Method and device for sintering ore blending based on liquid phase generation amount of mixed ore

Abstract

The application discloses a method and device for sintering ore blending based on mixed ore liquid phase generation amount. The method establishes a prediction model based on multiple ore blending schemes and corresponding strength data, converts the implicit correlation between the mixed ore liquid phase generation amount and the sinter strength into a quantifiable mathematical mapping, determines the target strength to be achieved for the current production to call the prediction model to reversely deduce the liquid phase generation amount constraint condition, and generates an optimized sintering ore blending scheme with the mixed ore liquid phase generation amount as the core constraint condition. The method forms a closed-loop regulation and control mechanism of data modeling-constraint generation-scheme optimization, can adjust and quickly iterate the constraint condition and the ore blending scheme according to different sinter strength requirements, improves the production adaptability, simultaneously takes the liquid phase generation amount as a link for the key parameter of sintering consolidation, provides theoretical and technical support for the fine control of the sintering process, and finally realizes the overall goal of improving the stability of the sinter strength.

Description

Technical Field

[0001] This application relates to the field of metallurgical technology, specifically to a method and apparatus for sintering blending based on the amount of mixed ore liquid phase generated. Background Technology

[0002] In steel production, sinter is a key raw material for blast furnace smelting, and its quality has a significant impact on steel production. During sintering, carbon combustion provides heat, iron oxides undergo redox reactions, and the formation of the liquid phase (such as calcium ferrite and silicates) is the core of sinter consolidation. The amount of effective liquid phase directly affects the quality of sinter. The particle size distribution of iron ore is an important indicator of the effective liquid phase. Particles of 1-3 mm are nucleating particles, reacting little with mineral flux and producing virtually no liquid phase. The liquid phase is mainly generated by the reaction of adhering particles smaller than 0-1 mm with calcareous flux. The amount of liquid phase in the raw material can be calculated by the mass content of these particles. After mixing and granulation, the mixture forms a specific particle size distribution (0-3 mm < 15%, 3-5 mm 40%-50%, 5-10 mm ≤ 30%, > 10 mm ≤ 10%) to form an effective binder phase, meeting the requirements for sinter consolidation strength.

[0003] Currently, my country needs to rationally blend ore to reduce costs and produce sinter that meets the requirements of blast furnaces. The current sinter blending largely relies on manual matching of raw materials, a cumbersome and complex process that cannot meet the requirements of automation and intelligent control. While some automated methods exist for generating blending constraints using the chemical composition of sinter as a constraint, in practice, this method struggles to guarantee drum strength, thus failing to ensure that the drum strength meets the requirements of blast furnace smelting.

[0004] Therefore, how to achieve more intelligent sintering and ore blending while ensuring that the strength of the sinter drum meets the requirements of blast furnace smelting is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0005] In view of the above-mentioned defects or deficiencies in the prior art, it is desirable to provide a method and apparatus for sintering blending based on the amount of mixed ore liquid phase generated, which can realize more intelligent sintering blending, while ensuring that the strength of the sinter drum meets the requirements of blast furnace smelting.

[0006] In a first aspect, embodiments of this application provide a method for sintering ore blending based on the amount of homogenized ore liquid phase generated, including:

[0007] Based on the strength data of finished sinter corresponding to various sintering blending schemes, a predictive model for the amount of mixed ore liquid phase generated and the strength of finished sinter is established.

[0008] Identify the intensity indicators to be achieved as the target intensity;

[0009] The prediction model is invoked to calculate the range of mixed mineral liquid phase generation required to achieve the target intensity, and constraints on the amount of mixed mineral liquid phase generation are generated.

[0010] An optimized sintering ore blending scheme is generated based on the aforementioned constraints.

[0011] In one embodiment, the step of establishing a predictive model for the amount of mixed liquid phase generated and the strength of the finished sinter based on the strength data of the finished sinter corresponding to multiple sintering blending schemes includes:

[0012] Acquire data on the chemical composition, burn-off rate, and particle size distribution of various raw materials and fuels as metallurgical performance parameters;

[0013] Collect data on the strength of finished sintered ore corresponding to various raw material and fuel ratios during the sintering production process;

[0014] Based on the raw material and fuel ratio and the corresponding metallurgical performance parameters, calculate the amount of liquid phase generated in the blended ore;

[0015] Organize the data on the amount of mixed ore liquid phase generated and the corresponding strength data of finished sintered ore to generate the original dataset;

[0016] A prediction model is built based on the original dataset.

