Gas phase drying parameter optimization method, device and system for kerosene

Abstract

The application relates to a kerosene gas phase drying parameter optimization method, device and system. Through establishment of a correlation model of a process parameter-adsorption coefficient, the process parameters of kerosene gas phase drying are changed from empirical judgment to scientific quantification, and the internal quantitative law between the process parameters and adsorption results is directly revealed. After a target kerosene residual amount is determined, a target adsorption coefficient is determined, and the established correlation model is used for reverse optimization to solve an optimal process parameter combination meeting the target. Since the model itself is fitted based on test data of multiple groups of parameter value range coverage, the reverse solving process can accurately locate the process parameters meeting the conditions in the global parameter space, and deviation of the parameters caused by artificial trial and error or conservative setting is avoided. The process parameter setting is changed from passive trial and error to active quantitative calculation, and finally, the optimal process parameter combination directly corresponds to the target residual amount requirement, so that the problem of inaccurate parameter setting is completely solved.

Description

Technical Field

[0001] This application relates to the field of transformer technology, and in particular to a method, apparatus and system for optimizing kerosene vapor phase drying parameters. Background Technology

[0002] Kerosene vapor phase drying is one of the core drying technologies in the manufacturing of large oil-immersed transformers. It uses kerosene vapor as a heating medium and solvent to heat the transformer body in a vacuum environment, causing the moisture inside the insulation material to evaporate, diffuse, and be extracted, thereby achieving the purpose of deep drying.

[0003] After the drying process is completed, some kerosene will remain in the microporous structure of the transformer's insulating components (such as crepe paper, cardboard, support bars, etc.). Under long-term operation in a high electric field and high temperature environment, this residual kerosene may gradually precipitate, decompose, or interact with the insulating materials, seriously threatening the long-term operational safety and lifespan of the transformer.

[0004] At present, the setting of process parameters for kerosene vapor phase drying mostly relies on historical experience or rough industry standards, lacking systematic experimental research. This leads to blind process adjustments and exacerbates the phenomenon of kerosene residue on insulating parts. Summary of the Invention

[0005] Therefore, it is necessary to address the problem of inaccurate setting of process parameters for kerosene vapor phase drying by proposing a method, apparatus, and system for optimizing kerosene vapor phase drying parameters.

[0006] This application provides a method for optimizing kerosene vapor phase drying parameters, comprising: determining basic parameters of a target insulating component to be tested; wherein the basic parameters include at least one of sample type, sample weight, and sample specific surface area; determining multiple process parameters affecting the kerosene adsorption capacity of the target insulating component to be tested, and the value range of each process parameter; determining multiple sets of process parameter combinations based on the process parameters and the value range, and performing kerosene vapor phase drying treatment on the target insulating component to be tested under each set of process parameter combinations; obtaining the kerosene residue of the target insulating component to be tested after kerosene vapor phase drying treatment, and determining the target insulating component to be tested based on the sample weight and the kerosene residue. The insulation adsorption coefficient of the insulating component is measured; using the combination of the basic parameters and the process parameters as input variables and the insulation adsorption coefficient as the output variable, a multiple regression analysis is performed to determine the correlation model between the process parameters and the adsorption coefficient of the target insulating component; wherein, the correlation model characterizes the quantitative influence of the basic parameters and the process parameters on the insulation adsorption coefficient; the target kerosene residue is obtained, and the target adsorption coefficient is determined based on the target kerosene residue; the optimal process parameter combination is determined by inverse optimization based on the target adsorption coefficient and the correlation model; wherein, the optimal process parameter combination is used for kerosene vapor-phase drying of insulating components of the same type as the target insulating component.

[0007] In some embodiments, the method further includes: performing kerosene vapor-phase drying treatment on the verification sample under a verification process parameter combination; wherein the verification sample is of the same sample type as the target insulation component to be tested, and the verification process parameter combination is different from the process parameter combination; obtaining the kerosene residue of the verification sample and determining the measured insulation adsorption coefficient; determining the predicted insulation adsorption coefficient of the verification sample according to the correlation model; and performing model verification on the correlation model according to the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient.

[0008] In some embodiments, the step of validating the correlation model based on the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient includes: determining the deviation between the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient; if the deviation is greater than a preset deviation threshold, re-determining multiple sets of process parameter combinations based on the process parameters and the value range, and returning to the step of performing kerosene vapor phase drying treatment on the target insulation component under each set of process parameter combinations; if the deviation is less than or equal to the preset deviation threshold, determining that the correlation model has passed the validation.

[0009] In some embodiments, the correlation model includes an independent influence term of the combination of the basic parameters and the process parameters on the insulation adsorption coefficient, and an interactive influence term of the combination of the basic parameters and the process parameters on the insulation adsorption coefficient.

[0010] In some embodiments, the method further includes: comparing the optimal process parameter combination with a preset drying process library; taking an existing process parameter combination in the preset drying process library that matches the optimal process parameter combination as the actual process parameter combination; and performing kerosene vapor phase drying on an insulating component of the same type as the target insulating component to be tested according to the actual process parameter combination.

[0011] In some embodiments, the step of using an existing process parameter combination that matches the optimal process parameter combination in the preset drying process library as the actual process parameter combination includes: determining the core influencing parameter that has the greatest impact on the insulation adsorption coefficient among the process parameters according to the correlation model; and selecting an existing process parameter combination from the preset drying process library that has the smallest difference from the optimal process parameter combination in terms of the core influencing parameter, and the difference is less than or equal to a preset difference threshold, as the actual process parameter combination.

[0012] In some embodiments, after comparing the optimal process parameter combination with a preset drying process library, the method further includes: updating the preset drying process library according to the optimal process parameter combination when there is no existing process parameter combination in the preset drying process library that matches the optimal process parameter combination.

[0013] In some embodiments, the method further includes: constructing a process scheme library; wherein the process scheme library is constructed hierarchically based on two dimensions: transformer voltage level and total weight range of insulation materials; from the process scheme library, calling the initial process parameter combination corresponding to the current transformer voltage level and the current total weight range of insulation materials, and performing kerosene vapor phase drying treatment on the insulation components; obtaining the kerosene residue amount of the insulation components after the kerosene vapor phase drying treatment; if the kerosene residue amount exceeds a preset safety threshold, using the process parameter combination corresponding to the previous level transformer voltage level and / or the previous level total weight range of insulation materials as the initial process parameter combination for the current transformer voltage level and the current total weight range of insulation materials.

[0014] This application provides a kerosene vapor phase drying parameter optimization device, comprising: a parameter acquisition module for determining basic parameters of a target insulating component to be tested; wherein the basic parameters include at least one of sample type, sample weight, and sample specific surface area; a process determination module for determining multiple process parameters affecting the kerosene adsorption capacity of the target insulating component to be tested, and the value range of each process parameter; a drying treatment module for determining multiple sets of process parameter combinations based on the process parameters and the value range, and performing kerosene vapor phase drying treatment on the target insulating component to be tested under each set of process parameter combinations; and a residue detection module for acquiring the amount of kerosene residue on the target insulating component to be tested after kerosene vapor phase drying treatment, and determining the residual amount of kerosene based on the sample weight and the amount of kerosene residue. The system comprises: an insulation adsorption coefficient for the target insulating component; a model building module for performing multiple regression analysis using the combination of basic parameters and process parameters as input variables and the insulation adsorption coefficient as output variable to determine the correlation model between process parameters and adsorption coefficient of the target insulating component; wherein the correlation model characterizes the quantitative influence of the basic parameters and process parameters on the insulation adsorption coefficient; a target analysis module for obtaining the target kerosene residue and determining the target adsorption coefficient based on the target kerosene residue; and a parameter optimization module for performing reverse optimization based on the target adsorption coefficient and the correlation model to determine the optimal combination of process parameters; wherein the optimal combination of process parameters is used for kerosene vapor-phase drying of insulating components of the same type as the target insulating component.

