A method for constructing a melting degree prediction model for high-temperature molten glassy end products

Abstract

The present invention discloses a method for constructing a melting degree prediction model of a high-temperature molten glassy terminal product, and relates to the field of melting degree prediction technology. The method for constructing a melting degree prediction model of a high-temperature molten glassy terminal product comprises the following steps: obtaining characteristic data; obtaining a related index; and obtaining a melting degree prediction index. The present invention obtains melting degree related data based on a pre-treated terminal product sample, then obtains characteristic data based on the melting degree related data, and then obtains a first related index, a second related index, and a third related index based on the characteristic data, and inputs the first related index, the second related index, and the third related index into a constructed melting degree prediction index model to obtain a melting degree prediction index, thereby achieving the effect of improving the accuracy of the prediction of the melting degree of the glassy terminal product, and solving the problem of low accuracy of the prediction of the melting degree of the glassy terminal product in the prior art.

Description

Technical Field

[0001] The invention relates to the technical field of melting degree prediction, and in particular to a method for constructing a melting degree prediction model of a high-temperature molten glassy terminal product. Background Art

[0002] With the rapid development of materials science and its increasing status in the development of the national economy, materials science has been increasingly valued and supported by various countries. The development and application of materials are ubiquitous in the national economy and people's lives. Whether it is the launch of high-tech products or the utensils in daily life, they are inseparable from various materials. The continuous progress of science and technology and the continuous improvement of human's living standards have made human's requirements for materials higher and higher. Glass, as an important inorganic non-metallic material, has excellent properties such as transparency, high hardness, and good chemical stability. It has a wide range of applications in optics, construction, electronics, medical and other fields. As an important link in the glass manufacturing process, the prediction of the melting degree of high-temperature molten glass-like terminal products is of great significance for controlling glass quality and optimizing production processes. Melting degree is an important indicator for evaluating the molten state of glass, which directly affects the quality and performance of glass products. By predicting the melting degree, the production process parameters can be adjusted in time to ensure that the glass products meet the predetermined quality requirements. The prediction of melting degree helps to optimize the process parameters such as heating temperature and time in the glass production process, improve production efficiency, and reduce energy consumption and costs. When developing new glass materials, the prediction of melting degree is an important basis for evaluating material properties and determining the preparation process. By predicting the melting point, the research and development and preparation process of new materials can be guided.

[0003] The existing method is to experimentally determine the melting temperature range of glass under different conditions to obtain experimental data of melting point, and then use the thermodynamic properties, molecular structure and other parameters of the glass to perform theoretical calculations to predict the melting point of the glass.

[0004] For example, the invention patent with announcement number: CN112881455B discloses a method for predicting the melting temperature of coal ash based on a composite model of mineral phase and neural network, comprising: firstly establishing a coal ash mineral phase composition sub-model, using the change in Gibbs free energy produced by the mutual reaction of chemical components at high temperature to establish a linear programming problem, establishing a prediction model for solving the coal ash mineral phase composition at a specified temperature, and testing the consistency of this model; on the basis of the mineral phase composition sub-model, establishing an ash melting point prediction sub-model; establishing a neural network model, adjusting various training parameters of the neural network, and using an iterative algorithm to further enhance the prediction accuracy of the prediction model, while adding a correction value to represent the influence of minor elements in the coal ash on the fusibility of the coal ash; finally, analyzing the accuracy and reliability of the established model, establishing an accuracy index for the prediction result, and comparing it with the prediction result of the support vector machine prediction method.

[0005] For example, the molten metal composition inference device, molten metal composition inference method and molten metal manufacturing method disclosed in the invention patent with publication number CN112154218B include: an input device, into which measurement information about a refining device is input, the measurement information includes measurement results of optical characteristics of a furnace mouth portion in a blowing process of the refining device; a model database, storing model formulas and model parameters about blowing process reactions, the model formulas and model parameters including model formulas and model parameters representing the relationship between the decarburization oxygen efficiency of the refining device and the carbon concentration in the molten metal; a model calculation unit, using the measurement information, the model formulas and the model parameters to infer the component concentration of the molten metal including the carbon concentration in the molten metal; and a model determination unit, inferring the carbon concentration in the molten metal based on the measurement results, determining the model formula and model parameters used by the model calculation unit based on the inference results, the measurement results including the intensity change rate of the spectrum of the reduction reaction of iron oxide from the slag, the model determination unit detects the timing when the intensity change rate changes from a positive value to a negative value as the timing when a critical carbon concentration is reached, and switches the model formula and model formula parameters based on the critical carbon concentration.

[0006] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, the present application found that the above technology has at least the following technical problems:

[0007] In the prior art, since tests can usually only be performed on specific samples and conditions, it is difficult to comprehensively cover all possible glass types and melting conditions, resulting in low accuracy of test results and low accuracy in predicting the melting point of glassy terminal products. Summary of the invention

[0008] The embodiment of the present application solves the problem of low accuracy in predicting the melting degree of glassy terminal products in the prior art by providing a method for constructing a melting degree prediction model for a high-temperature molten glassy terminal product, thereby improving the accuracy of predicting the melting degree of glassy terminal products.

[0009] The embodiment of the present application provides a method for constructing a melting degree prediction model for a high-temperature molten glass-like terminal product, comprising the following steps: S1, obtaining melting degree related data based on an acquired pre-treated terminal product sample, and performing data preprocessing and feature extraction on the melting degree related data to obtain feature data, wherein the pre-treated terminal product sample is a sample that has undergone a melting treatment; S2, obtaining a first correlation index, a second correlation index, and a third correlation index based on the feature data, wherein the first correlation index is used to evaluate the change in phase change rate during the melting treatment process, the second correlation index is used to evaluate the change in flow rate during the melting treatment process, and the third correlation index is used to evaluate the change in melting efficiency during the melting treatment process; S3, inputting the first correlation index, the second correlation index, and the third correlation index into the constructed melting degree prediction index model to obtain a melting degree prediction index, wherein the melting degree prediction index is used to quantify the melting degree of the high-temperature molten glass-like terminal product.