[0017] In one embodiment, calculating the amount of liquid phase generated from the blended ore based on the raw material and fuel ratio and the corresponding metallurgical performance parameters includes:

[0018] The amount of liquid phase generated in the blended ore is calculated based on the content of adhering particles, the burn-off rate, and the dry material amount in the raw material and fuel ratio.

[0019] In one embodiment, calculating the amount of liquid phase generated from the blended ore based on the content of adhering particles, the burn-off rate, and the dry weight in the raw material and fuel ratio includes:

[0020] Calculate the amount of liquid phase generated in the homogenized ore according to Formula 1;

[0021] Formula 1:

[0022] in, This represents the amount of mineral liquid phase generated during homogenization. This refers to the dry weight of each raw material. The content of 0 to 1 mm particles in the raw materials. The burn loss rate of each raw material in the stated raw material and fuel ratio.

[0023] In one embodiment, building a prediction model based on the original dataset includes:

[0024] The relationship between liquid phase generation and intensity was fitted using a nonlinear regression model.

[0025] The nonlinear regression model is trained and validated based on the original dataset.

[0026] The nonlinear regression model that meets the accuracy requirements after verification is used as the prediction model.

[0027] In one embodiment, before training and validating the nonlinear regression model based on the original dataset, the method further includes:

[0028] Perform data cleaning on the original dataset;

[0029] The cleaned dataset is then standardized.

[0030] In one embodiment, the step of calling the prediction model to calculate the range of mixed mineral-liquid phase generation required to achieve the target intensity, and generating constraints on the mixed mineral-liquid phase generation, includes:

[0031] The upper and lower limits of the amount of mixed mineral liquid phase required to achieve the target strength are calculated based on the prediction model.

[0032] Based on the upper limit and the lower limit, Formula 2 is generated as a constraint condition;

[0033] Formula 2:

[0034] Where A is the lower limit value and B is the upper limit value. The proportions of each raw material, The amount of liquid phase in each raw material.

[0035] In one embodiment, after generating the optimized sintering ore blending scheme based on the constraints, the method further includes:

[0036] A cost analysis is performed on the optimized sintering ore blending scheme, and a cost analysis report is generated.

[0037] The performance of the finished sinter is predicted based on the optimized sintering ore blending scheme, and a performance evaluation report is generated.

[0038] Output the optimized sintering ore blending scheme, along with the corresponding cost analysis report and performance evaluation report.

[0039] Secondly, embodiments of this application provide an apparatus for sintering blending based on the amount of homogenized ore liquid phase generated, comprising:

[0040] The model building module is used to establish a predictive model of the amount of mixed ore liquid phase generated and the strength of finished sinter based on the strength data of finished sinter corresponding to various sintering blending schemes.

[0041] The indicator determination module is used to determine the intensity indicator to be achieved, which serves as the target intensity.

[0042] The constraint calculation module is used to call the prediction model to calculate the range of mixed mineral liquid phase generation required to achieve the target intensity, and to generate the constraint conditions for the mixed mineral liquid phase generation.

[0043] The scheme generation module is used to generate an optimized sintering ore blending scheme based on the constraints.

[0044] Thirdly, embodiments of this application provide a server, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the sintering and blending method based on the amount of mixed ore liquid phase generated as described above.

[0045] The sintering blending method provided in this application, based on the amount of liquid phase generated in the mixed ore, establishes a predictive model based on multiple blending schemes and corresponding strength data. It transforms the implicit correlation between the amount of liquid phase generated in the mixed ore and the strength of the sinter into a quantifiable mathematical mapping. By determining the target strength to be achieved in the current production process, the predictive model is used to deduce the liquid phase generation constraint condition in reverse, generating an optimized sintering blending scheme with the liquid phase generation amount as the core constraint. This method forms a closed-loop control mechanism of data modeling, constraint generation, and scheme optimization. It can rapidly iterate the constraint conditions and blending scheme according to different sinter strength requirements, improving production adaptability. Simultaneously, using the liquid phase generation amount, a key parameter for sintering consolidation, as a link, it provides theoretical and technical support for the refined control of the sintering process, ultimately achieving the overall goal of improving the stability of sinter strength.

[0046] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0047] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0048] Figure 1 A schematic flowchart of a sintering blending method based on the amount of mixed ore liquid phase generated is shown in an embodiment of this application.

[0049] Figure 2 An exemplary structural block diagram of an apparatus for sintering blending based on the amount of mixed ore liquid phase generated, provided in an embodiment of this application, is shown.

