A real-time quantitative analysis and optimization method for sintering operation parameters

Abstract

The present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters, which belongs to the technical field of sintering process. The method comprises the following steps: collecting data related to the sintering process, identifying key parameters that affect the quality and technical indicators of the sintering process from the relevant data; obtaining the historical data range of the key parameters, dividing data intervals according to the historical data range, and combining the data intervals to obtain a sintering production category set; establishing a parameter prediction model based on the sintering production category set, and determining the quantitative mapping between key parameters according to the output result of the parameter prediction model; determining the current sintering production category according to the sintering process parameters currently collected in real time, and then determining the quantitative adjustment effect of the current key parameters according to the quantitative mapping, optimizing the sintering process, improving product quality and production efficiency, and improving the accuracy of prediction.

Description

Technical Field

[0001] The invention relates to the technical field of sintering process, and in particular to a real-time quantitative analysis and optimization method for sintering operation parameters. Background Art

[0002] The sintering process is an important process in materials science and engineering, and is widely used in the production of metals, ceramics and other materials. With the development of Industry 4.0 and intelligent manufacturing, real-time data monitoring and intelligent control have gradually become the key to improving production efficiency and product quality. During the sintering process, the mutual influence between various parameters is complex. Traditional statistical methods or simple machine learning models are difficult to capture these nonlinear and interactive effects, resulting in insufficient model accuracy.

[0003] Therefore, the present invention provides a method for real-time quantitative analysis and optimization of sintering operation parameters. Summary of the invention

[0004] The present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters. The method collects sintering process data through the system, identifies key parameters and divides historical data intervals to form a sintering production category set. A parameter prediction model is established based on the set to quantify the relationship between key parameters, monitor the current production parameters in real time, and dynamically adjust the key parameters in combination with the model output, thereby optimizing the sintering process, improving product quality and production efficiency, and improving the accuracy of the prediction.

[0005] The present invention provides a method for real-time quantitative analysis and optimization of sintering operation parameters, comprising:

[0006] Step 1: Collect data related to the sintering process, and identify key parameters that affect the quality and technical indicators of the sintering process from the relevant data;

[0007] Step 2: Obtain the historical data range of the key parameters, divide the data intervals according to the historical data range, and combine the data intervals to obtain a sintering production category set;

[0008] Step 3: establishing a parameter prediction model based on the sintering production category set, and determining the quantitative mapping between key parameters according to the output results of the parameter prediction model;

[0009] Step 4: Determine the current sintering production category based on the current real-time collected sintering process parameters, and then determine the quantitative adjustment effect of the current key parameters based on the quantitative mapping;

[0010] Wherein, a parameter prediction model is established based on the sintering production category set, and a quantitative mapping between key parameters is determined according to the output result of the parameter prediction model, including:

[0011] Based on each sintering production category data, a first quantitative relationship among sintering raw material parameters, operating parameters and state parameters is established;

[0012] Selecting a first feature related to a state parameter from sintering raw material parameters and operation parameters, creating an interaction item based on the first feature, and deriving a second feature according to the interaction item;

[0013] Establishing a first parameter prediction model based on the first quantitative relationship, the first feature and the second feature;

[0014] Based on each sintering production category data, a second quantitative relationship between the state parameter and the quality parameter is established, and a second parameter prediction model is established by combining the second quantitative relationship with the single sintering production category data;

[0015] The first parameter prediction model is used to quantitatively map the sintering raw material parameters and the operation parameters into state parameters, and the second parameter prediction model is used to quantitatively map the state parameters into quality parameters.

[0016] The present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters, collects data related to the sintering process, and identifies key parameters that affect the quality and technical indicators of the sintering process from the relevant data, including:

[0017] Collect first relevant data in the sintering process through laboratory experiments, collect second relevant data in the sintering process according to the actual production process, and obtain third relevant data in the sintering process according to relevant literature;

[0018] Performing a first classification on the first relevant data, the second relevant data, and the third relevant data, and performing a second classification based on the first classification to obtain quantitative data and qualitative data;

[0019] The distribution characteristics of the first classification results are determined based on the quantitative data, and the high-frequency categories in the first classification results are counted by frequency statistics of the qualitative data to determine the key parameters.

[0020] The present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters, which determines the distribution characteristics of the first classification result based on quantitative data, performs frequency statistics on the high-frequency categories in the first classification result of qualitative data, and then determines the key parameters, including:

[0021] Determine the independent variable according to the quantitative parameters obtained from the quantitative data, determine the dependent variable according to the quality of the sintering process, input the independent variable and the dependent variable into a linear regression model, and identify the key quantitative parameters affecting the quality and technical indicators of the sintering process based on the distribution characteristics;

[0022] Analyze the relationship between qualitative parameters and sintering process quality and technical indicators, and determine key qualitative parameters based on the relationship analysis results and the high-frequency categories;

[0023] Key parameters are derived according to the key quantitative parameters and the key qualitative parameters, wherein the key parameters include sintering raw material parameters, state parameters, operation parameters and quality parameters.