[0010] Furthermore, the specific process of obtaining the pre-treated terminal product sample is as follows: a high-temperature molten sample is obtained according to preset information, and the high-temperature molten sample is pre-treated to obtain a pre-treated sample, wherein the preset information includes sample hardness, raw material type and chemical composition of the raw material, and the pre-treatment represents crushing, screening and mixing of the high-temperature molten sample, and melting the pre-treated sample to obtain a pre-treated terminal product sample.

[0011] Furthermore, the specific steps of the data preprocessing are as follows: performing missing value processing on the melting-related data to obtain preprocessed first data, and the missing value processing means filling the missing values ​​in the melting-related data with the median; performing standardization processing on the preprocessed first data to obtain preprocessed data, and the standardization processing is used to eliminate different dimensions in the preprocessed first data to ensure that the format of the preprocessed data is consistent.

[0012] Furthermore, the specific steps of the feature extraction are as follows: performing melting point feature extraction on the preprocessed data according to the preset melting point feature to obtain the first feature data, the first feature data represents the preprocessed data after the melting point feature extraction; performing melting heat feature extraction on the first feature data according to the preset melting heat feature to obtain the second feature data, the second feature data represents the first feature data after the melting heat feature extraction; performing phase change feature extraction on the second feature data according to the preset phase change feature to obtain the third feature data; the feature data includes the first feature data, the second feature data and the third feature data.

[0013] Furthermore, the specific acquisition process of the first correlation index is as follows: measuring the first experimental data by a first measuring device at a preset time point, obtaining the first free energy of the pretreated sample by a contact angle measurement method, the first measuring device includes a transmission electron microscope, a melting point meter and a differential scanning calorimeter, the first experimental data includes a first radius, a zero-dimensional particle melting temperature and a first melting enthalpy, obtaining a preset first melting temperature threshold, a preset second melting temperature threshold, a preset first weight and a preset second weight from a database, and obtaining the first correlation index in combination with the first melting enthalpy, the first radius, the zero-dimensional particle melting temperature and the first free energy.

[0014] Furthermore, the specific acquisition process of the second related index is as follows: at a preset time point, the density of the pretreated sample is measured by a density meter, and the first molar volume of the pretreated sample is obtained accordingly; the reference constant, the property constant and the preset melting temperature threshold are obtained from the database, and the second related index is obtained in combination with the first molar volume; the specific acquisition process of the third related index is as follows: at a preset time point, the third experimental data is measured by a third measuring device, the third measuring device includes an electronic balance and a differential scanning calorimeter, and the third experimental data includes the mass, latent heat of melting and first heat of the pretreated sample; the preset minimum mass threshold, the preset minimum heat threshold and the preset maximum total heat threshold are obtained from the database, and the third related index is obtained in combination with the mass, latent heat of melting and the first heat.

[0015] Furthermore, the first correlation index is calculated using the following formula:

[0016]

[0017] In the formula, represents the first correlation index of the mth pre-treated sample, m = 1, 2, ..., n, m represents the number of the pre-treated sample, n represents the total number of pre-treated samples, γ im represents the first free energy of the mth pretreated sample at the ith preset time point, i = 1, 2, ..., r, i represents the number of the preset time point, r represents the total number of preset time points, H im represents the first melting enthalpy of the mth pretreated sample at the i-th preset time point, r im represents the first radius of the mth pre-treated sample at the ith preset time point, T im represents the zero-dimensional particle melting temperature of the mth pretreated sample at the i-th preset time point, T0 represents the preset first melting temperature threshold, T2 represents the preset second melting temperature threshold, σ1 represents the preset first weight, represents the preset second weight, and e represents a natural constant.

[0018] Furthermore, the second correlation index is calculated using the following formula:

[0019]

[0020] In the formula, represents the second correlation index of the mth pre-treated sample, m = 1, 2, ..., n, m represents the number of the pre-treated sample, n represents the total number of pre-treated samples, T 01 Represents the reference constant, V i represents the first molar volume of the mth pretreated sample at the i-th preset time point, i=1,2,...,r, i represents the number of the preset time point, r represents the total number of preset time points, A represents the material property constant, T1 represents the preset melting temperature threshold, and e represents the natural constant.

[0021] Furthermore, the third correlation index is calculated using the following formula:

[0022]

[0023] In the formula, represents the third correlation index of the mth pre-treated sample, m = 1, 2, ..., n, m represents the number of the pre-treated sample, n represents the total number of pre-treated samples, M im represents the mass of the mth pre-treated sample at the ith preset time point, i = 1, 2, ..., r, i represents the number of the preset time point, r represents the total number of preset time points, L im represents the melting latent heat of the mth pretreated sample at the i-th preset time point, Q im represents the first heat of the mth pretreated sample at the i-th preset time point, M0 represents the preset minimum mass threshold, L0 represents the preset minimum heat threshold, Q0 represents the preset maximum total heat threshold, and e represents a natural constant.

[0024] Furthermore, the specific acquisition process of the melting prediction index is as follows: obtain the preset allocation weights from the melting database, the preset allocation weights include a preset first allocation weight, a preset second allocation weight and a preset third allocation weight, the preset first allocation weight is used to evaluate the influence of the first correlation index on the melting prediction index model, the preset second allocation weight is used to evaluate the influence of the second correlation index on the melting prediction index model, and the preset third allocation weight is used to evaluate the influence of the third correlation index on the melting prediction index model. Input the preset allocation weights, the first correlation index, the second correlation index and the third correlation index into the constructed melting prediction index model to obtain the melting prediction index.