[0050] Figure 3 A schematic diagram of the structure of a computer system suitable for implementing the server of the present application is shown. Detailed Implementation

[0051] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0052] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present application will now be described in detail with reference to the accompanying drawings and embodiments. Although the embodiments of this application provide method operation instruction steps as shown in the following embodiments or drawings, more or fewer operation instruction steps may be included in the method based on conventional or non-inventive effort. In steps where there is no logically necessary causal relationship, the execution order of these steps is not limited to the execution order provided in the embodiments of this application. In actual processing or when the device executes, the method may be executed sequentially or in parallel according to the method shown in the embodiments or drawings.

[0053] Example 1:

[0054] This embodiment proposes a method for sintering ore blending based on the amount of mixed ore liquid phase generated. Please refer to [reference needed]. Figure 1 , Figure 1 A schematic flowchart of the sintering blending method based on the amount of homogenized ore liquid phase generated, as provided in this embodiment, is shown. Figure 1 As shown, the method includes:

[0055] S101. Based on the strength data of finished sinter corresponding to various sintering blending schemes, establish a prediction model for the amount of mixed ore liquid phase generated and the strength of finished sinter.

[0056] During sintering, the raw materials partially melt or soften at high temperatures to form a liquid phase, which is the key driving force for the consolidation of sinter particles. The amount of liquid phase generated directly affects the mineral phase composition and structural strength of the sinter. Insufficient liquid phase leads to inadequate particle consolidation and low strength; excessive liquid phase easily causes over-melting and adhesion, affecting permeability. Therefore, calculating the amount of liquid phase generated is a core intermediate variable for evaluating the quality of sinter.

[0057] This method proposes that during the sinter production process, multiple sets of ore blending schemes with different raw material ratios and process conditions should be collected first, and the amount of mixed ore liquid phase generated and the strength index of finished sinter (such as drum strength) corresponding to each scheme should be measured simultaneously to form an original dataset.

[0058] Then, nonlinear regression, machine learning, and other methods are used to analyze and fit the original dataset, constructing a mathematical model that can quantitatively describe the dynamic relationship between the amount of liquid phase generated in the mixed ore and the strength of the sinter, serving as a prediction model. This prediction model can be used to predict the strength of the sinter based on the amount of liquid phase generated, or to infer the required range of liquid phase generated based on the target strength, providing data support and decision-making basis for the optimal design of sinter blending.

[0059] S102. Determine the intensity index to be achieved as the target intensity;

[0060] During the optimization of sinter blending, based on blast furnace smelting process specifications, product quality standards, or customer-customized requirements, specific strength performance indicators that the finished sinter must achieve are defined, such as drum strength TI6.3≥75% and compressive strength, etc., and these indicators are used as the core target strength in the blending scheme design. In this embodiment, the specific strength quantification type of the strength indicators is not limited; only drum strength and compressive strength are used as examples. Other strength types can be referred to the description in this embodiment and will not be elaborated further here.

[0061] By quantifying the target intensity, a benchmark is provided for subsequently using predictive models to back-calculate the required range of liquid phase generation, thus giving the optimization of the ore blending scheme a clear performance orientation.

[0062] In this embodiment, the specific value of the target strength is not limited. It is necessary to comprehensively consider the minimum requirements of the smelting process for the strength of sinter, industry quality standards and raw material costs, etc., to ensure that the target value meets both production needs and economic feasibility, thereby laying the foundation for the constraint generation and optimization design of the subsequent ore blending scheme.

[0063] S103. Call the prediction model to calculate the range of mixed mineral liquid phase generation required to achieve the target strength, and generate the constraint conditions for the mixed mineral liquid phase generation.

[0064] Using a predetermined target strength of the finished sinter as input parameter, the established prediction model is invoked to inversely solve for the reasonable range of liquid phase generation required to achieve the target strength. This range is defined as the liquid phase generation range of the mixed ore. The model establishes a nonlinear correlation between liquid phase quantity and strength by fitting historical data. Based on the range of liquid phase generation, quantitative constraints expressed as mathematical inequalities are generated, explicitly requiring that the total liquid phase generation of each raw material and fuel in the blending scheme must fall within this range.

[0065] By using inversion calculations from the predictive model, the strength performance target of sinter is transformed into a specific constraint on the amount of liquid phase generated in the blended ore. This provides core process parameter constraints for the subsequent optimization design of the ore blending scheme, ensuring that the sinter strength meets the requirements while achieving a balance between multi-objective production optimization. This process embodies the reverse derivation logic from end-performance indicators to key intermediate variables and provides an engineering implementation path for the quantitative control and precise ore blending in sintering production.