[0024] The present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters, which obtains the historical data range of key parameters, divides data intervals according to the historical data range, and combines the data intervals to obtain a sintering production category set, including:

[0025] Based on the distribution characteristics, a division method is selected and determined, and according to the historical data ranges corresponding to the sintering raw material parameters, state parameters and operating parameters in the key parameters, the sintering raw material parameters, state parameters and operating parameters are divided into multiple data intervals using the division method, and the data in different intervals do not overlap with each other;

[0026] Set the parameter set and the interval of each parameter, generate all existing parameter combinations through Cartesian product, and then obtain the initial production category set;

[0027] A quality assessment is performed on each set in the initial production category set, and existing parameter combinations in the initial production category set are screened according to a preset quality threshold to obtain a sintering production category set, and the mean value of the parameters in each sintering production category is saved.

[0028] The present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters, sets a parameter set and an interval of each parameter, generates all existing parameter combinations through Cartesian products, and then obtains an initial production category set, including:

[0029] Determine a parameter set according to sintering raw material parameters, state parameters and operation parameters, define a value range for each parameter, and discretize the value range of each parameter into several specific values;

[0030] Generate a Cartesian product based on the specific values, and then obtain all existing parameter combinations;

[0031] According to the history-type table, the initial production category corresponding to each existing parameter combination is determined, and the initial production category set is obtained by combining all the initial production categories.

[0032] The present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters, which performs quality assessment on each set in the initial production category set, including:

[0033] ,in, Indicates that the condition Quality assessment value; Represents the scale index, including mi microscale, me mesoscale, and ma macroscale; n represents the total number of parameters in the parameter set; Represents the i-th parameter in the parameter set In scale The influence coefficient under Represents the i-th parameter in the parameter set In the conditions The impact function on quality assessment is as follows; Representation parameters With parameters The coupling influence function between them; Indicates that the condition Dynamic adjustment function of quality assessment based on historical quality assessment; represents the weight coefficient of the feedback mechanism; represents the weight coefficient of entropy in quality assessment; H represents the entropy during sintering; represents the jth operating condition in the sintering process.

[0034] The present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters, which determines the current sintering production category according to the current real-time collected sintering process parameters, and then determines the quantitative adjustment effect of the current key parameters according to the quantitative mapping, including:

[0035] Determine the range of current sintering process parameters and the current sintering production category;

[0036] Based on the first parameter prediction model and the second parameter prediction model corresponding to the current sintering production category, the first influence of the state parameter relative to the mean on the quality parameter, and the second influence of the operation parameter and the sintering raw material parameter relative to the mean on the state parameter are obtained, so as to obtain the quantitative adjustment effect of the operation parameter on the quality parameter under the current sintering production category;

[0037] According to the quantitative adjustment effect, the optimization direction and corresponding optimization values ​​of the operating parameters and sintering raw material parameters under the current sintering production category are fed back.

[0038] Compared with the prior art, the beneficial effects of the present application are as follows: by systematically collecting sintering process data, identifying key parameters and dividing historical data intervals, a sintering production category set is formed, a parameter prediction model is established based on this set, the relationship between key parameters is quantified, the current production parameters are monitored in real time, and the key parameters are dynamically adjusted in combination with the model output, thereby optimizing the sintering process, improving product quality and production efficiency, and improving the accuracy of the prediction.

[0039] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0040] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0042] Figure 1 It is a flow chart of a method for real-time quantitative analysis and optimization of sintering operation parameters provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0044] Embodiment 1:

[0045] The embodiment of the present invention provides a method for real-time quantitative analysis and optimization of sintering operation parameters, such as Figure 1 As shown, including:

[0046] Step 1: Collect data related to the sintering process, and identify key parameters that affect the quality and technical indicators of the sintering process from the relevant data;

[0047] Step 2: Obtain the historical data range of the key parameters, divide the data intervals according to the historical data range, and combine the data intervals to obtain a sintering production category set;

[0048] Step 3: establishing a parameter prediction model based on the sintering production category set, and determining the quantitative mapping between key parameters according to the output results of the parameter prediction model;

[0049] Step 4: Determine the current sintering production category based on the current real-time collected sintering process parameters, and then determine the quantitative adjustment effect of the current key parameters based on the quantitative mapping;

[0050] Wherein, a parameter prediction model is established based on the sintering production category set, and a quantitative mapping between key parameters is determined according to the output result of the parameter prediction model, including:

[0051] Based on each sintering production category data, a first quantitative relationship among sintering raw material parameters, operating parameters and state parameters is established;

[0052] Selecting a first feature related to a state parameter from sintering raw material parameters and operation parameters, creating an interaction item based on the first feature, and deriving a second feature according to the interaction item;

[0053] Establishing a first parameter prediction model based on the first quantitative relationship, the first feature and the second feature;

[0054] Based on each sintering production category data, a second quantitative relationship between the state parameter and the quality parameter is established, and a second parameter prediction model is established by combining the second quantitative relationship with the single sintering production category data;

[0055] The first parameter prediction model is used to quantitatively map the sintering raw material parameters and the operation parameters into state parameters, and the second parameter prediction model is used to quantitatively map the state parameters into quality parameters.