[0025] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0026] 1. The characteristic data is obtained by preprocessing the melting point related data and extracting the characteristics, and then the first correlation index, the second correlation index and the third correlation index are obtained according to the characteristic data. Finally, the first correlation index, the second correlation index and the third correlation index are input into the constructed melting point prediction index model to obtain the melting point prediction index, thereby improving the reliability of the prediction process, and then achieving the effect of improving the accuracy of the prediction of the melting point of the glassy terminal product, and effectively solving the problem of low accuracy of the prediction of the melting point of the glassy terminal product in the prior art.

[0027] 2. Preprocess first data by performing missing value processing on melting point related data, and then perform standardization processing on the preprocessed first data to obtain preprocessed data, and then perform melting point feature extraction on the preprocessed data according to preset melting point features to obtain first feature data, and then perform melting heat feature extraction on the first feature data according to preset melting heat features to obtain second feature data, and finally perform phase change feature extraction on the second feature data according to preset phase change features to obtain third feature data, thereby reducing the complexity of the data and improving the reliability of the acquired data.

[0028] 3. By measuring the first experimental data at a preset time point, and then obtaining the first free energy of the pretreated sample through the contact angle measurement method, and finally obtaining the preset first melting temperature threshold, the preset second melting temperature threshold, the preset first weight and the preset second weight from the database, and combining the first melting enthalpy, the first radius, the zero-dimensional particle melting temperature and the first free energy to obtain the first correlation index, thereby realizing the numerical evaluation of the phase change rate change during the melting process, and then realizing the improvement of the reliability of model prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 A flow chart of a method for constructing a melting degree prediction model for a high-temperature molten glassy end product provided in an embodiment of the present application;

[0030] Figure 2 A statistical graph of changes in the third correlation index provided in an embodiment of the present application. DETAILED DESCRIPTION

[0031] The embodiment of the present application solves the problem of low accuracy in the prediction of the melting degree of glass-like terminal products in the prior art by providing a method for constructing a melting degree prediction model for a high-temperature molten glass-like terminal product. Melting degree related data is obtained by obtaining a pretreated terminal product sample, and the melting degree related data is preprocessed and feature extracted to obtain feature data. Then, a first correlation index, a second correlation index and a third correlation index are obtained based on the feature data. Finally, the first correlation index, the second correlation index and the third correlation index are input into the constructed melting degree prediction index model to obtain the melting degree prediction index, thereby achieving the effect of improving the accuracy of the prediction of the melting degree of the glass-like terminal product.

[0032] The technical solution in the embodiment of the present application is to solve the problem that the prediction accuracy of the melting degree of the glassy terminal product is not high. The overall idea is as follows:

[0033] By obtaining characteristic data based on the pretreated terminal product sample, obtaining the first correlation index, the second correlation index and the third correlation index based on the characteristic data and inputting the first correlation index, the second correlation index and the third correlation index into the constructed melting degree prediction index model to obtain the melting degree prediction index, the effect of improving the accuracy of the prediction of the melting degree of the glassy terminal product is achieved.

[0034] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0035] like Figure 1 As shown, it is a flow chart of a method for constructing a melting degree prediction model of a high-temperature molten glass-like terminal product provided by an embodiment of the present application, the method comprising the following steps: S1, obtaining melting degree related data according to the obtained pre-treated terminal product sample, and performing data pre-processing and feature extraction on the melting degree related data to obtain feature data, the melting degree related data including the melting temperature, melting point, melting viscosity, melting enthalpy and melting entropy corresponding to the pre-treated sample when the pre-treated sample is subjected to melting treatment, the feature extraction is used to extract key features and remove redundant information, the data pre-processing is used to ensure the uniform format of the pre-treated data, and the pre-treated terminal product sample is a sample subjected to melting treatment; S2, obtaining a first correlation index, a second correlation index and a third correlation index according to the feature data, the first correlation index is used to evaluate the change of phase change rate during the melting treatment process, the second correlation index is used to evaluate the change of flow rate during the melting treatment process, and the third correlation index is used to evaluate the change of melting efficiency during the melting treatment process; S3, inputting the first correlation index, the second correlation index and the third correlation index into the constructed melting degree prediction index model to obtain a melting degree prediction index, and the melting degree prediction index is used to quantify the melting degree of the high-temperature molten glass-like terminal product.

[0036] In this embodiment, data acquisition and preprocessing are performed to ensure that the acquired preprocessed terminal product sample data is accurate and complete, and the data is cleaned and standardized to eliminate inconsistencies and outliers, remove redundant information, reduce data complexity, improve the generalization ability of the model, ensure that all preprocessed data are in a unified format, facilitate subsequent processing and analysis, and achieve improved accuracy in predicting the melting point of the glassy terminal product.

[0037] Furthermore, the specific process of obtaining the pretreatment terminal product sample is as follows: a high-temperature molten sample is obtained according to preset information, and the high-temperature molten sample is pretreated to obtain a pretreated sample, the preset information includes sample hardness, raw material type and chemical composition of the raw material, and the pretreatment means crushing, screening and mixing the high-temperature molten sample, crushing is used to ensure that the high-temperature molten sample meets the preset particle size range, screening is used to separate high-temperature molten samples of different particle sizes, and mixing is used to ensure the uniformity of the high-temperature molten sample. The pretreated sample means a high-temperature molten sample that has been pretreated; the pretreated sample is melt-treated to obtain a pretreatment terminal product sample, and the melting treatment is used to melt the pretreated sample to form a glassy state.