[0066] S104. Generate an optimized sintering ore blending scheme based on constraints.

[0067] Based on the established constraints on the amount of mixed ore liquid phase generated, the optimal raw material combination scheme is generated using the raw material ratio as the decision variable, which serves as the sintering ore blending scheme.

[0068] The sintering ore blending scheme takes the amount of mixed ore liquid phase generated as the core constraint. At the same time, it can also incorporate multiple constraints such as the feasibility of raw material supply (such as the available amount range of each raw material), process production requirements (such as the limits of indicators such as alkalinity and magnesium oxide content), and cost control targets. It should be noted that in this embodiment, apart from the constraint of the amount of mixed ore liquid phase generated as the core constraint, other constraints are not limited. The above conditions are only used as examples for introduction. The specific constraints can be set according to the actual application scenario and will not be elaborated here.

[0069] Based on the above introduction, the sintering ore blending method based on the amount of liquid phase generated in the mixed ore provided in this embodiment establishes a predictive model based on multiple ore blending schemes and corresponding strength data. This transforms the implicit correlation between the amount of liquid phase generated in the mixed ore and the strength of the sinter into a quantifiable mathematical mapping. By determining the target strength to be achieved in the current production, the predictive model is used to deduce the liquid phase generation constraint condition in reverse. Using the amount of liquid phase generated in the mixed ore as the core constraint condition, an optimized sintering ore blending scheme is generated. This method forms a closed-loop control mechanism of data modeling, constraint generation, and scheme optimization. It can rapidly iterate the constraint condition and ore blending scheme according to raw material fluctuations or process adjustments, improving production adaptability. Simultaneously, using the amount of liquid phase generated—a key parameter for sintering consolidation—as a link, it provides theoretical and technical support for the refined control of the sintering process, ultimately achieving the overall goal of improving the stability of the sinter strength.

[0070] Example 2:

[0071] The specific construction process of the prediction model is not limited in the above embodiments. In order to achieve accurate mapping of process parameters based on raw material characteristics, this embodiment proposes a prediction model construction process. Step S101 establishes a prediction model of the amount of mixed ore liquid phase generated and the strength of finished sinter based on the strength data of finished sinter corresponding to various sintering blending schemes. This can be performed according to the following steps:

[0072] Step S11: Obtain chemical composition, burn-off rate and particle size distribution data of various raw materials and fuels as metallurgical performance parameters;

[0073] The raw materials and fuels used in sintering refer to the mineral powder raw materials, sintering flux, and fuel used in the sintering production process. Their physicochemical properties directly determine the quality and performance of the sintered ore. In this step, specific data on the chemical composition, burn-off rate, and particle size distribution of the raw materials, sintering flux, and fuel are obtained as metallurgical performance parameters.

[0074] Among these parameters, chemical composition refers to the proportion of various elements and compounds in the raw materials (such as the content of Fe, SiO2, CaO, MgO, Al2O3, FeO, etc.), which is the basis for determining the thermodynamic conditions of mineral phase transformation and liquid phase formation during the sintering process. Burn-off rate refers to the proportion of mass loss of raw materials due to volatilization, decomposition, oxidation, and other reactions during high-temperature sintering, usually expressed as the percentage difference in mass before and after burning. Particle size distribution refers to the particle size composition and distribution characteristics of the raw materials, commonly expressed as the mass percentage of different particle size ranges (such as <3mm, 3-5mm, 5-10mm, etc.), reflecting the uniformity of particle size distribution. By quantifying the metallurgical performance parameters of each raw material, it is possible to analyze the liquid phase generation, strength, and other properties of sintered ore under different proportions, avoiding the failure of ore blending schemes due to differences in raw material characteristics.

[0075] Step S12: Collect the strength data of finished sinter corresponding to the various raw material and fuel ratios during the sintering production process;

[0076] The raw material and fuel ratio refers to the mass proportion of raw materials such as iron ore (e.g., hematite, magnetite), flux (e.g., limestone, dolomite), and fuel (e.g., coke powder, anthracite) used in sintering production. For example, iron ore powder A accounts for 40%, iron ore powder B accounts for 30%, limestone accounts for 15%, and coke powder accounts for 5%.