[0056] In this embodiment, the relevant data includes first relevant data, second relevant data and third relevant data.

[0057] In this embodiment, the key parameters are obtained by inputting the independent variables in the quantitative data and the sintering quality as dependent variables into the linear regression model, identifying the key quantitative parameters affecting the sintering quality, analyzing the relationship between the qualitative parameters and the sintering quality, determining the key qualitative parameters in combination with the frequency statistical results, integrating the key quantitative parameters and the qualitative parameters, and deriving the key parameters including sintering raw material parameters, state parameters, operating parameters and quality parameters.

[0058] In this embodiment, the technical indicators are specific standards or indicators used to evaluate the sintering process and product quality, and are usually expressed by quantitative data, such as the compressive strength, density, porosity, etc. of the sintered product.

[0059] In this embodiment, the sintering process quality refers to the quality characteristics of the product during the sintering process, which generally include strength, uniformity, appearance, etc. For example, the strength of the sintered product is 60 MPa and the porosity is 5%.

[0060] In this embodiment, the historical data range is the entire range of parameter values ​​collected in past production or experiments, for example, the historical data range of sintering temperature is 1000° C. to 1400° C., and the historical data range of raw material particle size is 30 microns to 50 microns.

[0061] In this embodiment, the data interval is a specific interval formed by dividing the historical data range according to the division method. For example, in the division of sintering temperature, the possible intervals include 1000-1100°C, 1100-1200°C, 1200-1300°C, and 1300-1400°C.

[0062] In this embodiment, by analyzing the historical data range of key parameters and selecting a suitable division method, the sintering raw materials, state and operating parameters are divided into non-overlapping data intervals, and the Cartesian product is used to generate all possible parameter combinations to form an initial production category set.

[0063] In this embodiment, the sintering production category set refers to data intervals and combinations divided according to key parameters (sintering raw material parameters, state parameters, and operation parameters), and the production categories that meet the quality standards are screened out after quality evaluation.

[0064] In this embodiment, the parameter prediction model includes a first parameter prediction model and a second parameter prediction model. By establishing a data model of the sintering production category, a first quantitative relationship between raw materials, operation and state parameters is first constructed, and relevant features are selected therefrom to generate interaction terms to form a second feature. Based on these features, a first parameter prediction model is established, and a second quantitative relationship between state parameters and quality parameters is established. The single category data is combined to form a second parameter prediction model.

[0065] In this embodiment, quantitative mapping is the process of mapping parameters to other parameters through a model. For example, the sintering raw material parameters (such as sintering temperature and raw material particle size) are quantitatively mapped to state parameters (such as sintering time) using the first parameter prediction model, and then the state parameters are quantitatively mapped to quality parameters (such as product strength) using the second parameter prediction model.

[0066] In this embodiment, the quantitative adjustment effect is the relationship between the state parameter and the quality parameter calculated by the model, which reflects the degree of influence of the change of the operating parameter and the raw material parameter on the quality parameter. For example, assuming that the first parameter prediction model of the current sintering production category shows that the relationship between the sintering time (state parameter) and the product strength (quality parameter) is: Y=0.5·X3+10, where Y is the product strength and X3 is the sintering time. If the mean value of the current state parameter is 120 minutes, the calculated quantitative adjustment effect is: ΔY=0.5·(X3−120).

[0067] In this embodiment, the first quantitative relationship describes the quantitative relationship between sintering raw material parameters, operating parameters and state parameters, which is usually established through regression analysis or other modeling methods. For example, assuming that the relationship between raw material particle size (X1), sintering temperature (X2) and sintering time (X3) can be expressed as: X3=a·X1+b·X2+cX, where a, b and c are coefficients obtained through data analysis.

[0068] In this embodiment, the first feature is a feature related to the state parameters selected from the sintering raw material parameters and the operating parameters. For example, if the sintering temperature and the raw material particle size have a significant effect on the sintering time, the sintering temperature (X2) and the raw material particle size (X1) can be selected as the first feature.

[0069] In this embodiment, the interaction term is a new feature formed by combining the first features, which represents the interaction between the two features. For example, if the selected first features are sintering temperature (X2) and raw material particle size (X1), the interaction term can be expressed as: interaction term = X1·X2.

[0070] In this embodiment, the second feature is a new feature generated based on the interaction term, which is used to enhance the expressive power of the model. For example, the second feature can be the interaction term itself, or a combination with other features, for example: the second feature = X1·X2+X1².