[0038] In this embodiment, the preset information is obtained from the melting database, the hardness of the sample will affect the selection of the raw materials and the melting conditions, the type of raw materials and the chemical composition of the raw materials can ensure the consistency and repeatability of the raw materials, the preset particle size range is obtained from the melting database, and the preset particle size range is generally 0-0.5 mm. The pre-treated terminal product sample is observed under a microscope to ensure the uniformity of the pre-treated terminal product sample, thereby improving the simplicity of subsequent processing; the accuracy of the prediction of the melting degree of the glassy terminal product is improved.

[0039] Furthermore, the specific steps of data preprocessing are as follows: performing missing value processing on the melting-related data to obtain preprocessed first data, the missing value processing means filling the missing values ​​in the melting-related data with the median, and the preprocessed first data means the melting-related data after the missing value processing; performing standardization processing on the preprocessed first data to obtain preprocessed data, the standardization processing is used to eliminate different dimensions in the preprocessed first data to ensure the consistency of the format of the preprocessed data, and the preprocessed data means the melting-related data after data preprocessing.

[0040] In this embodiment, before processing the data, the original data is backed up so that it can be restored to the original state when needed. Before missing value processing and standardization, the data is thoroughly cleaned to remove and correct errors. The median is insensitive to outliers and is suitable for processing outliers in the data set. The Z-score (z score) standardization method is used for standardization. The Z-score standardization converts the preprocessed first data into a standard normal distribution, in which the mean of the preprocessed first data is 0 and the standard deviation is 1, which can ensure that the data standardization is both accurate and effective, and provide a reliable data basis for subsequent analysis and application; the accuracy of the prediction of the melting point of the glassy terminal product is improved.

[0041] Furthermore, the specific steps of feature extraction are as follows: performing melting point feature extraction on the preprocessed data according to a preset melting point feature to obtain first feature data, the first feature data represents the preprocessed data after the melting point feature extraction, and the melting point feature extraction is used to obtain the melting point information in the preprocessed data; performing melting heat feature extraction on the first feature data according to a preset melting heat feature to obtain second feature data, the second feature data represents the first feature data after the melting heat feature extraction, and the melting heat feature extraction is used to obtain the melting heat information in the first feature data; performing phase change feature extraction on the second feature data according to a preset phase change feature to obtain third feature data, and the phase change feature extraction is used to obtain the phase change information in the second feature data; the feature data includes the first feature data, the second feature data and the third feature data.

[0042] In this embodiment, the preset melting point feature is obtained from the melting database, the preset melting heat feature is obtained from the melting database, and the preset phase change feature is obtained from the melting database. The melting point information includes the melting point temperature and the melting point range, the melting heat information includes the melting enthalpy and the melting heat rate, and the phase change information includes the phase change type, the phase change temperature and the phase change enthalpy. The melting point information, the melting heat information and the phase change information are crucial for understanding the processing performance, thermal stability and application potential of the glassy terminal product. The corresponding characteristic values ​​are extracted according to the melting point temperature and the melting point range in the historical time period, the corresponding characteristic values ​​are extracted according to the melting enthalpy and the melting heat rate in the historical time period, and the corresponding characteristic values ​​are extracted according to the phase change type, the phase change temperature and the phase change enthalpy in the historical time period; the accuracy of the prediction of the melting degree of the glassy terminal product is improved.

[0043] Furthermore, the specific acquisition process of the first correlation index is as follows: at a preset time point, the first experimental data is measured by a first measuring device, and the first free energy of the pretreated sample is obtained by a contact angle measurement method. The first measuring device includes a transmission electron microscope, a melting point meter and a differential scanning calorimeter. The first experimental data includes a first radius, a zero-dimensional particle melting temperature and a first melting enthalpy. The first radius represents the radius of a tiny crystal formed by molecular aggregation of the pretreated sample during the melting process. The first melting enthalpy represents the energy required for the pretreated sample to change from a solid state to a liquid state. The first free energy represents the melting temperature of the solid during the melting process. The free energy of the interface between the solid state and the liquid state, the zero-dimensional particle melting temperature represents the melting temperature of the zero-dimensional particles in the pretreated sample affected by the size effect; the preset first melting temperature threshold, the preset second melting temperature threshold, the preset first weight and the preset second weight are obtained from the database, and the first melting enthalpy, the first radius, the zero-dimensional particle melting temperature and the first free energy are combined to obtain the first correlation index, the preset first weight is used to evaluate the influence of the zero-dimensional particle melting temperature on the first correlation index, and the preset second weight is used to evaluate the influence of the first melting enthalpy, the first radius and the first free energy on the first correlation index.

[0044] In this embodiment, the preset first melting temperature threshold is represented by the mean of the first melting temperature data in the calculated database, and the preset second melting temperature threshold is represented by the mean of the second melting temperature data in the calculated database. The transmission electron microscope is a high-resolution, high-magnification microscope that uses an electron beam as a light source and focuses and images the electron beam through an electromagnetic field. The wavelength of the electron beam is much shorter than that of visible light and ultraviolet light, so it has a higher resolution and can observe fine structures less than 0.2 microns, that is, submicrostructures or ultrastructures. The differential scanning calorimeter is a thermal analysis instrument based on the measurement principle of heat flow differences. Under program-controlled temperature, the relationship between the power difference (such as in the form of heat) input to the sample and the reference and the temperature is measured. The sample and the reference are heated or cooled at the same time, and the temperature rise and fall rates are the same. When the sample undergoes a phase change or thermal reaction, it absorbs or releases heat, resulting in a temperature difference between the sample and the reference. This difference is detected by a thermocouple or a thermistor detector and converted into a heat signal, so that the thermodynamic properties of the sample, including the melting process, can be analyzed. The basic working principle of the melting point instrument is to determine the melting point by heating the pretreated sample and monitoring the thermal effect during its temperature change. During the heating process, when the pretreated sample changes from solid to liquid, it absorbs a large amount of heat, causing the temperature rise rate to slow down or a platform to appear. The melting point instrument determines the melting temperature of zero-dimensional particles by detecting the characteristic points of this temperature change. The contact angle is the angle between the tangent line of the droplet surface and the horizontal line of the solid surface at the gas-liquid-solid three-phase junction when the liquid is dropped on the solid surface. This angle is directly related to the free energy (or surface tension) of the solid surface, because the free energy determines the strength of the interaction between the liquid and the solid surface. By linking the contact angle with the surface tension of the liquid, the free energy of the pretreated sample can be calculated; the accuracy of the prediction of the melting degree of the glassy terminal product is improved.