[0015] This application provides a kerosene vapor phase drying parameter optimization system, comprising: a sample parameter acquisition device, a kerosene vapor phase drying treatment device, an adsorption measurement device, and a control device. The sample parameter acquisition device is used to determine the basic parameters of a target insulating component under test; the kerosene vapor phase drying treatment device is used to perform kerosene vapor phase drying treatment on the target insulating component under test; the adsorption measurement device is used to measure the kerosene residue in the target insulating component under test; the control device is connected to the sample parameter acquisition device, the kerosene vapor phase drying treatment device, and the adsorption measurement device, and is used to execute the steps of the above-described kerosene vapor phase drying parameter optimization method.

[0016] The aforementioned kerosene vapor-phase drying parameter optimization method, apparatus, and system systematically selects the basic parameters of the target insulating component (i.e., at least one of sample type, sample weight, and sample specific surface area), and combines multiple sets of process parameter combinations to conduct kerosene vapor-phase drying experiments, determine the kerosene residue, and calculate the insulation adsorption coefficient. Then, using the basic parameters and process parameter combinations as input variables and the insulation adsorption coefficient as the output variable, a multiple regression analysis is performed to establish a correlation model between the process parameters and the adsorption coefficient of the target insulating component. Based on this, the target adsorption coefficient is determined according to the target kerosene residue, and the established correlation model is used for inverse optimization to directly solve for the optimal combination of process parameters that meets the residue requirements.

[0017] This scheme, through the establishment of a correlation model between process parameters and adsorption coefficients, transforms the process parameters of kerosene gas-phase drying from empirical judgment to scientific quantification. It directly reveals the inherent quantitative law between process parameters and adsorption results, replacing the traditional approach that relies on fuzzy experience or rough specifications, thus ensuring the scientific basis for parameter setting in principle. After determining the target adsorption coefficient based on the target kerosene residue, the established correlation model is used for inverse optimization to solve for the optimal combination of process parameters that meets the target. Since the model itself is based on fitting multiple sets of experimental data covering the parameter value range, the inverse solution process can accurately locate the process parameters that meet the conditions within the global parameter space, avoiding parameter deviations caused by human trial and error or conservative settings. In this way, process parameter setting is transformed from passive trial and error to active quantitative calculation. The final output of the optimal combination of process parameters directly corresponds to the target residue requirement, thereby completely solving the problem of inaccurate parameter setting and achieving controllable and predictable drying effects. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the conventional technology, the drawings used in the description of the embodiments or the conventional technology will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the kerosene vapor phase drying parameter optimization method in one embodiment of this application;

[0020] Figure 2 This is a schematic diagram of the association model verification process in one embodiment of this application;

[0021] Figure 3 This is a schematic diagram of the kerosene vapor phase drying parameter optimization method in another embodiment of this application;

[0022] Figure 4This is a schematic diagram of the kerosene vapor phase drying parameter optimization method in another embodiment of this application;

[0023] Figure 5 This is a schematic diagram of the fault-tolerant optimization process in one embodiment of this application;

[0024] Figure 6 This is a schematic diagram of the kerosene vapor phase drying parameter optimization device in one embodiment of this application;

[0025] Figure 7 This is a schematic diagram of the kerosene vapor phase drying parameter optimization device in another embodiment of this application;

[0026] Figure 8 This is a schematic diagram of the kerosene vapor phase drying parameter optimization system in one embodiment of this application;

[0027] Figure 9 This is a schematic diagram of a process scheme library in one embodiment of this application. Detailed Implementation

[0028] To facilitate understanding of this application, a more complete description will be provided below with reference to the accompanying drawings. Preferred embodiments of this application are shown in the drawings. However, this application can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a thorough and complete understanding of the disclosure of this application.

[0029] Please see Figure 1 This application provides a method for optimizing kerosene vapor phase drying parameters, including steps 101, 102, 103, 104, 105, 106 and 107.

[0030] Step 101: Determine the basic parameters of the target insulating component to be tested.

[0031] Step 102: Determine multiple process parameters that affect the kerosene adsorption capacity of the target insulating component under test, as well as the value range of each process parameter.

[0032] Step 103: Based on the process parameters and their ranges, determine multiple combinations of process parameters, and perform kerosene vapor phase drying on the target insulation component under each combination of process parameters.

[0033] Step 104: Obtain the kerosene residue on the target insulating component after kerosene vapor phase drying treatment, and determine the insulation adsorption coefficient of the target insulating component based on the sample weight and kerosene residue.

[0034] Step 105: Using the combination of basic parameters and process parameters as input variables and the insulation adsorption coefficient as output variable, perform multiple regression analysis to determine the correlation model between process parameters and adsorption coefficient of the target insulation component to be tested.

[0035] Step 106: Obtain the target kerosene residue and determine the target adsorption coefficient based on the target kerosene residue.

[0036] Step 107: Perform reverse optimization based on the target adsorption coefficient and the correlation model to determine the optimal combination of process parameters.

[0037] The optimal combination of process parameters is used for kerosene vapor-phase drying of insulation components of the same type as the target insulation component. A correlation model characterizes the quantitative influence of basic parameters and process parameters on the insulation adsorption coefficient, and is used to guide the adjustment of process parameters for kerosene vapor-phase drying. Basic parameters include at least one of sample type, sample weight, and sample specific surface area.

[0038] The target insulation component to be tested refers to the insulation component of the transformer currently used for optimizing kerosene vapor phase drying parameters. It is understood that the type of insulation component is not unique; any component made from insulating material in the transformer can be considered an insulation component. For example, in one embodiment, the insulation component includes one or more of the following: winding turn insulation, lead insulation, phase spacers, shielding, pressure plates, support plates, end rings, support bars, and pads in the transformer body.

[0039] Basic parameters refer to the inherent properties of the target insulating component under test. These may include sample type (such as cellulose, epoxy composites, aromatic polyamide paper, etc.), sample weight (in kg), and sample specific surface area (in m² / g), used to characterize the basic properties of the insulating component. The method for obtaining sample weight is not unique; it can be obtained directly by weighing using an electronic balance, and is not limited here. Sample specific surface area refers to the surface area per unit mass of insulating component, usually measured using the Biosorption Spectrometry (BET) method, with units of square meters per gram (m² / g), reflecting the microstructural characteristics of the insulating component.

[0040] Process parameters refer to the controllable operational variables during kerosene vapor phase drying. It's important to understand that their types are not unique; process parameters will vary depending on the specific kerosene vapor phase drying stage. The kerosene vapor phase drying process typically includes the following stages: preparation, heating, depressurization, high vacuum, and gas filling. The heating stage further includes an intermediate depressurization stage and a distillation stage.

[0041] For example, the process parameters belonging to the heating stage and their value ranges include: the temperature control time for each step, the number of intermediate pressure reductions, and the total heating time; wherein, the total heating time is positively correlated with the temperature control time for each step and the number of intermediate pressure reductions, and the value range is 20 hours to 100 hours. The process parameters belonging to the high vacuum stage and their value ranges include: the vacuum level at the end of the high vacuum stage and the high vacuum stage time.

[0042] The value range refers to the numerical range that can be set for each process parameter, such as a total heating time of 20 hours to 100 hours, or a vacuum degree of 10 Pa to 100 Pa. A process parameter combination refers to a set of test conditions consisting of multiple process parameters with specific values, such as a combination of a heating time of 60 hours and a vacuum degree of 55 Pa.

[0043] Kerosene vapor drying is a drying process for transformer insulation components. It includes stages such as preparation, heating (including intermediate voltage reduction and distillation), high vacuum, and gas filling. Kerosene vapor is used as the heating medium to remove moisture from the insulation material under vacuum conditions, although some kerosene may remain. Kerosene residue refers to the mass of kerosene remaining in the microporous structure of the insulation component after drying, measured in grams (g). It is a key indicator for evaluating drying effectiveness and operational safety.

[0044] The insulation adsorption coefficient is defined as the amount of kerosene residue per unit mass of insulating component. The calculation formula is k=ΔW / m, and the unit is g / kg. It is used to quantitatively characterize the adsorption capacity of insulating materials under specific processes. Wherein, k represents the insulation adsorption coefficient, ΔW represents the amount of kerosene residue, and m represents the sample weight.

