A polynomial-based insulator structure optimization method and device

Abstract

This invention provides a method and apparatus for optimizing insulator structures based on a polynomial surrogate model, relating to the field of electrical engineering technology. The method includes acquiring first and second information; performing correlation analysis on the insulator icing thickness in the first and second information; inputting preset target meteorological observation data into an insulator model and using an optimization algorithm to optimize the insulator icing thickness, obtaining optimized initial insulator structure parameters and the corresponding target icing thickness; and using the initial insulator structure parameters as the target insulator structure parameters. This invention, by analyzing meteorological data, insulator structure parameters, and insulator icing thickness data, utilizes a polynomial surrogate model to construct and train an insulator structure model, effectively optimizing the insulator structure to achieve precise control of the target icing thickness. Furthermore, it evaluates the insulator structure performance through a physical model, thereby ensuring the reliability of the insulator structure design and effectively reducing the risk of icing flashover.

Description

Technical Field

[0001] This invention relates to the field of electrical engineering technology, and more specifically, to a method, apparatus, device, and readable storage medium for optimizing insulator structures based on polynomial vehicles. Background Technology

[0002] In power systems, vehicle roof insulators are critical electrical components, and their performance directly affects the stability and safety of the power system. Especially in cold, snowy regions, vehicle roof insulators are susceptible to icing, leading to decreased insulation performance and even flashover accidents, posing a serious threat to the normal operation of the power system. Therefore, optimizing the structure of vehicle roof insulators to improve their anti-icing capabilities has become an urgent problem to be solved.

[0003] Traditionally, the design of roof insulators has relied mainly on empirical formulas and experimental verification. This method is not only time-consuming and costly, but also makes it difficult to fully consider the influence of multiple physical fields such as electric field distribution, heat conduction, mechanical stress and ice thickness under different conditions. As a result, the accuracy of insulator structure optimization is not high. Therefore, we propose a polynomial-based structural optimization method for roof insulators to prevent ice flashover. Summary of the Invention

[0004] The purpose of this invention is to provide a method, apparatus, device, and readable storage medium for optimizing insulator structures based on polynomial vehicles, in order to improve the aforementioned problems. To achieve the above objective, the technical solution adopted by this invention is as follows:

[0005] In a first aspect, this application provides an insulator structure optimization method based on polynomial vehicles, including:

[0006] Acquire first information and second information, wherein the first information includes meteorological observation data of the insulator working area and the second information includes insulator structural parameters and insulator ice thickness;

[0007] A correlation analysis was performed on the insulator icing thickness in the first information and the second information to obtain the analyzed first information, which includes meteorological observation data that has a significant impact on the insulator icing thickness.

[0008] Based on the first and second information obtained from the analysis, a preset multinomial surrogate model is trained. When the preset loss function meets the set conditions, the training is stopped, and an insulator model is obtained.

[0009] The preset target meteorological observation data is input into the insulator model and the optimization algorithm is used to optimize the ice thickness of the insulator, so as to obtain the optimized initial insulator structure parameters and the corresponding target ice layer thickness.

[0010] The initial insulator structure parameters and the target icing layer thickness are verified based on a preset physical model to obtain verification results. When the verification results meet preset conditions, the initial insulator structure parameters are used as the target insulator structure parameters.

[0011] Secondly, this application also provides an insulator structure optimization device based on a polynomial vehicle, comprising:

[0012] The acquisition unit is used to acquire first information and second information. The first information includes meteorological observation data of the working area of ​​the insulator, and the second information includes insulator structural parameters and insulator ice thickness.

[0013] An analysis unit is used to perform correlation analysis on the insulator icing thickness in the first information and the second information to obtain the analyzed first information, which includes meteorological observation data that has a significant impact on the insulator icing thickness.

[0014] The training unit is used to train a preset polynomial surrogate model based on the analyzed first information and the second information. When the preset loss function meets the set conditions, the training stops and the insulator model is obtained.

[0015] The optimization unit is used to input the preset target meteorological observation data into the insulator model and use the optimization algorithm to optimize the ice thickness of the insulator, so as to obtain the optimized initial insulator structure parameters and the corresponding target ice layer thickness.

[0016] The verification unit is used to verify the initial insulator structure parameters and the target icing layer thickness based on a preset physical model to obtain the verification result. When the verification result meets the preset conditions, the initial insulator structure parameters are used as the target insulator structure parameters.

[0017] The beneficial effects of this invention are as follows:

[0018] This invention analyzes meteorological data, insulator structural parameters, and insulator icing thickness data. It uses a polynomial surrogate model to construct an insulator structural model and trains and optimizes it, which can effectively optimize the insulator structure to achieve precise control of the target icing thickness. Furthermore, it evaluates the performance of the insulator structure through a physical model, thereby ensuring the reliability of the insulator structure design and effectively reducing the risk of icing flashover.

[0019] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing embodiments of the invention. Attached Figure Description

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 This is a schematic diagram of the insulator structure optimization method based on polynomial vehicle described in an embodiment of the present invention;

[0022] Figure 2 This is a schematic diagram of the insulator structure optimization device based on polynomial vehicle described in an embodiment of the present invention.