[0045] Specifically, the preset first weight is obtained from a melting database; in a specific embodiment, a fitting curve is obtained by fitting the relationship between the preset melting temperature and the first correlation index to the preset melting temperature and the corresponding first correlation index within a historical period, and the preset melting temperature is input according to the fitting curve to obtain the preset first weight corresponding to the preset melting temperature.

[0046] Specifically, the preset second weight is obtained from a melting database; in a specific embodiment, a fitting curve is obtained by fitting the relationship between the first melting enthalpy, the first radius, the first free energy and the corresponding first correlation index within a historical period, and the first melting enthalpy, the first radius, the first free energy and the first correlation index are input according to the fitting curve to obtain the preset second weight corresponding to the first melting enthalpy, the first radius and the first free energy.

[0047] Furthermore, the specific acquisition process of the second related index is as follows: at a preset time point, the density of the pretreated sample is measured by a density meter, and the first molar volume of the pretreated sample is obtained, and the first molar volume represents the molar volume of the pretreated sample during the melting process; the reference constant, the property constant and the preset melting temperature threshold are obtained from the database, and the second related index is obtained in combination with the first molar volume, the reference constant is used for baseline adjustment, and the property constant is used to reflect the influence of the internal structure and interaction of the pretreated sample on the melting temperature; the specific acquisition process of the third related index is as follows: at a preset time point, the third experimental data is measured by a third measuring device, the third measuring device includes an electronic balance and a differential scanning calorimeter, the third experimental data includes the mass, latent heat of melting and first heat of the pretreated sample, the latent heat of melting represents the heat required to melt a unit mass of the pretreated sample, and the first heat represents the total heat in the melting process; the preset minimum mass threshold, the preset minimum heat threshold and the preset maximum total heat threshold are obtained from the database, and the third related index is obtained in combination with the mass, latent heat of melting and the first heat.

[0048] In this embodiment, the reference constant is represented by the result of calculating the mean of the reference constant data in the database, the property constant is represented by the result of calculating the mean of the property constant data in the database, the preset melting temperature threshold is represented by the result of calculating the mean of the melting temperature data in the database, the preset mass threshold is represented by the result of calculating the mean of the mass data in the database, the preset heat threshold is represented by the result of calculating the mean of the heat data in the database, and the preset total heat threshold is represented by the result of calculating the mean of the total heat data in the database. The design of the density meter utilizes the floating and sinking conditions of the object, that is, the principle that the buoyancy of the object in the liquid is equal to the gravity. When the object is placed in the liquid, if the buoyancy is greater than the gravity, the object will float; if the buoyancy is less than the gravity, the object will sink; when the buoyancy is equal to the gravity, the object will remain in a floating state. The density of the pre-treated sample is obtained by measuring the floating state of the pre-treated sample in the liquid. The electronic balance has a high-precision measurement capability, which can be accurate to several decimal places, avoiding reading errors; and the accuracy of the prediction of the melting degree of the glassy terminal product is improved.

[0049] Furthermore, the first correlation index is calculated using the following formula:

[0050]

[0051] In the formula, represents the first correlation index of the mth pre-treated sample, m = 1, 2, ..., n, m represents the number of the pre-treated sample, n represents the total number of pre-treated samples, γ imrepresents the first free energy of the mth pretreated sample at the ith preset time point, i = 1, 2, ..., r, i represents the number of the preset time point, r represents the total number of preset time points, H im represents the first melting enthalpy of the mth pretreated sample at the i-th preset time point, r im represents the first radius of the mth pre-treated sample at the ith preset time point, T im represents the zero-dimensional particle melting temperature of the mth pretreated sample at the i-th preset time point, T0 represents the preset first melting temperature threshold, T2 represents the preset second melting temperature threshold, σ1 represents the preset first weight, represents the preset second weight, and e represents a natural constant.

[0052] In this embodiment, the algorithm of this embodiment combines the melting temperature T of the zero-dimensional particles im , first free energy γ im , first melting enthalpy H im and the first radius r im , the first correlation index of the mth pretreated sample is obtained by comprehensive analysis. The melting temperature of the zero-dimensional particle T in this formula im , first free energy γ im , first melting enthalpy H im and the first radius r im There is a common regulatory relationship between them. When the melting temperature of the zero-dimensional particle is T im and the first free energy γ im Gradually increases, the first melting enthalpy H im and the first radius r im Gradually decrease, when the preset first melting temperature threshold T0 and the preset second melting temperature threshold T2 are small enough, the first correlation index may decrease, which means that the phase change rate in the melting process slows down and higher energy is required for melting processing.

[0053] It needs to be explained that the melting temperature of zero-dimensional particles is T im The expression is: Among them, T im ′ represents the zero-dimensional particle melting temperature data of the mth pretreated sample at the i-th preset time point, T im 0 represents the reference zero-dimensional particle melting temperature data of the mth pretreated sample at the i-th preset time point, ΔT im represents the reference deviation of the melting temperature of the zero-dimensional particle, the first free energy γ im The expression is: Among them, γ im ′ represents the first free energy data of the mth pretreated sample at the ith preset time point, γ im 0represents the reference first free energy data of the mth pretreated sample at the i-th preset time point, Δγ im represents the first free energy reference deviation, the first melting enthalpy H im The expression is: Among them, H im ′ represents the first melting enthalpy data of the mth pretreated sample at the i-th preset time point, H im 0 represents the reference first melting enthalpy data of the mth pretreated sample at the i-th preset time point, ΔH im represents the first melting enthalpy reference deviation, the first radius r im The expression is: Among them, r im ′ represents the first radius data of the mth pre-treated sample at the i-th preset time point, r im 0 represents the reference first radius data of the mth pre-treated sample at the i-th preset time point, Δr im Indicates the first radius reference deviation.