[0045] It should be noted that, in actual scenarios, the type of process parameters selected is not unique. It can be any process parameter in kerosene vapor-phase drying that affects kerosene residue, such as time, temperature, vacuum level, etc., at various stages, and is not limited here. For example, in one embodiment, the process parameters include the total heating time and the vacuum level during the high-vacuum stage.

[0046] When determining the combination of process parameters, more than two sets of process parameter combinations can be set according to actual needs, and there is no specific limitation. For example, in one embodiment, taking the above-mentioned process parameters including the total heating time and the vacuum degree of the high vacuum stage as an example, the total heating time and the vacuum degree of the high vacuum stage each include three values, and correspondingly, through orthogonal arrangement, nine sets of process parameter combinations can be formed.

[0047] The approach in this embodiment first determines the basic parameters of the target insulating component (sample type, weight, specific surface area), as these parameters directly affect the adsorption and desorption behavior of kerosene. Then, it determines the key process parameters affecting the kerosene adsorption capacity and their reasonable value ranges to avoid blindly setting parameters.

[0048] Multiple combinations of process parameters were generated based on experimental design methods (such as orthogonal experiments), and kerosene vapor-phase drying treatment was performed for each combination. The residual amount of kerosene under each combination was obtained by field measurement, and the insulation adsorption coefficient was calculated by combining the sample weight. Finally, a multiple regression analysis was performed with the basic parameters and process parameter combinations as input variables and the insulation adsorption coefficient as output variable to construct a correlation model between process parameters and adsorption coefficient.

[0049] The correlation model between process parameters and adsorption coefficients shows the correspondence between various basic parameters and process parameters and insulation adsorption coefficients. Based on the coefficients of each item in the correlation model, the quantitative influence of each parameter on the insulation adsorption coefficient can be determined.

[0050] For example, in one embodiment, the correlation model is: k = 13.0 + 3.2 × S - 0.020 × T + 0.040 × P, where k represents the insulation adsorption coefficient, S represents the sample specific surface area, T represents the total heating time, and P represents the vacuum degree during the high vacuum stage. It can be seen that the insulation adsorption coefficient k is positively correlated with the sample specific surface area S, negatively correlated with the total heating time T, and positively correlated with the vacuum degree P during the high vacuum stage. Since the coefficient of P is greater than the coefficient of T, the vacuum degree during the high vacuum stage has a greater impact on the insulation adsorption coefficient than the total heating time.

[0051] The target kerosene residue refers to the maximum permissible kerosene residue threshold set according to the transformer's operational safety requirements. For example, in one embodiment, it can be configured as 0.03%. The target adsorption coefficient refers to the target insulation adsorption coefficient value calculated from the target kerosene residue and the total weight of the insulation components, i.e., target adsorption coefficient = target kerosene residue / total weight of insulation components. This value serves as a constraint for reverse optimization. Reverse optimization is the process of using a pre-constructed correlation model to find the optimal combination of process parameters that meets the requirements, given the target adsorption coefficient (determined by the target kerosene residue). The optimal combination of process parameters (such as heating time, vacuum degree, and number of voltage reductions) is a set of process parameter values ​​(e.g., heating time, vacuum degree, number of voltage reductions) obtained through reverse optimization calculations, used to guide actual production.

[0052] This embodiment utilizes an established correlation model to achieve forward-looking design from target residual amount to process parameters. Specifically, firstly, the target kerosene residual amount is obtained based on product safety requirements (e.g., a 500kV transformer requires a residual amount ≤0.03%), and then the corresponding target adsorption coefficient is calculated based on the total weight of the insulating components. Next, the target adsorption coefficient is substituted into the correlation model, and using the model equation as a constraint, an optimization algorithm (such as numerical optimization or genetic algorithm) is employed for inverse optimization to solve for a set of optimal process parameter combinations.

[0053] This completely changes the traditional approach of relying on repeated trial and error to adjust the process, enabling engineers to predict and set parameters before actual drying, accurately control the amount of kerosene residue from the source, significantly shorten the process development cycle and reduce the risk of exceeding standards.

[0054] The aforementioned method for optimizing kerosene vapor-phase drying parameters systematically selects the basic parameters of the target insulating component and conducts kerosene vapor-phase drying experiments in combination with multiple sets of process parameters. It then measures the kerosene residue and calculates the insulation adsorption coefficient. Finally, using the basic parameters and process parameter combinations as input variables and the insulation adsorption coefficient as the output variable, a multiple regression analysis is performed to establish a correlation model between the process parameters and the adsorption coefficient of the target insulating component. This approach, by establishing the correlation model between process parameters and the adsorption coefficient, transforms the process parameters of kerosene vapor-phase drying from empirical judgment to scientific quantification, avoiding unclear mechanisms of process parameter influence. It provides a reliable predictive and guiding tool for precise control of kerosene residue, achieving accurate control of kerosene residue and effectively reducing the kerosene residue in insulating components.

[0055] In some embodiments, obtaining the kerosene residue in the target insulating component after kerosene vapor phase drying includes: determining the kerosene residue in the target insulating component after kerosene vapor phase drying by gravimetric analysis when the sample weight is less than or equal to a preset weight threshold; and determining the kerosene residue in the target insulating component after kerosene vapor phase drying by chromatographic analysis when the sample weight is greater than the preset weight threshold.

[0056] The preset weight threshold refers to the boundary value for dividing the measurement method based on the sample weight. The specific value can be set according to the measurement accuracy requirements. For example, in one embodiment, it can be 1 kg, and in another embodiment, it can be other values. The gravimetric method, also known as the direct gravimetric method, uses a high-precision analytical balance (e.g., accuracy 0.001g) to directly weigh the mass difference of the insulating parts before and after drying to obtain the kerosene residue. It is suitable for samples with small weights.

[0057] Chromatographic analysis is an indirect determination method based on the principle of dissolution equilibrium. The target insulating component is immersed in a suitable solvent until the kerosene dissolves to equilibrium. Then, the kerosene concentration is analyzed by gas chromatography, and the residual amount of kerosene is calculated based on the kerosene concentration. It is suitable for samples with large weights and avoids the problem of insufficient accuracy when directly weighing large samples.

[0058] This scheme adaptively selects the measurement method based on whether the sample weight exceeds a preset weight threshold when obtaining kerosene residue, ensuring measurement accuracy. For smaller samples, direct gravimetric measurement is used; for larger samples, indirect measurement is performed using chromatographic analysis based on the principle of dissolution equilibrium. This is because small samples (such as 0.25 kg of crepe paper) can obtain high-precision results using an analytical balance, which is simple to operate; while for large samples (such as insulation components weighing tens of kilograms), direct weighing has a large error. By dissolving and extracting the kerosene and then using chromatographic analysis, the kerosene can be separated from the complex insulation structure and accurately quantified. This adaptive strategy ensures that regardless of sample size, the measured kerosene residue meets the high-precision requirements for subsequent modeling and analysis, laying the foundation for accurate calculation of the insulation adsorption coefficient and the reliability of the correlation model.

[0059] The above scheme, when obtaining the kerosene residue after kerosene vapor-phase drying, adaptively selects either gravimetric determination or chromatographic analysis based on whether the sample weight is less than or equal to a preset weight threshold. This ensures the ease of measurement for small samples and overcomes the lack of accuracy in direct weighing of large samples through chromatographic analysis, thereby ensuring the accuracy of the measured kerosene residue and providing a reliable data foundation for subsequent calculation of the insulation adsorption coefficient and construction of the correlation model.

[0060] Please see Figure 2 In some embodiments, the kerosene vapor phase drying parameter optimization method further includes steps 201, 202, 203 and 204.

[0061] Step 201: Perform kerosene vapor phase drying on the verification sample under the verification process parameter combination.

[0062] Step 202: Obtain the kerosene residue of the verification sample and determine the measured insulation adsorption coefficient.

[0063] Step 203: Determine the predicted insulation adsorption coefficient of the verification sample based on the correlation model.