[0023] Marked in the image:

[0024] 10. Acquisition Unit; 20. Analysis Unit; 30. Training Unit; 40. Optimization Unit; 50. Validation Unit. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0026] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0027] Example 1:

[0028] This embodiment provides an insulator structure optimization method based on polynomial vehicles.

[0029] See Figure 1 The figure shows that the method includes steps S10, S20, S30, S40 and S50.

[0030] Step S10. Obtain first information and second information. The first information includes meteorological observation data of the insulator working area, and the second information includes insulator structural parameters and insulator icing thickness.

[0031] Specifically, the ice thickness of insulators under different climatic conditions and with varying structural parameters is collected to provide data support for subsequent model training. Climatic conditions refer to meteorological observation data, which can be obtained from publicly available data from weather stations, typically including ambient temperature, humidity, wind speed, wind direction, precipitation, and air pressure. Insulator structural parameters include the insulator's skirt curvature radius, skirt inclination angle, skirt spacing, and rod diameter. Ice thickness is usually collected using an ultrasonic thickness gauge, and cameras can also be used to periodically photograph the ice accumulation on vehicle roof insulators, recording changes in the ice's shape.

[0032] Step S20. Perform correlation analysis on the insulator icing thickness in the first information and the second information to obtain the analyzed first information, which includes meteorological observation data that have a significant impact on the insulator icing thickness;

[0033] Specifically, meteorological observation data contains multiple meteorological data, but this application only considers meteorological data that has a significant impact on the thickness of ice accretion on insulators for subsequent model building, thereby effectively improving the accuracy and computational efficiency of the model.

[0034] Specifically, step S20 includes steps S21, S22, S23, S24, S25, and S26:

[0035] Step S21. Divide the multiple meteorological data contained in the first information into multiple partitioning strategies to obtain multiple partitioning combinations. The partitioning strategy ensures that the combination of meteorological data in each partitioning combination is unique and that each meteorological data can appear in multiple partitioning combinations.

[0036] Specifically, a predetermined segmentation strategy is used to divide meteorological data into multiple unique segmentation combinations. Each combination contains a unique set of meteorological data, but each meteorological data point can appear in multiple segmentation combinations. This segmentation strategy helps to comprehensively consider the impact of different combinations of meteorological data on icing thickness, while avoiding overlooking certain specific data combinations.

[0037] Step S22. Input each partition combination into the preset feature transformation model for feature transformation to obtain multiple first feature vectors;

[0038] Specifically, through feature transformation, each partition combination can be effectively dimensionality reduced and key features can be extracted, thereby better reflecting the feature vectors corresponding to the meteorological data contained in each partition combination, providing more valuable data for subsequent correlation analysis and insulator structure optimization.

[0039] Step S23. Input the insulator ice thickness in each second piece of information into the preset feature extraction model to extract features and obtain multiple second feature vectors;

[0040] Specifically, feature extraction of insulator icing thickness can capture meteorological data that has the most significant impact on insulator icing thickness, thus helping to establish a more accurate insulator model.

[0041] Specifically, step S23 includes steps S231, S232, S233, and S234:

[0042] Step S231. Divide the second information into multiple levels, and compare the importance of the data corresponding to each pair of levels to construct a judgment matrix. There is an inclusion relationship between the multiple levels.

[0043] Step S232. Calculate the weight of each level in the judgment matrix based on the analytic hierarchy process (AHP) to determine the importance value of each data point;

[0044] Step S233. Sort the data from largest to smallest based on multiple importance values, and use the ice thickness of the insulators located before the preset sorting position as feature information;

[0045] Step S234. Normalize all feature information to obtain multiple second feature vectors;

[0046] Specifically, this application divides the insulator icing thickness data into multiple levels and calculates the weights using the analytic hierarchy process (AHP) to select the insulator icing thickness information that contributes the most to the features. By sorting and extracting key feature information and normalizing it, a second feature vector for model training is generated. This effectively highlights important data, reduces redundant information, improves the accuracy of feature extraction and training efficiency, and ensures the predictive performance and generalization ability of the constructed model.

[0047] Step S24. Perform correlation analysis based on the first feature vector of each partition combination and the second feature vector corresponding to each second piece of information to obtain multiple correlation values;

[0048] Step S25. Determine the maximum correlation value from multiple correlation values, and take all meteorological correlation data in the partition combination corresponding to the maximum correlation value as the target correlation data;

[0049] Step S26. Delete the meteorological data in the first information except for the target-related data to obtain the analyzed first information;

[0050] Specifically, correlation analysis was used to screen out some meteorological data that significantly affected the icing thickness of insulators, and meteorological data with less impact on the icing thickness of insulators were deleted. This optimized the primary information, reduced redundant information, and ensured the accurate capture of key meteorological data.

[0051] Step S30. Train the preset polynomial surrogate model based on the analyzed first and second information. When the preset loss function meets the set conditions, stop training and obtain the insulator model.

[0052] Specifically, the surrogate model, also known as the approximation model, works by constructing a relatively simple function to replace the real functional relationship. In this application, a polynomial surrogate model is used to approximate the insulator model, which has a certain interaction effect. The preset polynomial surrogate model consists of a first polynomial and a second polynomial. The first polynomial consists of insulator structural parameters, and the second polynomial consists of target-related data.

