ； in, ρ 0 indicates the density of the BOPP film before impregnation; ρi t0 represents the density of the BOPP film after impregnation; t0 represents the thickness of the BOPP film before impregnation; ti represents the thickness of the BOPP film after impregnation.

4. The method for evaluating the compatibility of BOPP film and impregnating agent according to claim 1, characterized in that, In step S2, the affinity index between the BOPP film and the impregnating agent The calculation method is as follows: ； in, , , These are the dispersion force, polar force, and hydrogen bonding force components, respectively; subscript 1 represents BOPP film, and subscript 2 represents impregnating agent.

5. The method for evaluating the compatibility of BOPP film and impregnating agent according to claim 1, characterized in that, The relative energy density of the BOPP film and the impregnating agent in step S2 The calculation method is as follows: ； Where R0 is the radius of the dissolution sphere of the BOPP film in the Hansen solubility parameter space.

6. The method for evaluating the compatibility of BOPP film and impregnating agent according to claim 1, characterized in that, In step S3, the normalization process specifically uses the minimum-maximum normalization method for standardization. The formula for calculating the standardized value is: ； in, X max and X min represents the maximum and minimum values ​​of the data sample on this indicator, x represents the corresponding indicator, and xnorm represents the normalized value of the corresponding indicator.

7. The method for evaluating the compatibility of BOPP film and impregnating agent according to claim 1, characterized in that, Step S4 specifically includes the following steps: The normalized dataset is divided into training, validation, and test sets; the gradient boosting tree regression prediction model is initialized; based on the training and validation sets, iterative training is performed with the goal of reducing the preset loss function to build a compatibility evaluation model; Input a test set, compare the predicted compatibility evaluation score output by the compatibility evaluation model with the benchmark compatibility evaluation score obtained based on the test set, and calculate the accuracy evaluation index of the compatibility evaluation model based on the comparison result. If all accuracy evaluation indicators reach or exceed the preset accuracy threshold, the compatibility evaluation model is considered to have completed training, and a trained compatibility evaluation model is obtained. If any key accuracy evaluation indicator fails to reach the preset accuracy threshold, the model training phase is returned, the model parameters are adjusted, and the model is retrained until the accuracy of the compatibility evaluation model meets the requirements.

8. The method for evaluating the compatibility of BOPP film and impregnating agent according to claim 7, characterized in that, The accuracy evaluation indicators include at least the coefficient of determination R² and the mean absolute error (MAE) or mean absolute percentage error (MAPE) between the predicted and actual values.

9. The method for evaluating the compatibility of BOPP film and impregnating agent according to claim 7, characterized in that, The loss function used is the least squares loss function, as follows: ； in, L ( y i , F ( X i The loss function is used to measure the error between the predicted value and the true value of a single sample. y i For the first i The true compatibility evaluation value of each sample; F ( x i ) for the current compatibility assessment model for the first i Predicted compatibility assessment scores for each sample; x i Let be the input feature vector of the i-th sample.

10. The method for evaluating the compatibility of BOPP film and impregnating agent according to claim 1, characterized in that, The evaluation function of the compatibility assessment model is: ； in, y The compatibility assessment score is the output of the compatibility assessment model. F 0( X () represents the initial predicted value of the gradient boosting tree regression prediction model; η To increase the learning rate of the tree regression prediction model for gradient gradation; f m ( X ) is the first m A regression decision tree for input features X The predicted output value.

Images