[0064] Step 204: Validate the correlation model based on the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient.

[0065] Among them, the verification sample is the same type as the target insulation component to be tested, but the combination of verification process parameters is different from the combination of process parameters.

[0066] Validation samples refer to independent insulation samples of the same type as the target insulation sample but not involved in modeling, used to verify the predictive accuracy of the correlation model. Validation process parameter combinations refer to a new set of process conditions different from the process parameter combinations used in modeling. Measured insulation adsorption coefficient refers to the insulation adsorption coefficient calculated by actually measuring the kerosene residue and dividing it by the sample weight after drying the validation sample under the validation process parameter combinations. Predicted insulation adsorption coefficient refers to the insulation adsorption coefficient calculated by substituting the basic parameters of the validation sample and the validation process parameter combinations into the constructed correlation model.

[0067] To address the issue of unconfirmed reliability of correlation model predictions, a separate model validation step was introduced. Specifically, a validation sample of the same type as the target insulation component but not involved in the modeling was selected, and kerosene vapor-phase drying was performed using a different combination of validation process parameters than the modeling experiment. The residual kerosene content was measured after treatment, and the measured insulation adsorption coefficient was calculated. Simultaneously, the basic parameters and validation process parameter combination of this validation sample were substituted into the constructed correlation model to calculate the predicted insulation adsorption coefficient.

[0068] The above scheme uses a combination of verification process parameters that are the same type as the target insulation component sample but different from the process parameters. The verification sample is subjected to kerosene vapor phase drying treatment to obtain the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient predicted by the correlation model. The results are compared and verified to solve the problem that the reliability of the constructed correlation model prediction cannot be confirmed. This ensures that the model has good generalization ability and provides a reliable basis for subsequent adjustment of process parameters based on the model to control the amount of kerosene residue.

[0069] In some embodiments, the correlation model is validated based on the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient, including: determining the deviation between the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient; if the deviation is greater than a preset deviation threshold, re-determining multiple sets of process parameter combinations based on process parameters and value ranges, and returning to the step of performing kerosene vapor phase drying treatment on the target insulation component under each set of process parameter combinations; if the deviation is less than or equal to the preset deviation threshold, the correlation model is determined to have passed validation.

[0070] The deviation value refers to the absolute value or relative error of the difference between the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient, used to quantify the model's prediction accuracy. The preset deviation threshold refers to the maximum allowable deviation range for model validation. For example, in one embodiment, the deviation value is the ratio of the absolute value of the difference between the measured and predicted insulation adsorption coefficients to the measured insulation adsorption coefficient; correspondingly, the preset deviation threshold can be set to 5%.

[0071] Re-determining multiple combinations of process parameters refers to returning to the experimental design stage when model validation fails, supplementing with new combinations of process parameters (such as increasing the number of orthogonal experimental points or adjusting parameter levels), and performing additional drying experiments to enrich the training data.

[0072] This scheme establishes an iterative self-correction mechanism for the correlation model to ensure that the model always meets the preset accuracy requirements. In the model validation step, the deviation (e.g., relative error) between the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient is calculated and compared with a preset deviation threshold. If the deviation exceeds the threshold, it indicates that the current model fails to fully cover the process parameter space or is underfitting. In this case, the system automatically returns to determine multiple combinations of process parameters, supplements new orthogonal experimental data, and re-performs regression analysis to optimize the correlation model until the deviation meets the requirements.

[0073] The above scheme calculates the deviation between the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient during the model verification process. If the deviation value is greater than the preset deviation threshold, it automatically returns to redetermine multiple sets of process parameter combinations and supplements the experiment until the deviation value meets the requirements. This realizes the iterative self-correction of the correlation model, solves the model failure problem caused by insufficient sample coverage or changes in process conditions, and ensures that the model always meets the preset accuracy requirements, thereby ensuring the continuous effectiveness of kerosene residue control.

[0074] In some embodiments, the correlation model includes independent influence terms of the combination of basic parameters and process parameters on the insulation adsorption coefficient, as well as interactive influence terms of the combination of basic parameters and process parameters on the insulation adsorption coefficient.

[0075] Independent influence terms refer to the effects of a single basic parameter or a single process parameter on the insulation adsorption coefficient in the correlation model; for example, -0.020·T represents the independent negative correlation between heating time T and the adsorption coefficient. Interactive influence terms, on the other hand, reflect the additional effects of two or more parameters acting together on the insulation adsorption coefficient in the correlation model, i.e., the synergistic or antagonistic effects between parameters, such as the product of heating time and vacuum degree.

[0076] The correlation model constructed in this scheme breaks through the limitations of traditional single-factor linear analysis. It not only includes the independent influence terms of each basic parameter and process parameter on the insulation adsorption coefficient, but also introduces the interaction influence terms between parameters. In the actual kerosene vapor phase drying process, there are often complex coupling relationships between parameters. For example, extending the heating time and increasing the vacuum degree may have a synergistic effect, jointly promoting kerosene desorption; and there may also be an interaction between the number of intermediate pressure reductions and the total heating time.

[0077] By incorporating interaction terms into the model through regression analysis, the physicochemical mechanisms of the drying process can be more accurately reflected, significantly improving the model's prediction accuracy for kerosene residue. This design allows process engineers to use the model to evaluate the overall effect of parameter combination adjustments, rather than adjusting a single parameter in isolation, thereby achieving more scientific and precise residue control.

[0078] The aforementioned scheme constructs a correlation model that includes not only the independent influence terms of the combination of basic parameters and process parameters on the insulation adsorption coefficient, but also the interaction terms between the two. This breaks through the limitation of considering only the linear influence of a single parameter, and can comprehensively reflect the synergistic or antagonistic effects between multiple parameters. This makes the model more consistent with the complex mechanism of the actual drying process, thereby significantly improving the prediction accuracy of kerosene residue and the scientific nature of process control.

[0079] Please see Figure 3 In some embodiments, the kerosene vapor phase drying parameter optimization method further includes steps 302, 304 and 306.

[0080] Step 302: Compare the optimal combination of process parameters with the preset drying process library.

[0081] Step 304: Take the existing process parameter combination that matches the optimal process parameter combination from the preset drying process library as the actual process parameter combination.

[0082] Step 306: Based on the actual combination of process parameters, perform kerosene vapor phase drying on the insulating component of the same type as the target insulating component to be tested.

[0083] The preset drying process library, also known as the equipment standard process scheme library, stores multiple sets of verified and directly executable existing process parameter combinations. Adaptation refers to comparing the optimal process parameter combination obtained through reverse optimization with existing schemes in the preset drying process library, selecting the existing scheme whose key parameters (especially core parameters that significantly affect the adsorption coefficient) are closest in value. The actual process parameter combination refers to the process parameter combination adapted from the preset drying process library that can be directly used in production equipment. It may not be exactly the theoretical optimal combination, but it ensures engineering feasibility.

[0084] After obtaining the optimal combination of process parameters, it is compared item by item with a preset drying process library. Prioritizing the selection of an existing solution from the library that most closely matches the optimal combination in terms of the core influencing parameters ranked highest, this solution is chosen as the recommended execution solution after adaptation. This is because the equipment typically only executes preset fixed process curves and cannot directly run arbitrary continuous parameter values. Through adaptation, the system can select a feasible solution natively supported by the equipment as the actual process parameter combination, while ensuring that key parameters (such as total heating time and vacuum level during the high-vacuum stage) are as close as possible to the theoretical optimal values. This actual combination is used for batch drying of insulating components of the same type as the target insulating component, ensuring both the theoretical optimality of residual amount control and the operability and equipment compatibility of production.

[0085] The above scheme, by comparing the optimal process parameter combination obtained through reverse optimization with a preset drying process library, selects an existing process parameter combination that matches the optimal process parameter combination from the library as the actual process parameter combination for drying the same type of insulating parts. This solves the problem of the difficulty in directly connecting the theoretical optimization results with the preset scheme of actual production equipment, ensures the engineering feasibility of the optimization scheme, and improves the implementation efficiency of kerosene residue control.