[0053] Specifically, step S30 includes steps S31, S32, S33, S34, S35, S36, S37, S38, and S39:

[0054] Step S31. Based on a single term composed of a single insulator structural parameter and a preset first coefficient, multiple terms are obtained, each with a different first coefficient and a different exponent value of the insulator structural parameter in each term.

[0055] Step S32. Based on the interaction terms composed of multiple insulator structure parameters and preset second coefficients, multiple interaction terms are obtained. In each pair of interaction terms, at least one insulator structure parameter is different. The exponent value of each insulator structure parameter in the interaction terms is a set threshold.

[0056] Step S33. Form a first polynomial based on a preset constant term, multiple single terms, and multiple interaction terms;

[0057] Step S34. Based on the single target-related data and the preset third coefficient, multiple environmental impact items are obtained. Each third coefficient in the environmental impact items is different, and the index value of the target-related data in each environmental impact item is a set threshold.

[0058] Step S35. Based on the action term composed of multiple target-related data and a preset fourth coefficient, an environmental action term is obtained, where the index value of each target-related data in the environmental action term is a set threshold.

[0059] Step S36. Construct a second polynomial based on multiple environmental impact terms and environmental action terms;

[0060] Step S37. Construct a preset polynomial proxy model based on the first polynomial, the second polynomial, and the insulator icing thickness;

[0061] Step S38. Input the analyzed first and second information into the preset multinomial proxy model for calculation, and calculate the preset loss function to obtain the loss function value;

[0062] Step S39. Adjust the values ​​of the coefficients in the preset polynomial surrogate model until the loss function value meets the set conditions to obtain the insulator model. The coefficients include the first coefficient, the second coefficient, the third coefficient, and the fourth coefficient.

[0063] Specifically, considering that there are five insulator structural parameters in this application, the highest degree of the polynomial is generally set to 4. The expression of the first polynomial constructed based on the insulator structural parameters is as follows:

[0064]

[0065] Where y1 is the first polynomial; b0 is the constant term; b i b is the coefficient when only the structural parameters of the i-th insulator are considered; ij To account for the coefficients when considering the i-th and j-th insulator structural parameters, b ijzs The coefficients for considering the i-th, j-th, z-th, and s-th insulator structural parameters; b ii b represents the coefficient of the quadratic term when only the structural parameters of the i-th insulator are considered; iiii Let n be the coefficient of the fourth term when only the structural parameters of the i-th insulator are considered; n = 5.

[0066] A second polynomial is constructed based on environmental target-related data that significantly affect the insulator icing thickness. In this application, only the effects of temperature and humidity on the insulator thickness are considered, and the corresponding expression is:

[0067]

[0068] Where y2 is the second polynomial; a is the coefficient when only the data related to the i-th target are considered; a ji The coefficients are used to consider the parameters related to the i-th and j-th objectives; m = 2.

[0069] The expression for the polynomial proxy model is:

[0070] Y = y1 + y2

[0071] Where Y is the ice thickness of the insulator; y1 is the first polynomial; and y2 is the second polynomial.

[0072] By setting the range of values ​​for each coefficient in the polynomial surrogate model, polynomial surrogate models with different combinations of polynomial coefficients are generated using SolidWorks computer-aided design software. The first and second information after analysis are then input into multiple polynomial surrogate models, and the combinations of polynomial coefficients are continuously adjusted based on the model prediction results. Finally, the polynomial surrogate model corresponding to the optimal combination of polynomial coefficients is determined as the insulator model.

[0073] Step S40. Input the preset target meteorological observation data into the insulator model and use the optimization algorithm to optimize the ice thickness of the insulator to obtain the optimized initial insulator structure parameters and the corresponding target ice layer thickness;

[0074] Specifically, when using the insulator model in practice, target meteorological data needs to be input. At this time, the minimum insulator icing thickness is used as the optimization target. The values ​​of each parameter in the insulator structure parameters will be continuously adjusted. Finally, the minimum insulator icing thickness is calculated, and the corresponding insulator structure parameters are the optimal target insulator structure parameters under this meteorological environment.

[0075] When adjusting the structural parameters of an insulator, it is necessary to ensure that the structural parameters of the insulator meet certain conditions to ensure the mechanical strength requirements of the insulator. In this application, after adjusting the structural parameters of the insulator, it is necessary to first calculate the creepage distance of the insulator. Only after the creepage distance meets the set conditions can the subsequent calculation of the ice thickness of the insulator be carried out.

[0076] Considering that the creepage distance of roof insulators is related not only to factors such as voltage, material, and pollution level, but also closely related to the insulator's structural parameters, such as the radius of curvature of the insulator skirts, the vertical tilt angle of the skirts, the spacing between the skirts, and the rod diameter, all of these factors significantly affect the creepage distance. Therefore, taking into account all five parameters of the insulator's structural parameters, the formula for calculating the creepage distance of roof insulators is as follows:

[0077]

[0078] Where C is the creepage distance of the insulator; k is an empirical coefficient; U r R is the voltage across the insulator; θ is the radius of curvature of the skirt; θ is the vertical tilt angle of the skirt; D is the rod diameter; d is the skirt spacing.