[0054] It should be understood that the zero-dimensional particle melting temperature data of the mth pretreated sample at the ith preset time point is obtained by measurement, the reference zero-dimensional particle melting temperature data is represented by the result of calculating the mean of the zero-dimensional particle melting temperature data at the ith preset time point, the zero-dimensional particle melting temperature reference deviation is obtained by calculating the zero-dimensional particle melting temperature data and the reference zero-dimensional particle melting temperature data through the deviation formula, the first free energy data of the mth pretreated sample at the ith preset time point is obtained by measurement, the reference first free energy data is represented by the result of calculating the mean of the first free energy data at the ith preset time point, the first free energy reference deviation is obtained by calculating the first free energy data and the reference first free energy data through the deviation formula, and the mth pretreated sample at the ith preset time point is obtained by measurement. The first melting enthalpy data of the point is obtained by measurement, the reference first melting enthalpy data is represented by the result of calculating the mean of the first melting enthalpy data at the i-th preset time point, the first melting enthalpy reference deviation is obtained by calculating the first melting enthalpy data and the reference first melting enthalpy data through a deviation formula, the first radius data of the m-th pretreated sample at the i-th preset time point is obtained by measurement, the reference first radius data is represented by the result of calculating the mean of the first radius data at the i-th preset time point, the first radius reference deviation is obtained by calculating the first radius data and the reference first radius data through a deviation formula, the deviation formula is used to measure the difference between the reference value and the true measurement value, and is represented by the difference between the reference value and the true measurement value; the accuracy of the prediction of the melting degree of the glassy terminal product is improved.

[0055] Furthermore, the second correlation index is calculated using the following formula:

[0056]

[0057] In the formula, represents the second correlation index of the mth pre-treated sample, m = 1, 2, ..., n, m represents the number of the pre-treated sample, n represents the total number of pre-treated samples, T 01 Represents the reference constant, V i represents the first molar volume of the mth pretreated sample at the i-th preset time point, i=1,2,...,r, i represents the number of the preset time point, r represents the total number of preset time points, A represents the material property constant, T1 represents the preset melting temperature threshold, and e represents the natural constant.

[0058] In this embodiment, the algorithm of this embodiment combines the first molar volume V i , reference constant T 01 and material property constant A, the second correlation index of the mth pretreated sample is obtained by comprehensive analysis. The first molar volume V in this formula i , reference constant T 01 There is a common regulatory relationship between the material property constant A and the reference constant T 01 and the material property constant A increases, the first molar volume V i When the preset melting temperature threshold is small enough, The overall value gradually increases, but the second correlation index gradually decreases. It is negatively correlated with the second correlation index, but when the preset melting temperature threshold is large enough, The overall value gradually decreases, and the second correlation index gradually increases. It is positively correlated with the second correlation index, so it can be accurately analyzed by establishing a mathematical form.

[0059] Assume r = 1, the reference constant T 01 The range of is 1000-5000 (Kelvin), and the range of material property constant A is 0.01-0.1 (cm 3 / mol), the first molar volume V i The range is 1-10 (cm 3 / mol), the preset melting temperature threshold T1 is 1000 (Kelvin), then the statistical variation table of the second correlation index is shown in Table 1:

[0060] Table 1 Statistical changes of the second correlation index

[0061]

[0062]

[0063] It can be seen from the table that as the reference constant T 01 and the material property constant A increases, the first molar volume V i The second correlation index gradually decreases, and the preset melting temperature threshold is small enough. It is negatively correlated with the second correlation index. The negative sign at this time represents the degree of reduction of the second correlation index. The smaller the second correlation index, the greater the degree of reduction, which means that the flow rate changes slower during the melting process. This is not only due to the increase in the melting point, but may also be related to the phase change and molecular rearrangement of the material during the melting process. More energy absorption means that the melting process requires more heat input, and may require higher temperatures and longer times to complete the melting; the accuracy of the prediction of the melting degree of the glassy terminal product is improved.

[0064] Furthermore, the third correlation index is calculated using the following formula:

[0065]

[0066] In the formula, represents the third correlation index of the mth pre-treated sample, m = 1, 2, ..., n, m represents the number of the pre-treated sample, n represents the total number of pre-treated samples, M im represents the mass of the mth pre-treated sample at the ith preset time point, i = 1, 2, ..., r, i represents the number of the preset time point, r represents the total number of preset time points, L im represents the melting latent heat of the mth pretreated sample at the i-th preset time point, Q im represents the first heat of the mth pretreated sample at the i-th preset time point, M0 represents the preset minimum mass threshold, L0 represents the preset minimum heat threshold, Q0 represents the preset maximum total heat threshold, and e represents a natural constant.

[0067] In this embodiment, the algorithm of this embodiment combines the first heat Q im 、Quality M im and latent heat of fusion L im , comprehensive analysis to obtain the third correlation index of the mth pre-treated sample, the first heat Q in this formula im 、Quality M im and latent heat of fusion L im There is a common regulatory association between im and latent heat of fusion L im The increase of the first heat Q im The gradual decrease of Enough hours, The overall value gradually increases, and the third related index gradually increases. It is positively correlated with the third correlation index, but when When it is big enough, The overall value gradually decreases, and the third correlation index gradually decreases. It is negatively correlated with the third correlation index, so it can be accurately analyzed by establishing a mathematical form.