[0086] In some embodiments, step 304 includes: determining the core influencing parameter that has the greatest impact on the insulation adsorption coefficient among the process parameters according to the correlation model; selecting from the preset drying process library the existing process parameter combination that has the smallest difference from the optimal process parameter combination in terms of the core influencing parameter and the difference is less than or equal to the preset difference threshold, as the actual process parameter combination.

[0087] The core influencing parameter refers to the process parameter that has the greatest impact on the insulation adsorption coefficient during kerosene vapor phase drying, as determined by the correlation model. The preset difference threshold is the upper limit of the allowable deviation for the core influencing parameter, used to determine whether the existing combination of process parameters can be used as an adaptation scheme.

[0088] An existing solution can be considered a candidate only if the absolute value of its value for the core influencing parameter deviates from that parameter in the optimal process parameter combination by no more than a preset difference threshold. If there are multiple candidates, the one with the smallest difference will be selected as the actual process parameter combination. If all deviations exceed the preset difference threshold, it is considered that there is no suitable solution, and the process of adding a new solution to the database should be triggered.

[0089] For example, in one embodiment, it is assumed that the correlation model indicates that the vacuum level (high vacuum stage) is the core influencing parameter with the greatest impact on the insulation adsorption coefficient. In the optimal process parameter combination, the vacuum level is 8 Pa, and the preset difference threshold is ±3 Pa. The preset drying process library includes scheme A with a vacuum level of 10 Pa (deviation 2 Pa), scheme B with a vacuum level of 9 Pa (deviation 1 Pa), and scheme C with a vacuum level of 5 Pa (deviation 4 Pa). All three schemes simultaneously satisfy the condition of being less than or equal to the preset difference threshold, but scheme B has the smallest deviation; therefore, scheme B is used as the actual process parameter combination.

[0090] This scheme uses a correlation model to quantitatively identify the core influencing parameter that has the greatest impact on the insulation adsorption coefficient. Using this as the core matching benchmark, it selects from a pre-set drying process library the existing process parameter combination that has the smallest deviation from the optimal process parameter combination on this core influencing parameter and meets a pre-set difference threshold as the actual process parameter combination. This solves the problem that directly selecting the scheme with the closest overall parameters may lead to a large deviation in the core parameter, thus significantly affecting the control effect of kerosene residue.

[0091] In this way, the actual process scheme can retain the optimization effect of reverse optimization to the greatest extent, significantly improve the accuracy of kerosene residue control and the rationality of engineering adaptation, and avoid the risk of exceeding the residue limit due to sacrificing core influencing parameters for matching secondary parameters.

[0092] Please see Figure 4 In some embodiments, after step 302, the method further includes step 402.

[0093] Step 402: If there is no existing process parameter combination in the preset drying process library that matches the optimal process parameter combination, update the preset drying process library according to the optimal process parameter combination.

[0094] When the deviation between the theoretically optimal combination of process parameters and all existing solutions in the library exceeds the preset proximity requirement (e.g., the difference in core parameters exceeds the allowable range), it is determined that there is no suitable solution. At this time, the theoretically optimal combination of process parameters will be standardized (e.g., rounding, segmentation and merging) and added to the preset drying process library as a new standard process solution, realizing the dynamic expansion of the process library.

[0095] This scheme endows the preset drying process library with the ability to learn and dynamically expand. After comparing the optimal combination of process parameters with the preset drying process library, if there is no existing combination of process parameters in the library that meets the proximity requirement (i.e., the difference between all existing schemes and the optimal combination exceeds the acceptable range), the process of adding a new process scheme is triggered.

[0096] For example, in one embodiment, when reverse optimization yields a combination of 95 hours of heating time and 8 Pa of vacuum, but the library only contains two other options, 90 hours / 10 Pa and 100 hours / 5 Pa, neither of which meets the proximity requirement, the system can standardize the new combination and add it to the library. This mechanism solves the problem of the inability to systematically accumulate and reuse new optimized solutions, allowing the process library to be continuously enriched with the accumulation of production experience. Subsequent control of kerosene residue in similar insulating components can directly utilize mature solutions already in the library, achieving continuous optimization.

[0097] The above scheme, after comparing the optimal combination of process parameters with the preset drying process library, updates the preset drying process library according to the optimal combination of process parameters if there is no suitable existing combination of process parameters in the library. This realizes the dynamic expansion and knowledge accumulation of the process library, solves the problem that new optimization schemes are difficult to systematically accumulate and reuse, and enables the control of kerosene residue in subsequent insulation components of the same type to continuously benefit from the constantly improving process library.

[0098] Please see Figure 5 In some embodiments, the kerosene vapor phase drying parameter optimization method further includes steps 502, 504, 506, and 508.

[0099] Step 502: Construct a process solution library.

[0100] Step 504: From the process scheme library, retrieve the initial process parameter combination corresponding to the current transformer voltage level and the current total weight range of insulation materials, and perform kerosene vapor phase drying treatment on the insulation components.

[0101] Step 506: Obtain the amount of kerosene residue on the insulating component after kerosene vapor phase drying treatment.

[0102] Step 508: If the kerosene residue exceeds the preset safety threshold, use the process parameter combination corresponding to the voltage level of the previous transformer and / or the total weight range of the previous insulation material as the initial process parameter combination for the current transformer voltage level and the current total weight range of the insulation material.

[0103] The process solution library is hierarchically constructed based on two dimensions: transformer voltage level and total weight range of insulation materials. The process solution library can be understood as a collection of standard process solutions hierarchically constructed based on these two dimensions. Each combination of transformer voltage level and total weight range of insulation materials maps to a set of core process parameters.

[0104] Transformer voltage level refers to the rated operating voltage of the transformer, such as 110kV, 220kV, 500kV, etc. In some embodiments, it can be divided into multiple continuous ranges, such as 132kV-250kV, 250kV-420kV, etc., which are not limited here. The total weight range of insulation materials refers to the numerical range of the total weight of all insulation materials in the transformer body.

[0105] It should be noted that in actual kerosene vapor phase drying treatment, a single transformer can be subjected to vapor phase drying treatment. In this case, the process can be constructed and analyzed based on the voltage level of a single transformer and the total weight of the insulation materials.

[0106] In another embodiment, multiple transformers may undergo kerosene vapor phase drying simultaneously. In this scenario, if the voltage levels of the transformers are different, the highest voltage level should be used as the transformer voltage level, serving as the basis for constructing the process scheme library and subsequent analysis. Correspondingly, in this scenario, the total weight range of insulation materials refers to the numerical range of the total weight of all insulation materials in the transformer bodies.

[0107] For example, participation can be combined Figure 9 In some embodiments, each transformer voltage level range can be further divided into multiple (two, three or more) continuous sub-ranges, such as 14t-20t, 20t-27t, etc., under the same transformer voltage level.

[0108] The initial process parameter combination refers to the standard process scheme retrieved from the process scheme library based on the current transformer voltage level and the total weight range of the insulation materials, which serves as the parameters for the first drying operation. The preset safety threshold is the maximum allowable kerosene residue set for a specific voltage level product; exceeding this value indicates non-compliance and triggers a fault-tolerant upgrade. The next higher transformer voltage level refers to the next higher range in the transformer voltage level classification, such as 250kV-420kV for the current level (e.g., 132kV-250kV for the current level). The next higher insulation material total weight range is the next higher range in the weight classification, such as 20t-27t for the current level (e.g., 20t-27t for the current level).

[0109] This embodiment constructs a process control system with intelligent fault tolerance capabilities, solving the problem of existing equipment lacking feedback decision-making and failure protection mechanisms. First, based on two dimensions—transformer voltage level and total weight range of insulation materials—a hierarchical process scheme library is pre-established. Each scheme is derived through the aforementioned modeling and optimization methods, with the safety margin increasing progressively at each level.