[0079] In this application, the values ​​of some parameters in the insulator structure parameters are as follows: the radius of curvature of the shed is 0.1cm to 1.0cm, the shed spacing is 0.05cm to 0.2cm, and the rod diameter is 20mm to 100mm.

[0080] Step S50. Based on the preset physical model, the initial insulator structure parameters and the target icing layer thickness are verified to obtain the verification results. When the verification results meet the preset conditions, the initial insulator structure parameters are used as the target insulator structure parameters.

[0081] Specifically, the initial insulator structural parameters calculated from the insulator model are imported into a preset physical model for simulation analysis. The preset physical model includes an electric field distribution model, a heat conduction model, and a mechanical stress model, which can calculate the corresponding electric field strength, surface temperature, and mechanical stress respectively. These physical parameters are then used to further determine whether the insulator structural parameters meet the actual usage requirements.

[0082] In this application, the electric field distribution model was established using ANSYS Maxwell software, the heat conduction model was established using ANSYS Thermal software, and the mechanical stress model was established using ANSYS Mechanical software.

[0083] In electric field distribution model simulations, the local electric field enhancement effect is an important physical phenomenon, especially in the presence of an icing layer. The combined effects of the icing layer's geometry, thickness, surface moisture, and temperature can cause the electric field to become abnormally strong in certain local areas, leading to electrical faults such as insulator surface breakdown or flashover. Therefore, accurately simulating and calculating the local electric field enhancement effect is crucial for predicting insulator performance under icing conditions.

[0084] Specifically, step S50 includes steps S51, S52, S53, S54, and S55:

[0085] Step S51. Estimate the radius of the ice tip based on the target ice layer thickness;

[0086] Step S52. Calculate the ratio of the radius of the ice tip to the thickness of the target ice layer, and perform an exponential operation on the ratio to obtain the first result. The value of the exponent is within the preset range.

[0087] Step S53. Calculate the water film thickness on the ice layer based on the target meteorological observation data and the target ice layer thickness;

[0088] Step S54. Calculate the ratio of water film thickness to target ice layer thickness, and perform exponential calculation on the ratio to obtain a second result. The value of the exponent is within a preset range.

[0089] Step S55. Calculate the product of the first result, the second result, and the overall electric field strength to obtain the local electric field strength;

[0090] Specifically, considering that the local electric field strength increases significantly at sharp points or in regions with high curvature, and the degree of increase is related to the ratio of ice thickness to the radius of curvature at the sharp point, under high humidity conditions, water films may form on the ice surface. These water films have high conductivity, which affects the distribution of the electric field and thus leads to an enhancement of the local electric field. The conductivity and dielectric constant of the water film differ significantly from those of the ice layer itself; therefore, the electric field may exhibit an enhanced effect in the water film region.

[0091] Therefore, combining the geometric effect and the water film effect, the mathematical model that comprehensively considers the local electric field enhancement effect is as follows:

[0092]

[0093] Among them, E local r represents the local electric field strength. tip h is the radius of the ice tip; ice h represents the target ice layer thickness. water E represents the thickness of the water film on the ice layer. global denoted as the overall electric field strength; o and p are correction coefficients, ranging from 0.5 to 1.

[0094] In the simulation of the electric field distribution model of the roof insulator, accurately describing the change in conductivity of the insulator's ice layer is crucial, as the conductivity of the ice layer varies significantly under different environmental conditions. The conductivity of the ice layer depends not only on temperature, ice purity, and density, but also on the ice layer's humidity, surface water layer, and climatic factors (such as precipitation and wind speed). Therefore, this application innovatively establishes a comprehensive model to dynamically simulate the change in ice layer conductivity and its impact on the electric field.

[0095] Specifically, step S50 includes steps S56, S57, S58, S59, S510, and S511:

[0096] Step S56. Calculate the product of the preset temperature coefficient and the temperature in the first information after analysis to obtain the first product. The preset temperature coefficient is used to characterize the effect of temperature on the conductivity of the ice layer.

[0097] Step S57. Calculate the product of the preset humidity coefficient and the humidity in the first information after analysis to obtain the second product. The preset humidity coefficient is used to characterize the effect of humidity on the conductivity of the ice layer.

[0098] Step S58. Calculate the product of the preset thickness coefficient and the target ice layer thickness to obtain the third product. The preset thickness coefficient is used to characterize the influence of ice layer thickness on the conductivity of the ice layer.

[0099] Step S59. Calculate the product of the Boltzmann constant and the temperature in the first piece of information after analysis to obtain the fourth product;

[0100] Step S510. Calculate the ratio of the preset activation energy of ice to the fourth product to obtain the target ratio. The preset activation energy of ice is used to characterize the energy threshold at which the conductivity of the ice layer changes significantly.

[0101] Step S511. Calculate the conductivity based on the preset basic ice conductivity, the first product, the second product, the third product, and the target ratio;

[0102] Specifically, this application mainly considers the effects of temperature, humidity and ice layer thickness on electrical conductivity.

[0103] (1) Temperature: As the temperature rises, the crystal structure of ice in the ice layer changes, especially the movement of water molecules in the ice intensifies, which leads to an increase in the electrical conductivity of the ice layer.