[0077] The strength data of finished sinter refers to the quantitative index characterizing the crush resistance of sinter. It is usually quantified by indicators such as drum strength (TI, such as TI6.3, which means the proportion of particles with a diameter > 6.3 mm), compressive strength, or drop strength.

[0078] Each raw material and fuel ratio is mapped one-to-one with the strength index of the sintered ore in the corresponding production batch, forming a dataset of ratio combination-strength value. Taking the strength value as the drum strength as an example, ratio scheme X corresponds to a drum strength of 78%, and ratio scheme Y corresponds to a drum strength of 80%, providing original data support for subsequent analysis of the impact of ratio changes on strength.

[0079] Step S13: Calculate the amount of liquid phase generated in the blended ore based on the raw material and fuel ratio and the corresponding metallurgical performance parameters.

[0080] In the sintering process, based on the mass ratio of the raw materials and their metallurgical characteristic parameters, the theoretical amount of high-temperature liquid phase generated during sintering is quantitatively predicted using metallurgical thermodynamic models or empirical formulas.

[0081] The amount of liquid phase generated is a core factor determining the strength of sinter. Accurate calculation can establish a direct correlation between it and strength. By converting the raw material ratio and reaction characteristics into quantifiable liquid phase generation indicators, and transforming the chemical composition and physical properties of the raw materials into process parameters, key theoretical calculation support is provided for subsequent back-calculation of liquid phase amount constraints and generation of optimized ore blending schemes based on target strength.

[0082] Step S14: Organize the data on the amount of mixed ore liquid phase generated and the corresponding strength data of finished sintered ore to generate the original dataset;

[0083] The theoretical calculation parameters (the amount of mixed ore liquid phase generated) under different ore blending schemes are systematically correlated with the actual production indicators (the strength of the finished sinter) to form a standardized data set.

[0084] By establishing a data pair mapping between ore blending scheme, liquid phase quantity, and intensity, key parameters in the metallurgical process are transformed into calculable and analyzable quantitative data, providing a foundation for the establishment of subsequent prediction models.

[0085] Step S15: Build a prediction model based on the original dataset.

[0086] Based on historical data on the relationship between the amount of liquid phase generated in the sintering process and the strength of the finished sinter, machine learning algorithms or statistical analysis methods are used to explore the nonlinear correlation between the amount of liquid phase generated and the strength of the sinter, forming a quantitative mapping relationship between the input amount of liquid phase and the output predicted strength. A mathematical model with mapping prediction function is then constructed as a prediction model.

[0087] The prediction model construction method provided in this embodiment is based on the metallurgical performance parameters of raw materials and fuels. It transforms the thermodynamic conditions and kinetic characteristics of liquid phase generation during sintering into quantifiable calculation parameters. By collecting intensity data corresponding to multiple ore blending schemes and associating them with the amount of liquid phase generation, the amount of liquid phase generation is used as a key intermediate variable connecting raw material characteristics and product performance during data processing and model building. Through the construction of structured datasets, high-quality training samples are provided for machine learning algorithms, enabling the model to accurately capture the nonlinear influence of the amount of liquid phase generation on intensity.

[0088] The prediction model ultimately formed by this method not only inherits the theoretical depth of metallurgical processes but also possesses the prediction accuracy of data science. It provides a reliable quantitative tool for subsequent optimization design of ore blending schemes by back-calculating liquid phase quantity constraints based on target strength, significantly improving the scientificity, adaptability, and engineering application value of sinter strength prediction.

[0089] Example 3:

[0090] To reduce experimental calculations, improve data acquisition efficiency, and lower experimental costs, this embodiment proposes that step S13 in the above embodiment two, which calculates the amount of liquid phase generated in the mixed ore based on the raw material and fuel ratio and the corresponding metallurgical performance parameters, can specifically calculate the amount of liquid phase generated in the mixed ore based on the content of adhering particles, burn-off rate, and dry material amount in the raw material and fuel ratio.

[0091] The amount of liquid phase generated from the raw material is directly calculated by taking into account the mass content of adhering particles with a particle size of 0 to 1 mm, the burn-off rate, and the dry material quantity. The reactivity of fine-particle materials is quantified by the adhering particle content. Compared with the traditional calculation method based solely on chemical composition, this method more accurately captures the influence of particle size distribution on liquid phase generation kinetics. Furthermore, the burn-off rate is combined to correct for the mass loss in the high-temperature reaction of the raw material, so that the calculation of liquid phase generation conforms to the principle of material balance. Using the dry material quantity as the measurement basis can eliminate the interference of moisture fluctuations on the calculation basis.