[0110] In actual production, the system automatically calls upon the corresponding initial process parameter combination for drying based on the current transformer voltage level and the total weight range of the insulation material. After drying, the residual kerosene level is checked. If it exceeds the preset safety threshold, the system automatically triggers the scheme upgrade process: the process parameter combination with more stringent process conditions corresponding to the next higher transformer voltage level and / or the next higher insulation material total weight range is used as the new initial process parameter combination for supplementary drying of the current product or for subsequent batches.

[0111] The above solution first constructs a process scheme library based on two dimensions: transformer voltage level and total weight range of insulation materials. Then, it calls the initial process parameter combination corresponding to the current transformer to perform kerosene vapor phase drying treatment. When the kerosene residue exceeds the preset safety threshold, it automatically upgrades to the process parameter combination corresponding to the next higher voltage level and / or the next higher total weight range of insulation materials as the new initial process parameter combination. This solves the problem of lacking feedback decision-making and failure protection mechanisms, and realizes dynamic fault tolerance and safety redundancy assurance.

[0112] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0113] Based on the same inventive concept, this application also provides a kerosene vapor phase drying parameter optimization device for implementing the above-mentioned kerosene vapor phase drying parameter optimization method. The solution provided by this device is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the kerosene vapor phase drying parameter optimization device provided below can be found in the limitations of the kerosene vapor phase drying parameter optimization method described above, and will not be repeated here.

[0114] Please see Figure 6This application provides a kerosene vapor phase drying parameter optimization device, including: a parameter acquisition module 601, a process determination module 602, a drying treatment module 603, a residue detection module 604, a model construction module 605, a target analysis module 606, and a parameter optimization module 607.

[0115] The parameter acquisition module 601 is used to determine the basic parameters of the target insulating component under test; the process determination module 602 is used to determine multiple process parameters affecting the kerosene adsorption capacity of the target insulating component under test, as well as the value range of each process parameter; the drying treatment module 603 is used to determine multiple sets of process parameter combinations based on the process parameters and value ranges, and to perform kerosene vapor-phase drying treatment on the target insulating component under each set of process parameter combinations; the residue detection module 604 is used to obtain the kerosene residue of the target insulating component under test after kerosene vapor-phase drying treatment, and to determine the insulation adsorption coefficient of the target insulating component under test based on the sample weight and the kerosene residue; the model construction module 605 is used to perform multiple regression analysis with the basic parameters and process parameter combinations as input variables and the insulation adsorption coefficient as output variable to determine the correlation model of process parameters-adsorption coefficient of the target insulating component under test; the target analysis module 606 is used to obtain the target kerosene residue and determine the target adsorption coefficient based on the target kerosene residue; the parameter optimization module 607 is used to perform reverse optimization based on the target adsorption coefficient and the correlation model to determine the optimal combination of process parameters.

[0116] Please see Figure 7 In some embodiments, the device also includes a model verification module 705.

[0117] The model validation module 705 is used to perform kerosene vapor phase drying treatment on the validation sample under the validation process parameter combination; obtain the kerosene residue of the validation sample and determine the measured insulation adsorption coefficient; determine the predicted insulation adsorption coefficient of the validation sample according to the correlation model; and validate the correlation model according to the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient.

[0118] In one embodiment, the model verification module 705 is further configured to determine the deviation between the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient; if the deviation is greater than a preset deviation threshold, multiple sets of process parameter combinations are redefined according to the process parameters and the value range, and the drying module 603 is controlled to perform kerosene vapor phase drying on the target insulation component under each set of process parameter combinations; if the deviation is less than or equal to the preset deviation threshold, the correlation model is determined to have passed verification.

[0119] In one embodiment, the device further includes an optimized drying module. The optimized drying module compares the optimal combination of process parameters with a preset drying process library; uses existing process parameter combinations in the preset drying process library that match the optimal combination of process parameters as the actual process parameter combination; and performs kerosene vapor-phase drying on insulating components of the same type as the target insulating component under test according to the actual process parameter combination.

[0120] In one embodiment, the optimized drying module is further configured to determine, based on the correlation model, the core influencing parameter that has the greatest impact on the insulation adsorption coefficient among all process parameters; and select from the preset drying process library the existing process parameter combination that has the smallest difference from the optimal process parameter combination in terms of the core influencing parameter, and the difference is less than or equal to the preset difference threshold, as the actual process parameter combination.

[0121] In one embodiment, the optimized drying module is further configured to update the preset drying process library according to the optimal process parameter combination when there is no existing process parameter combination in the preset drying process library that matches the optimal process parameter combination.

[0122] In one embodiment, the optimized drying module is also used to construct a process scheme library; from the process scheme library, it calls the initial process parameter combination corresponding to the current transformer voltage level and the current total weight range of insulation materials to perform kerosene vapor phase drying treatment on the insulation components; it obtains the kerosene residue on the insulation components after the kerosene vapor phase drying treatment; if the kerosene residue exceeds a preset safety threshold, it uses the process parameter combination corresponding to the previous level transformer voltage level and / or the previous level insulation material total weight range as the initial process parameter combination for the current transformer voltage level and the current total weight range of insulation materials.

[0123] Please see Figure 8 This application provides a kerosene vapor phase drying parameter optimization system, comprising: a sample parameter acquisition device 701, a kerosene vapor phase drying treatment device 702, an adsorption measurement device 703, and a control device 704. The sample parameter acquisition device 701 is used to determine the basic parameters of the target insulation component to be tested; the kerosene vapor phase drying treatment device 702 is used to perform kerosene vapor phase drying treatment on the target insulation component to be tested; the adsorption measurement device 703 is used to measure the kerosene residue in the target insulation component to be tested; the control device 704 is connected to the sample parameter acquisition device 701, the kerosene vapor phase drying treatment device 702, and the adsorption measurement device 703, and the control device 704 is used to execute the steps of the above-described kerosene vapor phase drying parameter optimization method.

[0124] The sample parameter acquisition device 701 refers to the hardware equipment used to determine the basic parameters (sample type, weight, specific surface area) of the target insulation component under test, such as an electronic balance or a specific surface area analyzer. The kerosene vapor phase drying treatment device 702 is a test or production equipment that can provide a controllable kerosene vapor phase drying environment for the target insulation component sample, including a heating system, a vacuum system, and a kerosene vapor generation and recovery system.

[0125] The adsorption measuring apparatus 703 refers to an instrument combination for determining the amount of kerosene residue in insulating components, including a high-precision analytical balance (for gravimetric analysis) and a gas chromatograph (for chromatographic analysis).

[0126] The control device 704 is a central control unit, which is electrically connected to each of the above-mentioned devices. It is used to execute the parameter optimization method and the steps of the parameter optimization method, coordinate the collaborative work of each unit, and includes a process scheme library module and a decision control module to realize the kerosene gas phase drying parameter optimization method.

[0127] This embodiment integrates the aforementioned measurement and optimization methods into a unified hardware and software system, solving the problem of isolated components and difficulty in forming a closed loop. The system includes a sample parameter acquisition device 701 (measuring the type, weight, and specific surface area of ​​the insulating component), a kerosene vapor phase drying treatment device 702 (providing a controllable drying environment), an adsorption measurement device 703 (measuring kerosene residue), and a control device 704. The control device 704, as the core, connects and coordinates all the above devices, automatically executing the aforementioned parameter optimization method steps (including experimental design, data acquisition, regression modeling, model validation, etc.), and / or executing the aforementioned parameter optimization method steps (including reverse optimization, process library comparison and adaptation, fault-tolerant upgrade decision-making, etc.).

[0128] The above solution, through the organic connection of the sample parameter acquisition device 701, the kerosene vapor phase drying treatment device 702, the adsorption measurement device 703, and the control device 704, allows the control device 704 to automatically execute the parameter optimization method and its steps, forming a complete system combining hardware and software. This solves the problem of separation between measurement, modeling, and optimization, making it difficult to form a closed loop. It achieves integrated intelligent control from basic parameter acquisition, drying treatment, residual amount measurement to model construction and parameter optimization, providing a system-level engineering solution for precise control of kerosene residual amount.

[0129] To facilitate understanding of the technical solution of this application, the following detailed embodiments will be used to explain and illustrate this application.