[0104] (2) Humidity: When the surface moisture of the ice layer increases, i.e., when the humidity is high, water molecules will promote ionic conductivity, so the conductivity of the ice layer will increase significantly. In addition, under high humidity conditions, the ice layer may form a thin water film, the conductivity of which is much higher than that of the ice body.

[0105] (3) Ice layer thickness: Generally, a thinner ice layer has a stronger local electric field enhancement effect, while a thicker ice layer makes the current more evenly distributed throughout the ice layer.

[0106] In summary, the variation in electrical conductivity of the ice layer is influenced by the combined effects of temperature, humidity, and ice thickness. During the simulation of the electric field distribution model, the following formula can be used to accurately describe the conductivity:

[0107]

[0108] Where, σ ice σ0 is the electrical conductivity of the ice layer; σ0 is the basic electrical conductivity of the ice; β is the temperature coefficient; T is the temperature; α is the humidity coefficient. γ represents humidity; h represents thickness coefficient; γ represents humidity. ice E represents the target ice layer thickness. a The activation energy for ice; k B is the Boltzmann constant; exp is the exponential function.

[0109] After determining the target insulator structure parameters that satisfy both the minimum icing thickness and the physical model verification through the above model, further verification can be carried out through experiments. Specifically, based on the target insulator structure parameters, corresponding insulator samples are made and tested in the laboratory. The tests include detecting the electrical properties of the samples, such as insulation resistance, breakdown voltage, and leakage current, as well as the mechanical properties, such as tensile strength, flexural strength, and impact toughness. The samples are then tested under icing conditions in an artificial climate chamber to simulate the target meteorological observation conditions. The insulator structure parameters can be optimized again based on the test results.

[0110] Example 2:

[0111] like Figure 2 As shown, this embodiment provides an insulator structure optimization device based on a polynomial vehicle. The device includes:

[0112] The acquisition unit is used to acquire first information and second information. The first information includes meteorological observation data of the working area of ​​the insulator, and the second information includes insulator structural parameters and insulator ice thickness.

[0113] The analysis unit is used to perform correlation analysis on the insulator icing thickness in the first information and the second information to obtain the analyzed first information, which includes meteorological observation data that have a significant impact on the insulator icing thickness.

[0114] The training unit is used to train a preset polynomial surrogate model based on the analyzed first and second information. When the preset loss function meets the set conditions, the training stops and the insulator model is obtained.

[0115] The optimization unit is used to input the preset target meteorological observation data into the insulator model and use the optimization algorithm to optimize the ice thickness of the insulator, so as to obtain the optimized initial insulator structure parameters and the corresponding target ice layer thickness.

[0116] The verification unit is used to verify the initial insulator structure parameters and the target icing layer thickness based on a preset physical model to obtain the verification results. When the verification results meet the preset conditions, the initial insulator structure parameters are used as the target insulator structure parameters.

[0117] In one specific embodiment disclosed in this application, the analysis unit includes:

[0118] The first partitioning unit is used to partition the multiple meteorological data contained in the first information using a predetermined partitioning strategy to obtain multiple partitioning combinations. The partitioning strategy ensures that the combination of meteorological data in each partitioning combination is unique, and each meteorological data can appear in multiple partitioning combinations.

[0119] The transformation unit is used to input each partition combination into a preset feature transformation model for feature transformation to obtain multiple first feature vectors;

[0120] The extraction unit is used to input the insulator ice thickness in each second piece of information into a preset feature extraction model to extract features and obtain multiple second feature vectors.

[0121] The correlation unit is used to perform correlation analysis based on the first feature vector of each partition combination and the second feature vector corresponding to each second information, and to obtain multiple correlation values.

[0122] The determination unit is used to determine the maximum correlation value from multiple correlation values, and to take all meteorological correlation data in the partition combination corresponding to the maximum correlation value as the target correlation data;

[0123] The deletion unit is used to delete meteorological data other than target-related data from the first information to obtain the analyzed first information.

[0124] In one specific embodiment disclosed in this application, the extraction unit includes:

[0125] The second partitioning unit is used to divide the second information into multiple levels, and to construct a judgment matrix based on the importance of the data corresponding to each pair of levels. There is an inclusion relationship between the multiple levels.

[0126] The judgment unit is used to calculate the weight of each level in the judgment matrix based on the analytic hierarchy process (AHP) to determine the importance value of each data point.

[0127] The sorting unit is used to sort the data from largest to smallest based on multiple importance values, and uses the ice thickness of the insulators located before the preset sorting position as feature information.

[0128] The normalization unit is used to normalize all feature information to obtain multiple second feature vectors.

[0129] In one specific embodiment disclosed in this application, the training unit includes:

[0130] The first component unit is used to obtain multiple individual terms based on a single insulator structural parameter and a preset first coefficient. Each first coefficient in an individual term is different, and the exponent value of the insulator structural parameter in each individual term is different.

[0131] The second component unit is used to obtain multiple interaction terms based on multiple insulator structure parameters and preset second coefficients. In each pair of interaction terms, at least one insulator structure parameter is different. The exponent value of each insulator structure parameter in the interaction term is a set threshold.

[0132] The third component unit is used to form the first polynomial based on a preset constant term, multiple single terms, and multiple interaction terms;

[0133] The fourth component unit is used to obtain multiple environmental impact items based on the impact items composed of a single target-related data and a preset third coefficient. Each third coefficient in the environmental impact items is different, and the index value of the target-related data in each environmental impact item is a set threshold.