[0092] This calculation method requires no additional testing costs, ensuring the rigor of the metallurgical mechanism of the calculation model while improving the real-time performance and adaptability of engineering applications, providing a quantitative basis for the optimization of ore blending schemes that combines theoretical depth and industrial applicability.

[0093] This embodiment further proposes a calculation formula for the liquid phase generation of the blended ore based on the content of adhering particles, burn-off rate and dry material amount in the raw material and fuel ratio, as shown in Formula 1 below.

[0094] Formula 1:

[0095] in, This represents the amount of mineral liquid phase generated during homogenization. This refers to the dry weight of each raw material. The content of 0 to 1 mm particles in the raw materials. This represents the burn loss rate of each raw material in the raw material and fuel blend.

[0096] This formula uses the content of 0-1mm particles ( As a quantitative indicator of the activity of adhering particles, it accurately anchors the nucleation-melting core role of fine-grained materials during sintering. Particles in this size range, due to their large specific surface area and high interfacial energy, easily form low-melting-point eutectics, making characterization more mechanistically targeted than that of a wider particle size range; through... The burn-off rate correction term incorporates the mass loss caused by high-temperature decomposition of raw materials (such as CO2 escape from carbonates and removal of water of crystallization) into the material balance calculation, converting the nominal dry material quantity into the actual effective quantity participating in the liquid phase reaction, thus avoiding overestimation or underestimation of the liquid phase quantity due to burn-off errors; based on the dry material quantity ( Using this as the measurement benchmark, the interference of raw material moisture fluctuations on the proportion calculation is eliminated. Moreover, all parameters (particle size sieving data, burn loss rate test value, dry material ratio) are routine testing items in sintering production, requiring no additional sensor investment. Through the mathematical expression of linear summation, it not only conforms to the metallurgical essence of the synergistic effect of multi-component raw materials, and the liquid phase generation of each raw material is additive, but also facilitates real-time calculation by embedding it into the industrial control system. It provides a direct quantitative tool for optimizing the proportion of fine-grained materials and controlling the amount of high burn loss raw materials in the ore blending scheme, so that the prediction of liquid phase generation and production practice form a closed-loop feedback.

[0097] Example 4:

[0098] The above embodiments do not limit the specific model type of the prediction model. In order to fully fit the nonlinear metallurgical mechanism of liquid phase generation and strength formation during sintering, this embodiment proposes that a nonlinear regression model can be used to establish the prediction model.

[0099] Specifically, step S15, which builds a prediction model based on the original dataset, can be performed according to the following steps:

[0100] Step S151: Fit the relationship between liquid phase generation amount and intensity based on a nonlinear regression model;

[0101] When the amount of liquid phase is insufficient, the strength increases exponentially with the amount of liquid phase generated (due to weak interparticle bonding caused by insufficient binder phase). However, when the amount of liquid phase exceeds a critical value, the strength growth slows down, and overmelting may even lead to a porous structure. During sintering, the relationship between liquid phase generation and strength forms a nonlinear metallurgical mechanism. Nonlinear models can accurately characterize this type of curve relationship using polynomials, exponential functions, etc., avoiding the mechanism distortion and prediction bias caused by linear models forcibly fitting nonlinear data.

[0102] Nonlinear fitting can capture higher-order interaction effects in data, such as the combined influence of the synergistic effect of different raw material liquid phases on intensity. Compared with linear models that only describe a single linear relationship, it can significantly improve the goodness of fit in complex industrial scenarios (e.g., R²). 2 (Value increases by 15%-20%), especially suitable for the quantitative characterization of nonlinear coupling effects in multi-group ore allocation schemes.

[0103] Step S152: Train and validate the nonlinear regression model based on the original dataset;

[0104] Step S153: Use the nonlinear regression model that meets the accuracy requirements after verification as the prediction model.

[0105] The model training process can optimize nonlinear parameters through mechanisms such as cross-validation, and can adapt to the strength response characteristics under different raw material structures. In this embodiment, the training and validation process of the model is not limited. You can refer to the introduction of relevant implementation methods, which will not be repeated here.

[0106] Before training and validating the nonlinear regression model based on the original dataset, data cleaning and standardization can be performed on the original dataset to ensure the reliability and generalization ability of the prediction model. Data cleaning removes outliers and repairs missing values ​​through methods such as quantile truncation and interpolation completion, avoiding interference from abnormal data on model training. For example, it prevents a set of erroneous liquid phase quantity data from causing the intensity prediction curve to shift, ensuring that the dataset input to the model conforms to the actual laws of metallurgical production.