[0068] like Figure 2 As shown, it is a statistical diagram of the change of the third correlation index provided in the embodiment of the present application, wherein it is assumed that: the total number of preset time points r=200, and Q1=M im *L im , Q1 is the initial energy threshold, then the third correlation index of the mth pre-processed sample can be simplified as: The initial energy threshold Q1 ranges from 4000 to 5000 (J), and the first heat Q im The range is 1000-1200 (J), the preset mass threshold M0 is 0.5 (g), the preset heat threshold L0 is 100 (J / g), and the preset total heat threshold Q0 is 50 (J). It can be seen from the figure that as the initial energy threshold Q1 gradually increases, the first heat Q im The gradually decrease of the third correlation index of the mth pretreated sample gradually increases, which means that the melting efficiency changes faster during the melting process, which may mean that the pretreated sample has better thermal stability and mechanical properties; the accuracy of the prediction of the melting degree of the glassy terminal product is improved.

[0069] Furthermore, the specific process of obtaining the melting prediction index is as follows: obtain the preset allocation weights from the melting database, the preset allocation weights include a preset first allocation weight, a preset second allocation weight and a preset third allocation weight, the preset first allocation weight is used to evaluate the influence of the first relevant index on the melting prediction index model, the preset second allocation weight is used to evaluate the influence of the second relevant index on the melting prediction index model, and the preset third allocation weight is used to evaluate the influence of the third relevant index on the melting prediction index model; input the preset allocation weights, the first relevant index, the second relevant index and the third relevant index into the constructed melting prediction index model to obtain the melting prediction index.

[0070] In this embodiment, changes in the first correlation index and the second correlation index will cause changes in the third correlation index. In the actual melting process, these three indexes are interdependent and influence each other. The faster phase change rate and the stable flow rate jointly promote the improvement of the melting efficiency, so that the third correlation index increases accordingly. Conversely, any deficiency in any aspect may lead to a decrease in the melting efficiency, thereby causing the third correlation index to decrease accordingly. The three need to be combined in turn to obtain the melting degree prediction index to comprehensively consider their mutual influence; the accuracy of the prediction of the melting degree of the glassy terminal product is improved.

[0071] Specifically, the preset allocation weight, the first correlation index, the second correlation index and the third correlation index are input into the constructed melting prediction index model to obtain the melting prediction index. The melting prediction index is obtained through the melting prediction index model in the following manner:

[0072]

[0073] Where Y m represents the melting prediction index of the mth pretreated sample, m=1,2,...,n, m represents the number of the pretreated sample, n represents the total number of pretreated samples, τ1 is the preset first allocation weight, τ2 is the preset second allocation weight, τ3 is the preset third allocation weight, represents the first correlation index of the mth pre-processed sample, represents the second correlation index of the mth pre-treated sample, represents the third correlation index of the mth pre-treated sample.

[0074] Specifically, the preset first allocation weight is obtained from a melting database; in a specific embodiment, a fitting curve is obtained by fitting the relationship between the first correlation index and the corresponding melting prediction index within a historical period, and the first correlation index is input according to the fitting curve to obtain the preset first allocation weight corresponding to the first correlation index.

[0075] Specifically, the preset second allocation weight is obtained from a melting database; in a specific embodiment, a fitting curve is obtained by fitting the relationship between the second correlation index and the corresponding melting prediction index within a historical period, and the second correlation index is input according to the fitting curve to obtain the preset second allocation weight corresponding to the second correlation index.

[0076] Specifically, the preset third allocation weight is obtained from the melting database; in a specific embodiment, a fitting curve is obtained by fitting the relationship between the third correlation index and the corresponding melting prediction index within a historical period, and the third correlation index is input according to the fitting curve to obtain the preset third allocation weight corresponding to the third correlation index.

[0077] In summary, the embodiment of the present application obtains characteristic data by performing data preprocessing and feature extraction on melting-related data, and then obtains a first correlation index, a second correlation index and a third correlation index based on the characteristic data. Finally, the first correlation index, the second correlation index and the third correlation index are input into the constructed melting prediction index model to obtain the melting prediction index, thereby improving the reliability of the prediction process, and further achieving the effect of improving the accuracy of the prediction of the melting point of the glassy terminal product, effectively solving the problem of low accuracy of the prediction of the melting point of the glassy terminal product in the prior art.

[0078] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0079] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0080] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0081] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1The steps for the functions specified in one or more boxes.