[0134] The fifth component unit is used to obtain the environmental action item based on the action item composed of multiple target-related data and the preset fourth coefficient. The index value of each target-related data in the environmental action item is a set threshold.

[0135] The sixth component is used to compose a second polynomial based on multiple environmental impact terms and environmental action terms;

[0136] The building unit is used to construct a predefined polynomial proxy model based on the first polynomial, the second polynomial, and the insulator icing thickness.

[0137] The first calculation unit is used to input the analyzed first and second information into a preset polynomial proxy model for calculation, and to calculate a preset loss function to obtain the loss function value.

[0138] The adjustment unit is used to adjust the values ​​of the coefficients in the preset polynomial surrogate model until the loss function value meets the set conditions to obtain the insulator model. The coefficients include the first coefficient, the second coefficient, the third coefficient, and the fourth coefficient.

[0139] In one specific embodiment disclosed in this application, the verification unit includes:

[0140] The second calculation unit is used to calculate the overall electric field strength of the insulator based on the target meteorological observation data, the initial insulator structure parameters and the target ice layer thickness.

[0141] An estimation unit is used to estimate the radius of the ice tip based on the target ice layer thickness.

[0142] The third calculation unit is used to calculate the ratio of the radius of the ice tip to the thickness of the target ice layer, and to perform an exponential operation on the ratio to obtain the first result. The value of the exponent is within a preset range.

[0143] The fourth calculation unit is used to calculate the thickness of the water film on the ice layer based on the target meteorological observation data and the target ice layer thickness;

[0144] The fifth calculation unit is used to calculate the ratio of the water film thickness to the target ice layer thickness, and to perform an exponential operation on the ratio to obtain the second result. The value of the exponent is within a preset range.

[0145] The sixth calculation unit is used to calculate the product of the first result, the second result, and the overall electric field strength to obtain the local electric field strength.

[0146] In one specific embodiment disclosed in this application, the verification unit includes:

[0147] The seventh calculation unit is used to calculate the product of the preset temperature coefficient and the temperature in the first information after analysis, and obtain the first product. The preset temperature coefficient is used to characterize the effect of temperature on the conductivity of the ice layer.

[0148] The eighth calculation unit is used to calculate the product of the preset humidity coefficient and the humidity in the first information after analysis, and obtain the second product. The preset humidity coefficient is used to characterize the influence of humidity on the conductivity of the ice layer.

[0149] The ninth calculation unit is used to calculate the product of the preset thickness coefficient and the target ice layer thickness to obtain the third product. The preset thickness coefficient is used to characterize the influence of the ice layer thickness on the conductivity of the ice layer.

[0150] The tenth calculation unit is used to calculate the product of the Boltzmann constant and the temperature in the first piece of information after analysis, to obtain the fourth product;

[0151] The eleventh calculation unit is used to calculate the ratio of the preset ice activation energy to the fourth product to obtain the target ratio. The preset ice activation energy is used to characterize the energy threshold at which the conductivity of the ice layer changes significantly.

[0152] The twelfth calculation unit is used to calculate the conductivity based on the preset ice basic conductivity, the first product, the second product, the third product, and the target ratio.

[0153] It should be noted that the specific manner in which each module performs its operation in the apparatus described in the above embodiments has been described in detail in the embodiments of the method, and will not be elaborated here.

[0154] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

[0155] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for optimizing insulator structure based on polynomial vehicles, characterized in that, include: Acquire first information and second information, wherein the first information includes meteorological observation data of the insulator working area and the second information includes insulator structural parameters and insulator ice thickness; A correlation analysis was performed on the insulator icing thickness in the first information and the second information to obtain the analyzed first information, which includes meteorological observation data that has a significant impact on the insulator icing thickness. Based on the first and second information obtained from the analysis, a preset multinomial surrogate model is trained. When the preset loss function meets the set conditions, the training is stopped, and an insulator model is obtained. The preset target meteorological observation data is input into the insulator model and the optimization algorithm is used to optimize the ice thickness of the insulator, so as to obtain the optimized initial insulator structure parameters and the corresponding target ice layer thickness. The initial insulator structure parameters and the target icing layer thickness are verified based on a preset physical model to obtain verification results. When the verification results meet preset conditions, the initial insulator structure parameters are used as the target insulator structure parameters. Correlation analysis is performed on the insulator icing thickness based on the first and second information to obtain the analyzed first information. The meteorological observation data includes multiple meteorological related data, including: The multiple meteorological data contained in the first information are divided using a predetermined division strategy to obtain multiple division combinations. The division strategy ensures that the combination of meteorological data in each division combination is unique, and each meteorological data can appear in multiple division combinations. Each partition combination is fed into a preset feature transformation model for feature transformation, resulting in multiple first feature vectors; The ice thickness of the insulator in each of the second pieces of information is input into a preset feature extraction model for feature extraction to obtain multiple second feature vectors; Correlation analysis is performed based on the first feature vector of each partition combination and the second feature vector corresponding to each second piece of information to obtain multiple correlation values; The maximum correlation value is determined from multiple correlation values, and all meteorological related data in the partition combination corresponding to the maximum correlation value are used as target related data. Meteorological data, excluding target-related data, are deleted from the first information to obtain the analyzed first information.