[0107] Meanwhile, the dimensions and numerical ranges of the features in the original data differ significantly. Directly inputting these features into the model can cause the gradient descent process to favor features with large values ​​(such as dry matter content) while ignoring small but crucial features (such as the proportion of fine-grained content). Standardization (such as Z-score standardization or Min-Max normalization) can map all features to a uniform scale, enabling the model to capture the influence of each feature on the intensity equally. For example, it can prevent the large value of liquid phase generation from masking the nonlinear contribution of fine-grained content to the intensity.

[0108] Of course, in addition to data cleaning and standardization, other data preprocessing methods can also be used. This embodiment does not limit these methods. The data can be preprocessed according to the actual application scenario. These methods will not be elaborated here.

[0109] Example 5:

[0110] In order to transform the range of mixed mineral liquid phase generation into a quantifiable control boundary, step S103 calls the prediction model to calculate the range of mixed mineral liquid phase generation required to reach the target intensity. The process of generating the constraint conditions for the mixed mineral liquid phase generation can be specifically executed using the following formula 2. First, the upper limit and lower limit of the mixed mineral liquid phase generation required to reach the target intensity are calculated according to the prediction model. Then, based on the upper limit and lower limit, formula 2 is generated as the constraint conditions.

[0111] Formula 2 is as follows:

[0112] Where A is the lower limit and B is the upper limit. The proportions of each raw material, The amount of liquid phase in each raw material.

[0113] Formula 2 is used to convert this range into the proportions of each raw material. ) and contribution of liquid phase formation ( The linear constraints of the strength index enable the reverse derivation of the liquid phase quantity control and raw material ratio optimization.

[0114] Formula 2 transforms the liquid phase quantity constraint into a linear inequality constraint, which can be embedded into a linear programming (LP) or nonlinear programming (NLP) optimization model to solve for the optimal ore blending scheme in conjunction with constraints such as raw material cost and inventory. For example, under the premise of satisfying the liquid phase quantity constraint (Formula 2), the raw material procurement cost can be optimized simultaneously to achieve multi-objective optimization of meeting strength requirements and minimizing costs. Of course, the above embedding is not required, and this embodiment does not impose any restrictions on it.

[0115] Example 6:

[0116] After generating the optimized sintering ore blending scheme, to facilitate user decision-making, a cost analysis can be further performed on the optimized sintering ore blending scheme to generate a cost analysis report; at the same time, the performance of the finished sintered ore can be predicted based on the optimized sintering ore blending scheme to generate a performance evaluation report; then the optimized sintering ore blending scheme, along with the corresponding cost analysis report and performance evaluation report, can be output simultaneously.

[0117] The core content of a cost analysis report may include, but is not limited to: raw material cost quantification, cost structure breakdown, cost comparison, and optimization value. Through cost analysis, the economic feasibility of a solution is verified, providing production managers with intuitive cost data to help determine the economic viability of different formulation schemes, thereby balancing the conflict between performance requirements and cost control.

[0118] The core content of a performance evaluation report may include, but is not limited to: quantitative prediction of sinter performance indicators, prediction methods and models, performance comparison, and risk assessment. Performance prediction, by quantitatively predicting key indicators of sinter, ensures that the optimized scheme meets the production needs of the blast furnace while also identifying potential quality risks in the ore blending scheme in advance. Furthermore, it can be used to guide the secondary optimization of the ore blending scheme; for example, when the performance prediction fails to meet the standards, the raw material ratio can be adjusted in reverse.

[0119] This embodiment generates an optimized sintering ore blending scheme, then uses a cost analysis report to quantify the entire cost chain, including raw material input and energy consumption, and dynamically evaluates the scheme's economic viability based on real-time price data. A performance evaluation report pre-verifies the physical properties (such as drum strength) and metallurgical indicators (such as reducibility) of the sinter. Based on historical data and mechanistic models, it simulates the impact of raw material fluctuations on quality, proactively mitigating risks such as insufficient strength and excessive composition. Finally, through the synergistic output of dual reports and the optimized scheme, the traditional post-event verification of experience-based ore blending is transformed into data-driven pre-event control. This not only enhances the cost competitiveness and quality reliability of the ore blending scheme but also provides parameter traceability support for smooth blast furnace operation, promoting the upgrade of sintering production towards digital and intelligent decision-making.