[0071] In this embodiment, the input of the first parameter prediction model is the sintering raw material parameters and the operation parameters, and the output is the state parameters (such as sintering time). For example, the input is (X1, X2, X1 \cdot X2), and the output is the sintering time (X3).

[0072] In this embodiment, the second quantitative relationship is a quantitative relationship describing the state parameter and the quality parameter, which is usually established through regression analysis or other modeling methods. For example, it is assumed that the relationship between the sintering time (X3) and the product strength (Y) can be expressed as Y=d·X3+e, where d and e are coefficients obtained through data analysis.

[0073] In this embodiment, the input of the second parameter prediction model is a state parameter (such as sintering time), and the output is a quality parameter (such as product strength). For example, the input is (X3) and the output is product strength (Y).

[0074] In this embodiment, by establishing a data model of the sintering production category, the first quantitative relationship between raw materials, operation and state parameters is first constructed, and relevant features are selected therefrom to generate interaction terms to form second features. Based on these features, a first parameter prediction model is established, and a second quantitative relationship between state parameters and quality parameters is established. The single category data is combined to form a second parameter prediction model. Through these two models, quantitative mapping of parameters is achieved, thereby enhancing the model's prediction ability for state and quality.

[0075] The working principle and beneficial effects of the above technical solution are: by systematically collecting sintering process data, identifying key parameters and dividing historical data intervals, forming a sintering production category set, establishing a parameter prediction model based on this set, quantifying the relationship between key parameters, monitoring current production parameters in real time, combining model output, and dynamically adjusting key parameters, thereby optimizing the sintering process, improving product quality and production efficiency, and improving the accuracy of prediction.

[0076] Embodiment 2:

[0077] The embodiment of the present invention provides a method for real-time quantitative analysis and optimization of sintering operation parameters, collects data related to the sintering process, and identifies key parameters that affect the quality and technical indicators of the sintering process from the relevant data, including:

[0078] Collect first relevant data in the sintering process through laboratory experiments, collect second relevant data in the sintering process according to the actual production process, and obtain third relevant data in the sintering process according to relevant literature;

[0079] Performing a first classification on the first relevant data, the second relevant data, and the third relevant data, and performing a second classification based on the first classification to obtain quantitative data and qualitative data;

[0080] The distribution characteristics of the first classification results are determined based on the quantitative data, and the high-frequency categories in the first classification results are counted by frequency statistics of the qualitative data to determine the key parameters.

[0081] In this embodiment, the first relevant data is direct data obtained through laboratory experiments, which are usually collected under controlled conditions, such as temperature, pressure, time, raw material composition and other data measured during sintering experiments in the laboratory.

[0082] In this embodiment, the second relevant data is data collected based on the actual production process, reflecting the actual situation in the production environment, such as the actual operating temperature, atmosphere, production speed, product quality (such as strength, density) and other data of the sintering furnace recorded in the production line.

[0083] In this embodiment, the third relevant data is data obtained based on relevant literature, usually data from existing research or standards, such as research results in academic papers on the impact of sintering temperature on material properties, recommended parameter values ​​for the sintering process in industry standards, etc.

[0084] In this embodiment, the first classification is a preliminary classification of the collected data, usually based on certain significant features, for example, the parameters of the sintering process are divided into sintering raw material parameters (such as composition, particle size) and operating parameters (such as temperature, time).

[0085] In this embodiment, the second classification is a more detailed classification based on the first classification, and may be based on other characteristics or relationships, for example, the sintering raw material parameters are further classified into chemical composition (such as oxide content) and physical properties (such as particle size distribution).

[0086] In this embodiment, quantitative data is data that can be represented by numbers and subjected to mathematical operations, such as sintering temperature (such as 1200°C), time (such as 2 hours), and material strength (such as 50 MPa); qualitative data is data that describes properties or characteristics and is usually represented by categories or labels, such as the appearance of the sintered product (such as smooth, rough), and whether it is qualified or not (qualified, unqualified).

[0087] In this embodiment, the distribution feature is the distribution of quantitative data in different categories, which is used to understand the central tendency and variability of the data. For example, by analyzing the distribution of material strength at different sintering temperatures, it may be found that the strength is highest within a certain temperature range.

[0088] In this embodiment, the high frequency category is a category that appears more frequently in the qualitative data, indicating its importance in the sample. For example, in product quality assessment, it is found that the frequency of occurrence of the "qualified" category is higher than that of "minor defect" or "serious defect".

[0089] The working principle and beneficial effects of the above technical solution are: by collecting three types of relevant data from the laboratory, actual production and literature, classifying and analyzing them. First, the data is classified into the first category, and then the second category is classified based on this to distinguish quantitative and qualitative data. By analyzing the distribution characteristics of quantitative data and the frequency statistics of qualitative data, high-frequency categories are identified, and then the key parameters affecting the sintering process are determined. Combining data from multiple sources provides a more comprehensive perspective and enhances the accuracy of the analysis.