2. The insulator structure optimization method based on polynomial vehicle according to claim 1, characterized in that... Based on the analyzed first and second information, a preset polynomial surrogate model is trained. Training stops when the preset loss function meets set conditions, resulting in an insulator model. The preset polynomial surrogate model consists of a first polynomial and a second polynomial. The first polynomial is composed of insulator structural parameters, and the second polynomial is composed of target-related data, including: Based on a single term composed of a single insulator structural parameter and a preset first coefficient, multiple terms are obtained, wherein each first coefficient in a term is different, and the exponent value of the insulator structural parameter in each term is different. Multiple interaction terms are obtained based on multiple insulator structure parameters and preset second coefficients. In each pair of interaction terms, at least one insulator structure parameter is different. The exponent value of each insulator structure parameter in each interaction term is a set threshold. The first polynomial is composed of a preset constant term, multiple single terms, and multiple interaction terms; Multiple environmental impact items are obtained based on the impact items composed of single target-related data and preset third coefficients. Each third coefficient in the environmental impact items is different, and the index value of the target-related data in each environmental impact item is a set threshold. An environmental action term is obtained based on multiple target-related data and a preset fourth coefficient. The index value of each target-related data in the environmental action term is a set threshold. The second polynomial is composed of multiple environmental impact terms and the environmental action terms; The preset polynomial proxy model is constructed based on the first polynomial, the second polynomial, and the ice thickness of the insulator; The analyzed first and second information are input into the preset polynomial proxy model for calculation, and the preset loss function is calculated to obtain the loss function value; Adjust the values ​​of the coefficients in the preset polynomial surrogate model until the loss function value meets the set conditions to obtain the insulator model. The coefficients include the first coefficient, the second coefficient, the third coefficient, and the fourth coefficient.

3. The insulator structure optimization method based on polynomial vehicle according to claim 1, characterized in that... The initial insulator structure parameters are verified based on a preset physical model, which includes an electric field distribution model, a heat conduction model, and a mechanical stress model. The electric field distribution model consists at least of the local electric field strength and the conductivity of the icing layer. The local electric field strength includes: Based on the target meteorological observation data, the initial insulator structural parameters, and the target ice layer thickness, the overall electric field strength of the insulator is calculated. The radius of the ice tip is estimated based on the target ice layer thickness. Calculate the ratio of the radius of the ice tip to the thickness of the target ice layer, and perform an exponential operation on the ratio to obtain a first result, wherein the value of the exponent is within a preset range; The thickness of the water film on the ice layer is calculated based on the meteorological observation data of the target and the thickness of the target ice layer. Calculate the ratio of the water film thickness to the target ice layer thickness, and perform an exponential operation on the ratio to obtain a second result. The value of the exponent is within a preset range. The local electric field strength is obtained by multiplying the first result, the second result, and the overall electric field strength.

4. The insulator structure optimization method based on polynomial vehicle according to claim 3, characterized in that... The initial insulator structure parameters are verified based on a preset physical model, which includes an electric field distribution model, a heat conduction model, and a mechanical stress model. The electric field distribution model consists at least of the conductivity of the icing layer, which includes: The product of the preset temperature coefficient and the temperature in the first information after analysis is calculated to obtain the first product. The preset temperature coefficient is used to characterize the effect of temperature on the conductivity of the ice layer. The product of the preset humidity coefficient and the humidity in the first information after analysis is calculated to obtain the second product. The preset humidity coefficient is used to characterize the effect of humidity on the conductivity of the ice layer. The product of the preset thickness coefficient and the target ice layer thickness is calculated to obtain a third product. The preset thickness coefficient is used to characterize the effect of ice layer thickness on the conductivity of the ice layer. The fourth product is obtained by multiplying the Boltzmann constant by the temperature from the first piece of information after analysis; Calculate the ratio of the preset ice activation energy to the fourth product to obtain the target ratio. The preset ice activation energy is used to characterize the energy threshold at which the conductivity of the ice layer changes significantly. The conductivity is calculated based on the preset basic ice conductivity, the first product, the second product, the third product, and the target ratio.

5. An insulator structure optimization device based on a polynomial vehicle, characterized in that, include: The acquisition unit is used to acquire first information and second information. The first information includes meteorological observation data of the working area of ​​the insulator, and the second information includes insulator structural parameters and insulator ice thickness. An analysis unit is used to perform correlation analysis on the insulator icing thickness in the first information and the second information to obtain the analyzed first information, which includes meteorological observation data that has a significant impact on the insulator icing thickness. The training unit is used to train a preset polynomial surrogate model based on the analyzed first information and the second information. When the preset loss function meets the set conditions, the training stops and the insulator model is obtained. The optimization unit is used to input the preset target meteorological observation data into the insulator model and use the optimization algorithm to optimize the ice thickness of the insulator, so as to obtain the optimized initial insulator structure parameters and the corresponding target ice layer thickness. The verification unit is used to verify the initial insulator structure parameters and the target icing layer thickness based on a preset physical model to obtain the verification result. When the verification result meets the preset conditions, the initial insulator structure parameters are used as the target insulator structure parameters. The analysis unit includes: The first partitioning unit is used to partition the multiple meteorological data contained in the first information using a predetermined partitioning strategy to obtain multiple partitioning combinations. The partitioning strategy ensures that the combination of meteorological data in each partitioning combination is unique, and each meteorological data can appear in multiple partitioning combinations. The transformation unit is used to input each partition combination into a preset feature transformation model for feature transformation to obtain multiple first feature vectors; An extraction unit is used to input the insulator ice thickness in each of the second information into a preset feature extraction model for feature extraction, and obtain multiple second feature vectors. The correlation unit is used to perform correlation analysis based on the first feature vector of each partition combination and the second feature vector corresponding to each second information, and to obtain multiple correlation values. The determination unit is used to determine the maximum correlation value from multiple correlation values, and to take all meteorological correlation data in the partition combination corresponding to the maximum correlation value as the target correlation data; The deletion unit is used to delete meteorological data other than target-related data from the first information to obtain the analyzed first information.