[0082] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0083] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for constructing a melting degree prediction model for a high-temperature molten glassy end product, characterized in that: The following steps are involved: S1, obtaining melting point related data according to the obtained pre-processed terminal product sample, and performing data pre-processing and feature extraction on the melting point related data to obtain feature data, wherein the pre-processed terminal product sample is a sample that has been subjected to melting treatment; S2, acquiring a first correlation index, a second correlation index and a third correlation index according to the characteristic data, wherein the first correlation index is used to evaluate the change of the phase change rate during the melting process, the second correlation index is used to evaluate the change of the flow rate during the melting process, and the third correlation index is used to evaluate the change of the melting efficiency during the melting process; S3, inputting the first correlation index, the second correlation index and the third correlation index into the constructed melting degree prediction index model to obtain a melting degree prediction index, wherein the melting degree prediction index is used to quantify the melting degree of the high-temperature molten glass-like terminal product; The specific process of obtaining the first correlation index is as follows: At a preset time point, first experimental data is measured by a first measuring device, and a first free energy of the pretreated sample is obtained by a contact angle measurement method, wherein the first measuring device includes a transmission electron microscope, a melting point meter, and a differential scanning calorimeter, and the first experimental data includes a first radius, a zero-dimensional particle melting temperature, and a first melting enthalpy; Obtaining a preset first melting temperature threshold, a preset second melting temperature threshold, a preset first weight, and a preset second weight from a database, and obtaining a first correlation index by combining a first melting enthalpy, a first radius, a zero-dimensional particle melting temperature, and a first free energy; The specific process of obtaining the second correlation index is as follows: At a preset time point, the density of the pretreated sample is measured by a densitometer, thereby obtaining a first molar volume of the pretreated sample; Obtaining a reference constant, a property constant, and a preset melting temperature threshold from a database, and obtaining a second correlation index in combination with the first molar volume; The specific process of obtaining the third correlation index is as follows: measuring third experimental data at a preset time point by a third measuring device, wherein the third measuring device includes an electronic balance and a differential scanning calorimeter, and the third experimental data includes the mass, melting latent heat and first heat of the pretreated sample; Obtaining a preset minimum mass threshold, a preset minimum heat threshold, and a preset maximum total heat threshold from a database, and obtaining a third correlation index by combining the mass, the latent heat of fusion, and the first heat; The first correlation index is calculated using the following formula: In the formula, represents the first correlation index of the mth pre-treated sample, m = 1, 2, ..., n, m represents the number of the pre-treated sample, n represents the total number of pre-treated samples, γ im represents the first free energy of the mth pretreated sample at the ith preset time point, i = 1, 2, ..., r, i represents the number of the preset time point, r represents the total number of preset time points, H im represents the first melting enthalpy of the mth pretreated sample at the i-th preset time point, r im represents the first radius of the mth pre-treated sample at the ith preset time point, T im represents the zero-dimensional particle melting temperature of the m-th pretreated sample at the i-th preset time point, T0 represents the preset first melting temperature threshold, T2 represents the preset second melting temperature threshold, σ1 represents the preset first weight, represents the preset second weight, and e represents a natural constant; The second correlation index is calculated using the following formula: In the formula, represents the second correlation index of the mth pre-treated sample, m = 1, 2, ..., n, m represents the number of the pre-treated sample, n represents the total number of pre-treated samples, T 01 Represents the reference constant, V i represents the first molar volume of the mth pretreated sample at the i-th preset time point, i=1,2,...,r, i represents the number of the preset time point, r represents the total number of the preset time points, A represents the material property constant, T1 represents the preset melting temperature threshold, and e represents the natural constant; The third correlation index is calculated using the following formula: In the formula, represents the third correlation index of the mth pre-treated sample, m = 1, 2, ..., n, m represents the number of the pre-treated sample, n represents the total number of pre-treated samples, M im represents the mass of the mth pre-treated sample at the ith preset time point, i = 1, 2, ..., r, i represents the number of the preset time point, r represents the total number of preset time points, L im represents the melting latent heat of the mth pretreated sample at the i-th preset time point, Q im represents the first heat of the mth pretreated sample at the i-th preset time point, M0 represents the preset minimum mass threshold, L0 represents the preset minimum heat threshold, Q0 represents the preset maximum total heat threshold, and e represents a natural constant; The specific acquisition process of the melting degree prediction index is as follows: Acquire preset allocation weights from a melting database, wherein the preset allocation weights include a preset first allocation weight, a preset second allocation weight, and a preset third allocation weight, wherein the preset first allocation weight is used to evaluate the influence of the first correlation index on the melting prediction index model, the preset second allocation weight is used to evaluate the influence of the second correlation index on the melting prediction index model, and the preset third allocation weight is used to evaluate the influence of the third correlation index on the melting prediction index model; Inputting the preset allocation weight, the first correlation index, the second correlation index and the third correlation index into the constructed melting degree prediction index model to obtain the melting degree prediction index; The melting point prediction index is obtained by the melting point prediction index model as follows: Where Y m represents the melting prediction index of the mth pretreated sample, m=1,2,...,n, m represents the number of the pretreated sample, n represents the total number of pretreated samples, τ1 is the preset first allocation weight, τ2 is the preset second allocation weight, τ3 is the preset third allocation weight, represents the first correlation index of the mth pre-processed sample, represents the second correlation index of the mth pre-treated sample, represents the third correlation index of the mth pre-treated sample.

2. A method for constructing a melting degree prediction model for a high-temperature molten glassy end product according to claim 1, characterized in that: The specific process of obtaining the pre-processed terminal product sample is as follows: Acquire a high temperature molten sample according to preset information, and pre-treat the high temperature molten sample to obtain a pre-treated sample, wherein the preset information includes sample hardness, raw material type and chemical composition of the raw material, and the pre-treating means crushing, screening and mixing the high temperature molten sample; The pretreated sample is melt-treated to obtain a pretreated terminal product sample.

3. The method for constructing a melting degree prediction model for a high-temperature molten glassy end product according to claim 1, characterized in that: The specific steps of data preprocessing are as follows: Performing missing value processing on the melting-related data to obtain preprocessed first data, wherein the missing value processing means filling the missing values ​​in the melting-related data by using the median; The preprocessed first data is subjected to standardization processing to obtain preprocessed data, wherein the standardization processing is used to eliminate different dimensions in the preprocessed first data to ensure that the format of the preprocessed data is consistent.

4. A method for constructing a melting degree prediction model for a high-temperature molten glassy end product as claimed in claim 3, characterized in that: The specific steps of the feature extraction are as follows: Extracting the melting point feature of the preprocessed data according to the preset melting point feature to obtain first feature data, wherein the first feature data represents the preprocessed data after the melting point feature extraction; Performing melting heat feature extraction on the first feature data according to the preset melting heat feature to obtain second feature data, wherein the second feature data represents the first feature data after the melting heat feature extraction; Performing phase change feature extraction on the second feature data according to the preset phase change feature to obtain third feature data; The feature data includes first feature data, second feature data and third feature data.

Images