[0090] Embodiment 3:

[0091] The embodiment of the present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters, which determines the distribution characteristics of the first classification result based on quantitative data, performs frequency statistics on the qualitative data to obtain high-frequency categories in the first classification result, and then determines the key parameters, including:

[0092] Determine the independent variable according to the quantitative parameters obtained from the quantitative data, determine the dependent variable according to the quality of the sintering process, input the independent variable and the dependent variable into a linear regression model, and identify the key quantitative parameters affecting the quality and technical indicators of the sintering process based on the distribution characteristics;

[0093] Analyze the relationship between qualitative parameters and sintering process quality and technical indicators, and determine key qualitative parameters based on the relationship analysis results and the high-frequency categories;

[0094] Key parameters are derived according to the key quantitative parameters and the key qualitative parameters, wherein the key parameters include sintering raw material parameters, state parameters, operation parameters and quality parameters.

[0095] In this embodiment, the linear regression model is a statistical method used to analyze the linear relationship between the independent variable (input) and the dependent variable (output) by fitting a best straight line to predict the dependent variable. For example, a linear regression model is used to analyze the effects of sintering temperature, time and raw material particle size (independent variables) on sintering strength (dependent variable).

[0096] In this embodiment, the key quantitative parameters are quantitative data parameters identified in the analysis that have a significant impact on the sintering quality, such as sintering temperature (such as 1250°C), sintering time (such as 3 hours), raw material particle size (such as 40 microns), etc.

[0097] In this embodiment, the relationship analysis result is a conclusion or finding obtained by analyzing the relationship between qualitative parameters and the quality of the sintering process. For example, the analysis found that the use of a specific type of raw material (such as high-purity alumina) is positively correlated with product quality.

[0098] In this embodiment, the key qualitative parameters are qualitative characteristics or categories identified in the analysis that have a significant impact on the sintering quality, such as the type of raw materials (such as alumina, silicate), the operation mode during the production process (such as fast cooling or slow cooling).

[0099] The working principle and beneficial effects of the above technical solution are: by inputting the independent variables in the quantitative data and the sintering quality as dependent variables into the linear regression model, the key quantitative parameters affecting the sintering quality are identified, the relationship between the qualitative parameters and the sintering quality is analyzed, and the key qualitative parameters are determined in combination with the frequency statistical results. The key quantitative parameters and qualitative parameters are integrated to obtain key parameters including sintering raw material parameters, state parameters, operating parameters and quality parameters, which is helpful to effectively control the production process, ensure product quality, systematically identify the key parameters affecting the sintering quality, and improve the optimization effect.

[0100] Embodiment 4:

[0101] The embodiment of the present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters, which obtains the historical data range of key parameters, divides data intervals according to the historical data range, and combines the data intervals to obtain a sintering production category set, including:

[0102] Based on the distribution characteristics, a division method is selected and determined, and according to the historical data ranges corresponding to the sintering raw material parameters, state parameters and operating parameters in the key parameters, the sintering raw material parameters, state parameters and operating parameters are divided into multiple data intervals using the division method, and the data in different intervals do not overlap with each other;

[0103] Set the parameter set and the interval of each parameter, generate all existing parameter combinations through Cartesian product, and then obtain the initial production category set;

[0104] A quality assessment is performed on each set in the initial production category set, and existing parameter combinations in the initial production category set are screened according to a preset quality threshold to obtain a sintering production category set, and the mean value of the parameters in each sintering production category is saved.

[0105] In this embodiment, the partitioning method is a strategy or algorithm for dividing the data range into multiple non-overlapping intervals, for example, dividing the sintering temperature range (such as 1000°C to 1400°C) into several intervals, such as 1000-1100°C, 1100-1200°C, 1200-1300°C, and 1300-1400°C.

[0106] In this embodiment, the Cartesian product is for multiple sets, and the Cartesian product is all possible combinations, that is, the combination of each element in each set with each element in other sets. For example, if the sintering temperature range is [1000-1100°C, 1100-1200°C], and the raw material particle size range is [30-40 microns, 40-50 microns], then the Cartesian product will generate combinations such as (1000-1100°C, 30-40 microns), (1000-1100°C, 40-50 microns), (1100-1200°C, 30-40 microns), (1100-1200°C, 40-50 microns).

[0107] In this embodiment, an existing parameter combination is a valid combination that meets certain conditions among all combinations generated by the Cartesian product. For example, if a combination has a corresponding production record in the historical data, the combination is considered to be an existing parameter combination, such as (1100-1200°C, 30-40 microns).

[0108] In this embodiment, the initial production category set is a set of all existing parameter combinations, representing different production conditions. For example, if the generated valid combinations are (1100-1200°C, 30-40 microns) and (1000-1100°C, 40-50 microns), these combinations constitute the initial production category set.