6. The insulator structure optimization device based on polynomial vehicle according to claim 5, characterized in that, The preset polynomial proxy model consists of a first polynomial and a second polynomial. The first polynomial is composed of insulator structure parameters, and the second polynomial is composed of target-related data. The training unit includes: The first component unit is used to obtain multiple individual terms based on a single term composed of a single insulator structural parameter and a preset first coefficient. Each first coefficient in the individual term is different, and the exponent value of the insulator structural parameter in each individual term is different. The second component unit is used to obtain multiple interaction terms based on multiple insulator structure parameters and preset second coefficients. In each pair of interaction terms, at least one insulator structure parameter is different. The exponent value of each insulator structure parameter in each interaction term is a set threshold. The third component unit is used to form the first polynomial based on a preset constant term, multiple single terms, and multiple interaction terms; The fourth component unit is used to obtain multiple environmental impact items based on the impact items composed of a single target-related data and a preset third coefficient. Each third coefficient in the environmental impact items is different, and the index value of the target-related data in each environmental impact item is a set threshold. The fifth component unit is used to obtain an environmental action item based on multiple target-related data and a preset fourth coefficient. The index value of each target-related data in the environmental action item is a set threshold. The sixth component is used to compose the second polynomial based on multiple environmental impact terms and the environmental action terms; A construction unit is used to construct the preset polynomial proxy model based on the first polynomial, the second polynomial, and the ice thickness of the insulator; The first calculation unit is used to input the analyzed first information and second information into the preset polynomial proxy model for calculation, and calculate the preset loss function to obtain the loss function value; An adjustment unit is used to adjust the values ​​of the coefficients in the preset polynomial surrogate model until the loss function value meets the set conditions to obtain the insulator model. The coefficients include the first coefficient, the second coefficient, the third coefficient, and the fourth coefficient.

7. The insulator structure optimization device based on polynomial vehicle according to claim 5, characterized in that, The physical model includes an electric field distribution model, a heat conduction model, and a mechanical stress model. The electric field distribution model consists at least of the local electric field strength and the conductivity of the ice layer. The verification unit includes: The second calculation unit is used to calculate the overall electric field strength of the insulator based on the target meteorological observation data, the initial insulator structure parameters, and the target ice layer thickness. An estimation unit is used to estimate the radius of the ice tip based on the target ice layer thickness; The third calculation unit is used to calculate the ratio of the radius of the ice tip to the thickness of the target ice layer, and to perform an exponential operation on the ratio to obtain a first result, wherein the value of the exponent is within a preset range. The fourth calculation unit is used to calculate the thickness of the water film on the ice layer based on the target meteorological observation data and the target ice layer thickness; The fifth calculation unit is used to calculate the ratio of the water film thickness to the target ice layer thickness, and perform an exponential operation on the ratio to obtain a second result, wherein the value of the exponent is within a preset range; The sixth calculation unit is used to calculate the product of the first result, the second result, and the overall electric field strength to obtain the local electric field strength.

8. The insulator structure optimization device based on polynomial vehicle according to claim 7, characterized in that, The verification unit includes: The seventh calculation unit is used to calculate the product of the preset temperature coefficient and the temperature in the first information after analysis, to obtain the first product. The preset temperature coefficient is used to characterize the effect of temperature on the conductivity of the ice layer. The eighth calculation unit is used to calculate the product of the preset humidity coefficient and the humidity in the first information after analysis to obtain the second product. The preset humidity coefficient is used to characterize the influence of humidity on the conductivity of the ice layer. The ninth calculation unit is used to calculate the product of the preset thickness coefficient and the target ice layer thickness to obtain the third product. The preset thickness coefficient is used to characterize the influence of the ice layer thickness on the conductivity of the ice layer. The tenth calculation unit is used to calculate the product of the Boltzmann constant and the temperature in the first piece of information after analysis, to obtain the fourth product; The eleventh calculation unit is used to calculate the ratio of the activation energy of the preset ice to the fourth product to obtain the target ratio. The activation energy of the preset ice is used to characterize the energy threshold at which the conductivity of the ice layer changes significantly. The twelfth calculation unit is used to calculate the conductivity based on the preset ice basic conductivity, the first product, the second product, the third product, and the target ratio.

Images