[0130] First, select an insulation component used in a typical transformer as the target insulation component to be tested, such as cellulose-based insulating crepe paper. Determine the basic parameters of this insulation component using a sample parameter acquisition device (such as an electronic balance and a specific surface area analyzer), including: sample type (e.g., cellulose), sample weight, and sample specific surface area (measured by gas adsorption method, in square meters per gram).

[0131] Then, based on the mechanism analysis of the kerosene vapor-phase drying process, several process parameters that significantly affect the kerosene adsorption capacity of insulating components were identified. For example, the total duration of the heating stage and the vacuum level of the high-vacuum stage were selected as key process parameters. For each process parameter, a reasonable range of values ​​was determined based on equipment capacity and process knowledge. For example, the range of the total heating duration was set from short to long (e.g., tens to hundreds of hours); the range of the vacuum level of the high-vacuum stage was set from low to high (e.g., tens to hundreds of Pascals).

[0132] Subsequently, experimental design methods (such as orthogonal experiments or uniform design) were employed to determine multiple different combinations of process parameters based on their ranges. Each combination included a specific value for the total heating time and the vacuum level during the high-vacuum stage. For each combination, a kerosene vapor phase drying process was used to dry the target insulation sample. The process strictly followed the standard procedures for kerosene vapor phase drying, including preparation, heating (including intermediate pressure reduction and distillation), high vacuum, and gas filling stages.

[0133] After drying, the residual kerosene content of the target insulating component is obtained. The determination method is adaptively selected based on the sample weight: if the sample weight is less than or equal to a preset weight threshold, a gravimetric method is used, where a high-precision analytical balance is used to directly weigh the mass difference of the sample before and after drying to obtain the residual kerosene content. If the sample weight is greater than the preset threshold, a chromatographic analysis method is used, where the sample is immersed in a suitable solvent to dissolve the kerosene, and then a gas chromatograph is used to determine the kerosene concentration in the solution, thereby calculating the residual kerosene content.

[0134] Based on the measured kerosene residue and the known sample weight, the insulation adsorption coefficient is calculated. It is defined as the kerosene residue divided by the sample weight. This coefficient quantitatively characterizes the degree of kerosene adsorption by the insulation component under the given set of process parameters.

[0135] Finally, the data obtained from all experimental groups were summarized. Using the basic parameters (such as sample specific surface area) and process parameter combinations (such as total heating time and vacuum degree) corresponding to each experimental group as input variables, and the calculated insulation adsorption coefficient as the output variable, a multiple regression analysis was performed. The regression model included not only independent influence terms for each input variable but also interaction terms between parameters (e.g., the product of total heating time and vacuum degree, or the product of sample specific surface area and total heating time). The coefficients of each term were determined through regression analysis, thereby constructing a correlation model between the process parameters and the adsorption coefficient of the target insulating component.

[0136] After parameter optimization, the correlation model can be validated. Specifically, an independent validation sample of the same type as the target insulation component sample but not involved in the modeling is selected. A different combination of validation process parameters than those used in the aforementioned modeling experiment is employed to perform kerosene vapor-phase drying on this validation sample. After treatment, the residual kerosene content is measured, and the measured insulation adsorption coefficient is calculated. Simultaneously, the basic parameters and validation process parameters of the validation sample are substituted into the established correlation model to calculate the predicted insulation adsorption coefficient.

[0137] Calculate the deviation between the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient, and compare this deviation with a pre-set deviation threshold. If the deviation is greater than the threshold, it indicates that the current model's prediction accuracy is insufficient. Then, based on the original process parameters and value ranges, determine more combinations of process parameters or different levels, conduct supplementary kerosene vapor phase drying experiments, and perform regression analysis again to optimize the correlation model. If the deviation is less than or equal to the threshold, the correlation model is considered validated and can be used for subsequent process parameter optimization.

[0138] If the correlation model is validated, the optimal combination of process parameters can be determined through reverse optimization. Specifically, for the insulating component to be produced that is the same type as the target insulating component to be tested, the target kerosene residue requirement is obtained. Based on the target kerosene residue and the total weight of the insulating component to be dried, the corresponding target adsorption coefficient (i.e., the target kerosene residue divided by the total weight) is calculated.

[0139] The target adsorption coefficient is used as a constraint and substituted into the validated correlation model. A numerical optimization algorithm is then used for inverse optimization. Within the range of process parameter values, a set of process parameter combinations that makes the insulation adsorption coefficient output by the model closest to the target adsorption coefficient is found and taken as the optimal process parameter combination.

[0140] The preset drying process library stores multiple sets of existing process parameter combinations that have been engineering-verified and can be directly executed on the equipment. The optimal process parameter combination obtained through reverse optimization is compared one by one with the existing solutions in this process library. Based on the coefficients in the correlation model, the parameter with the greatest impact on the insulation adsorption coefficient is determined, called the core influencing parameter. Then, from the preset drying process library, those existing process parameter combinations that have the smallest difference between the core influencing parameter and the optimal process parameter combination, and that difference is less than or equal to a preset difference threshold, are selected. If multiple solutions meet the conditions, the one with the smallest deviation in the core influencing parameter is selected as the actual process parameter combination.

[0141] If, after comparison, it is found that there is no existing process parameter combination in the preset drying process library that matches the optimal process parameter combination (i.e., the deviation of all existing schemes in the core influencing parameters exceeds the preset difference threshold), then the optimal process parameter combination is added to the preset drying process library as a new standard process scheme after standardization processing (e.g., rounding or segmentation and merging), thereby realizing the dynamic expansion of the process library.

[0142] Furthermore, a process scheme library can be pre-built, which is hierarchically constructed according to two dimensions: the voltage level range of the transformer and the total weight range of the insulation material. Each cell corresponds to a set of recommended initial process parameter combinations.

[0143] In actual production, for the transformer currently awaiting drying, based on its voltage level and the total weight range of the insulation material, the corresponding initial process parameter combination is retrieved from the process scheme library to perform kerosene vapor phase drying on the insulation components. After drying, the residual kerosene content in the insulation components is checked. If the residual content exceeds the preset safety threshold for products of that voltage level, a fault-tolerant upgrade process is automatically triggered. Specifically, a more stringent process parameter combination corresponding to the next higher transformer voltage level (i.e., a higher voltage level range) and / or the next higher insulation material total weight range (i.e., a larger weight range) is used as a new initial process parameter combination for the current transformer voltage level and the current insulation material total weight range. This combination is used for supplementary drying of the current product or for the production of subsequent products of the same specification.

[0144] In some embodiments, this application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0145] The process involves: determining the basic parameters of the target insulating component; identifying multiple process parameters affecting the kerosene adsorption capacity of the target insulating component, and defining the value range of each parameter; determining multiple combinations of process parameters based on these parameters and their ranges, and subjecting the target insulating component to kerosene vapor-phase drying treatment under each parameter combination; obtaining the kerosene residue on the target insulating component after kerosene vapor-phase drying treatment, and determining the insulation adsorption coefficient based on the sample weight and kerosene residue; performing a multiple regression analysis using the basic parameters and process parameter combinations as input variables and the insulation adsorption coefficient as the output variable to determine the correlation model between the process parameters and the adsorption coefficient of the target insulating component; obtaining the target kerosene residue and determining the target adsorption coefficient based on the target kerosene residue; and performing inverse optimization based on the target adsorption coefficient and the correlation model to determine the optimal combination of process parameters.

[0146] In some embodiments, this application provides a computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0147] The process involves: determining the basic parameters of the target insulating component; identifying multiple process parameters affecting the kerosene adsorption capacity of the target insulating component, and defining the value range of each parameter; determining multiple combinations of process parameters based on these parameters and their ranges, and subjecting the target insulating component to kerosene vapor-phase drying treatment under each parameter combination; obtaining the kerosene residue on the target insulating component after kerosene vapor-phase drying treatment, and determining the insulation adsorption coefficient based on the sample weight and kerosene residue; performing a multiple regression analysis using the basic parameters and process parameter combinations as input variables and the insulation adsorption coefficient as the output variable to determine the correlation model between the process parameters and the adsorption coefficient of the target insulating component; obtaining the target kerosene residue and determining the target adsorption coefficient based on the target kerosene residue; and performing inverse optimization based on the target adsorption coefficient and the correlation model to determine the optimal combination of process parameters.