[0109] In this embodiment, the preset quality threshold is a standard value set in the quality assessment, which is used to determine whether the production category is qualified. For example, the set quality threshold is that the product strength is not less than 50 MPa.

[0110] In this embodiment, screening is to evaluate the parameter combinations in the initial production category set according to a preset quality threshold and retain the combinations that meet the conditions. For example, if the quality evaluation result of a combination is a strength of 55 MPa, which meets the quality threshold, the combination is retained; if the strength is 45 MPa, it is screened out.

[0111] The working principle and beneficial effects of the above technical solution are: by analyzing the historical data range of key parameters and selecting a suitable division method, the sintering raw materials, status and operating parameters are divided into non-overlapping data intervals, and all possible parameter combinations are generated using Cartesian products to form an initial production category set. The quality of each category is evaluated, and qualified sintering production categories are screened out according to preset thresholds. Finally, the mean value of the parameters in each category is saved to help optimize the production process, reduce the risk of relying on experience, and improve production efficiency and product quality.

[0112] Embodiment 5:

[0113] The embodiment of the present invention provides a real-time quantitative analysis and optimization method for sintering operation parameters, sets a parameter set and an interval of each parameter, generates all existing parameter combinations through Cartesian products, and then obtains an initial production category set, including:

[0114] Determine a parameter set according to sintering raw material parameters, state parameters and operation parameters, define a value range for each parameter, and discretize the value range of each parameter into several specific values;

[0115] Generate a Cartesian product based on the specific values, and then obtain all existing parameter combinations;

[0116] According to the history-type table, the initial production category corresponding to each existing parameter combination is determined, and the initial production category set is obtained by combining all the initial production categories.

[0117] In this embodiment, the value interval is the range of allowable values ​​set for each parameter, usually expressed as a minimum value and a maximum value. For example, for the sintering temperature, the value interval can be set to 1000°C to 1400°C; for the raw material particle size, the value interval can be set to 30 microns to 50 microns.

[0118] In this embodiment, the specific values ​​are specific values ​​obtained by discretizing the value range, which represent the actual values ​​that the parameters can take. For example, if the value range of the sintering temperature is 1000°C to 1400°C, it can be discretized into specific values ​​[1000°C, 1100°C, 1200°C, 1300°C, 1400°C]; for the raw material particle size, if the value range is 30 microns to 50 microns, it can be discretized into [30 microns, 35 microns, 40 microns, 45 microns, 50 microns].

[0119] In this embodiment, the history-type table is a table containing historical data and corresponding production categories, which is usually used to find the production results or categories of a specific parameter combination, the sintering temperature is 1100°C, the raw material particle size is 30 microns, and the production category is category A.

[0120] The working principle and beneficial effects of the above technical solution are: by analyzing the historical data of sintering raw materials, status and operating parameters, these parameters are divided into multiple non-overlapping data intervals using a specific division method. After setting the parameter set and its interval, all possible parameter combinations are generated through Cartesian product to form an initial production category set. The quality of each combination is evaluated, and qualified sintering production categories are screened out according to the preset quality threshold. Finally, the mean value of the parameters in each category is saved to improve the validity of the data, improve product quality and production efficiency, and reduce the unqualified rate.

[0121] Embodiment 6:

[0122] The embodiment of the present invention provides a method for real-time quantitative analysis and optimization of sintering operation parameters, which performs quality assessment on each set in the initial production category set, including:

[0123] ,in, Indicates that the condition Quality assessment value; Represents the scale index, including mi microscale, me mesoscale, and ma macroscale; n represents the total number of parameters in the parameter set; Represents the i-th parameter in the parameter set In scale The influence coefficient under Represents the i-th parameter in the parameter set In the conditions The impact function on quality assessment is as follows; Representation parameters With parameters The coupling influence function between them; Indicates that the condition Dynamic adjustment function of quality assessment based on historical quality assessment; represents the weight coefficient of the feedback mechanism; represents the weight coefficient of entropy in quality assessment; H represents the entropy during sintering; represents the jth operating condition in the sintering process.

[0124] In this embodiment, ,in, Representation parameters and scale The coefficient of the relevant quadratic term; Representation parameters and scale The coefficient of the related first-order term; Representation parameters and scale The associated constant term.

[0125] In this embodiment, ,in, Representation parameters With parameters The strength of the interaction between them.

[0126] In this embodiment, ,in, Indicates that the condition Lower quality indicators; Represents the historical average quality index.

[0127] The working principle and beneficial effects of the above technical solution are: by establishing a multi-scale quality assessment model, comprehensively considering each parameter in the parameter set and its influence coefficient, using influence functions and coupled influence functions, analyzing the influence of different parameters on quality at micro, meso and macro scales, and optimizing the quality assessment process through dynamic adjustment functions and feedback mechanisms, combined with historical quality assessment data, ensuring the accuracy and real-time nature of the assessment results, making the quality assessment more accurate and able to reflect the production status in a timely manner.

[0128] Embodiment 7:

[0129] The embodiment of the present invention provides a method for real-time quantitative analysis and optimization of sintering operation parameters, which determines the current sintering production category according to the current real-time collected sintering process parameters, and then determines the quantitative adjustment effect of the current key parameters according to the quantitative mapping, including:

[0130] Determine the range of current sintering process parameters and the current sintering production category;

[0131] Based on the first parameter prediction model and the second parameter prediction model corresponding to the current sintering production category, the first influence of the state parameter relative to the mean on the quality parameter, and the second influence of the operation parameter and the sintering raw material parameter relative to the mean on the state parameter are obtained, so as to obtain the quantitative adjustment effect of the operation parameter on the quality parameter under the current sintering production category;

[0132] According to the quantitative adjustment effect, the optimization direction and corresponding optimization values ​​of the operating parameters and sintering raw material parameters under the current sintering production category are fed back.

[0133] In this embodiment, the first impact refers to the degree of influence of the state parameter (such as sintering time) on the quality parameter (such as product strength) relative to its mean under the current sintering production category. This impact can be quantified by the first parameter prediction model, assuming that the first parameter prediction model is: Y = 0.6 · (t-120) + 30, where Y is the product strength and t is the sintering time. If the current state parameter (sintering time) is 120 minutes, the first impact is calculated: when t = 120 minutes: Y = 0.6 · (120-120) + 30 = 30MPa; when t = 130 minutes: Y = 0.6 · (130-120) + 30 = 36MPa, therefore, increasing the sintering time by 10 minutes (relative to the mean) will increase the product strength by 6MPa, reflecting the first impact.

[0134] In this embodiment, the second impact refers to the influence of operating parameters (such as sintering temperature) and sintering raw material parameters (such as raw material particle size) on state parameters (such as sintering time) relative to their mean values. This impact can be quantified by a second parameter prediction model. Assume that the second parameter prediction model is: t=0.4·(T−1250)+120, where T is the sintering temperature. If the current operating parameter (sintering temperature) is 1280°C, the second impact is calculated: when T=1280°C: t=0.4·(1280−1250)+120=124 minutes; when T=1300°C: t=0.4·(1300−1250)+120=128 minutes. Therefore, increasing the sintering temperature by 20°C (relative to the mean value) will increase the sintering time by 4 minutes, reflecting the second impact.

[0135] In this embodiment, the optimization direction is to determine the direction in which the operating parameters and raw material parameters need to be adjusted under the current sintering production category based on the quantitative adjustment effect. For example, if the quantitative adjustment effect shows that increasing the sintering temperature (operating parameter) will lead to an increase in product strength, then the optimization direction is "increasing the sintering temperature."

[0136] In this embodiment, the corresponding optimization value is the parameter value that needs to be adjusted to achieve the optimization goal. For example, if the current sintering temperature is 1250°C, it is recommended to increase it to 1300°C in the hope of improving the product strength. The specific optimization value is "increase by 50°C".

[0137] The working principle and beneficial effects of the above technical solution are: to compare the mean values ​​of the operating parameters of the current sintering production category, evaluate their quantitative adjustment effects, calculate the range of the current parameters to determine the production category, use the first and second parameter prediction models to analyze the influence relationship between the state parameters and the quality parameters, obtain the influence of the operating parameters and the sintering raw material parameters on the state parameters, and based on the quantitative adjustment effects, feedback the optimization direction and values ​​to guide production improvements and improve the efficiency of the sintering process and product quality.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for real-time quantitative analysis and optimization of sintering operation parameters, characterized in that: include: Step 1: Collect data related to the sintering process, and identify key parameters that affect the quality and technical indicators of the sintering process from the relevant data; Step 2: Obtain the historical data range of the key parameters, divide the data intervals according to the historical data range, and combine the data intervals to obtain a sintering production category set; Step 3: establishing a parameter prediction model based on the sintering production category set, and determining the quantitative mapping between key parameters according to the output results of the parameter prediction model; Step 4: Determine the current sintering production category based on the current real-time collected sintering process parameters, and then determine the quantitative adjustment effect of the current key parameters based on the quantitative mapping; Wherein, a parameter prediction model is established based on the sintering production category set, and a quantitative mapping between key parameters is determined according to the output result of the parameter prediction model, including: Based on each sintering production category data, a first quantitative relationship among sintering raw material parameters, operating parameters and state parameters is established; Selecting a first feature related to a state parameter from sintering raw material parameters and operation parameters, creating an interaction item based on the first feature, and deriving a second feature according to the interaction item; Establishing a first parameter prediction model based on the first quantitative relationship, the first feature and the second feature; Based on each sintering production category data, a second quantitative relationship between the state parameter and the quality parameter is established, and a second parameter prediction model is established by combining the second quantitative relationship with the single sintering production category data; The first parameter prediction model is used to quantitatively map the sintering raw material parameters and the operation parameters into state parameters, and the second parameter prediction model is used to quantitatively map the state parameters into quality parameters.