[0148] In some embodiments, this application provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0149] The process involves: determining the basic parameters of the target insulating component; identifying multiple process parameters affecting the kerosene adsorption capacity of the target insulating component, and defining the value range of each parameter; determining multiple combinations of process parameters based on these parameters and their ranges, and subjecting the target insulating component to kerosene vapor-phase drying treatment under each parameter combination; obtaining the kerosene residue on the target insulating component after kerosene vapor-phase drying treatment, and determining the insulation adsorption coefficient based on the sample weight and kerosene residue; performing a multiple regression analysis using the basic parameters and process parameter combinations as input variables and the insulation adsorption coefficient as the output variable to determine the correlation model between the process parameters and the adsorption coefficient of the target insulating component; obtaining the target kerosene residue and determining the target adsorption coefficient based on the target kerosene residue; and performing inverse optimization based on the target adsorption coefficient and the correlation model to determine the optimal combination of process parameters.

[0150] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0151] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for optimizing kerosene vapor phase drying parameters, characterized in that, include: Determine the basic parameters of the target insulating component to be tested; wherein, the basic parameters include at least one of sample type, sample weight, and sample specific surface area; Determine multiple process parameters that affect the kerosene adsorption capacity of the target insulating component under test, and the value range of each process parameter; Based on the process parameters and the value range, multiple sets of process parameter combinations are determined, and the target insulation component to be tested is subjected to kerosene vapor phase drying treatment under each set of process parameter combinations. After obtaining the kerosene vapor phase drying treatment, the kerosene residue of the target insulating component is obtained, and the insulation adsorption coefficient of the target insulating component is determined based on the sample weight and the kerosene residue. Using the combination of the basic parameters and the process parameters as input variables and the insulation adsorption coefficient as the output variable, a multiple regression analysis is performed to determine the correlation model between the process parameters and the adsorption coefficient of the target insulating component under test; wherein, the correlation model characterizes the quantitative influence of the basic parameters and the process parameters on the insulation adsorption coefficient; Obtain the target kerosene residue amount, and determine the target adsorption coefficient based on the target kerosene residue amount; The optimal combination of process parameters is determined by reverse optimization based on the target adsorption coefficient and the correlation model; wherein the optimal combination of process parameters is used to perform kerosene vapor phase drying on insulating components of the same type as the target insulating component to be tested.

2. The method for optimizing kerosene vapor phase drying parameters according to claim 1, characterized in that, The method further includes: The verification sample is subjected to kerosene vapor phase drying under the verification process parameter combination; wherein the verification sample is of the same type as the target insulation component to be tested, and the verification process parameter combination is different from the process parameter combination. Obtain the kerosene residue of the verification sample and determine the measured insulation adsorption coefficient; Based on the correlation model, the predicted insulation adsorption coefficient of the verification sample is determined; The correlation model is validated based on the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient.

3. The method for optimizing kerosene vapor phase drying parameters according to claim 2, characterized in that, The step of validating the correlation model based on the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient includes: Determine the deviation between the measured insulation adsorption coefficient and the predicted insulation adsorption coefficient; If the deviation value is greater than the preset deviation threshold, multiple sets of process parameter combinations are re-determined according to the process parameters and the value range, and the process returns to the step of performing kerosene vapor phase drying treatment on the target insulation component under each set of process parameter combinations. If the deviation value is less than or equal to the preset deviation threshold, the correlation model is determined to have passed verification.

4. The method for optimizing kerosene vapor phase drying parameters according to any one of claims 1-3, characterized in that, The correlation model includes independent influence terms of the combination of the basic parameters and the process parameters on the insulation adsorption coefficient, as well as interactive influence terms of the combination of the basic parameters and the process parameters on the insulation adsorption coefficient.

5. The method for optimizing kerosene vapor phase drying parameters according to any one of claims 1-3, characterized in that, The method further includes: The optimal combination of process parameters is compared with a preset drying process library; The existing process parameter combinations that match the optimal process parameter combination in the preset drying process library are used as the actual process parameter combinations. Based on the actual process parameter combination, kerosene vapor phase drying is performed on the insulating component of the same type as the target insulating component to be tested.

6. The method for optimizing kerosene vapor phase drying parameters according to claim 5, characterized in that, The step of using existing process parameter combinations from the preset drying process library that match the optimal process parameter combination as the actual process parameter combination includes: Based on the correlation model, the core influencing parameter with the greatest impact on the insulation adsorption coefficient among the process parameters is determined; From the preset drying process library, select the existing process parameter combination that has the smallest difference from the optimal process parameter combination in terms of the core influencing parameter, and whose difference is less than or equal to the preset difference threshold, as the actual process parameter combination.

7. The method for optimizing kerosene vapor phase drying parameters according to claim 5, characterized in that, After comparing the optimal combination of process parameters with a preset drying process library, the method further includes: If there is no existing process parameter combination in the preset drying process library that matches the optimal process parameter combination, the preset drying process library is updated according to the optimal process parameter combination.

8. The method for optimizing kerosene vapor phase drying parameters according to any one of claims 1-3, characterized in that, Also includes: A process solution library is constructed; wherein, the process solution library is constructed hierarchically based on two dimensions: transformer voltage level and total weight range of insulation materials. From the process scheme library, the initial process parameter combination corresponding to the current transformer voltage level and the current total weight range of insulation material is called to perform kerosene vapor phase drying treatment on the insulation components; The amount of kerosene residue in the insulating component after kerosene vapor phase drying treatment was obtained; If the kerosene residue exceeds a preset safety threshold, the combination of process parameters corresponding to the voltage level of the previous transformer and / or the total weight range of the previous insulation material shall be used as the initial combination of process parameters for the current transformer voltage level and the current total weight range of the insulation material.

9. A device for optimizing kerosene vapor phase drying parameters, characterized in that, include: The parameter acquisition module is used to determine the basic parameters of the target insulating component under test; wherein, the basic parameters include at least one of sample type, sample weight, and sample specific surface area; The process determination module is used to determine multiple process parameters that affect the kerosene adsorption capacity of the target insulation component under test, as well as the value range of each process parameter. The drying module is used to determine multiple combinations of process parameters based on the process parameters and the value range, and to perform kerosene vapor phase drying on the target insulation component under each combination of process parameters. The residual detection module is used to obtain the amount of kerosene residue in the target insulating component after kerosene vapor phase drying treatment, and to determine the insulation adsorption coefficient of the target insulating component based on the sample weight and the amount of kerosene residue. The model building module is used to perform multiple regression analysis with the combination of the basic parameters and the process parameters as input variables and the insulation adsorption coefficient as output variable to determine the correlation model of the process parameters-adsorption coefficient of the target insulation component under test; wherein, the correlation model characterizes the quantitative influence of the basic parameters and the process parameters on the insulation adsorption coefficient; The target analysis module is used to obtain the target kerosene residue and determine the target adsorption coefficient based on the target kerosene residue. The parameter optimization module is used to perform reverse optimization based on the target adsorption coefficient and the correlation model to determine the optimal combination of process parameters; wherein, the optimal combination of process parameters is used to perform kerosene vapor phase drying on insulating components of the same type as the target insulating component to be tested.

10. A kerosene vapor phase drying parameter optimization system, characterized in that, include: The sample parameter acquisition device is used to determine the basic parameters of the target insulating component under test. A kerosene vapor phase drying treatment device is used to perform kerosene vapor phase drying treatment on the target insulating component to be tested; An adsorption measuring device is used to determine the amount of kerosene residue in the target insulating component. A control device is connected to the sample parameter acquisition device, the kerosene vapor phase drying device, and the adsorption measurement device. The control device is used to execute the steps of the kerosene vapor phase drying parameter optimization method according to any one of claims 1-8.

Images