2. A method for real-time quantitative analysis and optimization of sintering operation parameters according to claim 1, characterized in that: Collect data related to the sintering process and identify key parameters that affect the quality and technical indicators of the sintering process from the relevant data, including: Collect first relevant data in the sintering process through laboratory experiments, collect second relevant data in the sintering process according to the actual production process, and obtain third relevant data in the sintering process according to relevant literature; Performing a first classification on the first relevant data, the second relevant data, and the third relevant data, and performing a second classification based on the first classification to obtain quantitative data and qualitative data; The distribution characteristics of the first classification results are determined based on the quantitative data, and the high-frequency categories in the first classification results are counted by frequency statistics of the qualitative data to determine the key parameters.

3. A method for real-time quantitative analysis and optimization of sintering operation parameters according to claim 2, characterized in that: Based on the quantitative data, the distribution characteristics of the first classification results are determined. The high-frequency categories in the first classification results are counted for the qualitative data, and then the key parameters are determined, including: Determine the independent variable according to the quantitative parameters obtained from the quantitative data, determine the dependent variable according to the quality of the sintering process, input the independent variable and the dependent variable into a linear regression model, and identify the key quantitative parameters affecting the quality and technical indicators of the sintering process based on the distribution characteristics; Analyze the relationship between qualitative parameters and sintering process quality and technical indicators, and determine key qualitative parameters based on the relationship analysis results and the high-frequency categories; Key parameters are derived according to the key quantitative parameters and the key qualitative parameters, wherein the key parameters include sintering raw material parameters, state parameters, operation parameters and quality parameters.

4. The method for real-time quantitative analysis and optimization of sintering operation parameters according to claim 2, characterized in that: Obtain the historical data range of the key parameters, divide the data intervals according to the historical data range, and combine the data intervals to obtain a sintering production category set, including: Based on the distribution characteristics, a division method is selected and determined, and according to the historical data ranges corresponding to the sintering raw material parameters, state parameters and operating parameters in the key parameters, the sintering raw material parameters, state parameters and operating parameters are divided into multiple data intervals using the division method, and the data in different intervals do not overlap with each other; Set the parameter set and the interval of each parameter, generate all existing parameter combinations through Cartesian product, and then obtain the initial production category set; A quality assessment is performed on each set in the initial production category set, and existing parameter combinations in the initial production category set are screened according to a preset quality threshold to obtain a sintering production category set, and the mean value of the parameters in each sintering production category is saved.

5. According to the method for real-time quantitative analysis and optimization of sintering operation parameters as claimed in claim 4, a parameter set and an interval of each parameter are set, all existing parameter combinations are generated by Cartesian product, and then an initial production category set is obtained, including: Determine a parameter set according to sintering raw material parameters, state parameters and operation parameters, define a value range for each parameter, and discretize the value range of each parameter into several specific values; Generate a Cartesian product based on the specific values, and then obtain all existing parameter combinations; According to the history-type table, the initial production category corresponding to each existing parameter combination is determined, and the initial production category set is obtained by combining all the initial production categories.

6. A method for real-time quantitative analysis and optimization of sintering operation parameters according to claim 4, characterized in that: A quality assessment is performed on each set of said initial production categories, including: in, Indicates that the condition Quality assessment value; Represents the scale index, including mi microscale, me mesoscale, and ma macroscale; n represents the total number of parameters in the parameter set; Represents the i-th parameter in the parameter set In scale The influence coefficient under Represents the i-th parameter in the parameter set In the conditions The impact function on quality assessment is as follows; Representation parameters With parameters The coupling influence function between them; Indicates that the condition Dynamic adjustment function of quality assessment based on historical quality assessment; represents the weight coefficient of the feedback mechanism; represents the weight coefficient of entropy in quality assessment; H represents the entropy during sintering; represents the jth operating condition in the sintering process.

7. The method for real-time quantitative analysis and optimization of sintering operation parameters according to claim 1, characterized in that: The current sintering production category is determined based on the current real-time collected sintering process parameters, and then the quantitative adjustment effect of the current key parameters is determined based on the quantitative mapping, including: Determine the range of current sintering process parameters and the current sintering production category; Based on the first parameter prediction model and the second parameter prediction model corresponding to the current sintering production category, the first influence of the state parameter relative to the mean on the quality parameter, and the second influence of the operation parameter and the sintering raw material parameter relative to the mean on the state parameter are obtained, so as to obtain the quantitative adjustment effect of the operation parameter on the quality parameter under the current sintering production category; According to the quantitative adjustment effect, the optimization direction and corresponding optimization values ​​of the operating parameters and sintering raw material parameters under the current sintering production category are fed back.

Images