[0120] Example 7:

[0121] Further reference Figure 2The diagram illustrates an exemplary structural block diagram of an apparatus for sintering and blending based on the amount of mixed ore liquid phase generated according to an embodiment of the present application. The apparatus mainly includes a model building module, an index determination module, a constraint calculation module, and a scheme generation module. The apparatus for sintering and blending based on the amount of mixed ore liquid phase generated adopts a modular design and achieves intelligent sintering and blending through four core units.

[0122] Among them, the model building module is used to establish a predictive model of the amount of mixed ore liquid phase generated and the strength of finished sinter based on the strength data of finished sinter corresponding to various sintering blending schemes.

[0123] The indicator determination module is used to determine the intensity indicator to be achieved, which serves as the target intensity.

[0124] The constraint calculation module is used to call the prediction model to calculate the range of mixed mineral liquid phase generation required to achieve the target strength, and to generate the constraint conditions for the mixed mineral liquid phase generation.

[0125] The scheme generation module is used to generate optimized sintering ore blending schemes based on constraints.

[0126] In the system meltdown processing device provided in this embodiment, the model building module, based on historical data from multiple ore blending schemes, deeply integrates metallurgical mechanisms with data-driven algorithms to establish a predictive model that can accurately capture the nonlinear relationship between liquid phase generation and intensity. The index determination module transforms production requirements into specific target intensity values ​​and, through linkage with downstream blast furnace ironmaking process parameters, enables the ore blending scheme to directly serve the optimization of the entire production process. The constraint calculation module uses the predictive model to infer the upper and lower limits of liquid phase generation, transforming the abstract intensity target into quantifiable process constraints. Under the constraint of liquid phase quantity, the scheme generation module, combined with multi-dimensional constraints such as raw material costs and inventory status, quickly searches for the optimal blending ratio combination through linear programming or intelligent algorithms. This device transforms the traditional sintering ore blending from an extensive mode relying on manual experience to a data-driven precise control mode, providing core technical support for steel enterprises to improve quality and reduce costs.

[0127] It should be noted that the system fuse-breaking device described in this embodiment can be referred to in conjunction with the system fuse-breaking method described in the above embodiments, and the repeated parts will not be described again in this embodiment.

[0128] Example 8:

[0129] The following is for reference. Figure 3 , Figure 3 A schematic diagram of the structure of a computer system suitable for implementing the server of the present application is shown.

[0130] like Figure 3As shown, the computer system includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 302 or programs loaded from storage section 308 into random access memory (RAM) 303. RAM 303 also stores various programs and data required for the system's operating instructions. CPU 301, ROM 302, and RAM 303 are interconnected via bus 304. Input / output (I / O) interface 305 is also connected to bus 304.

[0131] The following components are connected to I / O interface 305: an input section 306 including a keyboard, mouse, etc.; an output section 307 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN card, modem, etc. The communication section 309 performs communication processing via a network such as the Internet. Drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0132] Specifically, according to embodiments of this application, the flowchart above refers to... Figure 1 The described process can be implemented as a computer software program. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowchart. In such an embodiment, the computer program contains program code for performing the methods shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by central processing unit (CPU) 301, it performs the functions defined in the system of this application.

[0133] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operational instructions of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two connected blocks may actually be executed substantially in parallel, or they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified functions or operational instructions, or using a combination of dedicated hardware and computer instructions.

[0134] The units or modules described in the embodiments of this application can be implemented in software or hardware. The described units or modules can also be located in a processor. The names of these units or modules do not, in certain circumstances, constitute a limitation on the unit or module itself.

[0135] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the foregoing disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A method for sintering ore blending based on the amount of mixed ore liquid phase generated, characterized in that, include: Based on the strength data of finished sinter corresponding to various sintering blending schemes, a predictive model for the amount of mixed ore liquid phase generated and the strength of finished sinter is established. Identify the intensity indicators to be achieved as the target intensity; The prediction model is invoked to calculate the range of mixed mineral liquid phase generation required to achieve the target intensity, and constraints on the amount of mixed mineral liquid phase generation are generated. An optimized sintering ore blending scheme is generated based on the aforementioned constraints; The step of establishing a predictive model for the relationship between the amount of mixed liquid phase generated and the strength of the finished sinter based on the strength data of the finished sinter corresponding to various sintering blending schemes includes: Acquire data on the chemical composition, burn-off rate, and particle size distribution of various raw materials and fuels as metallurgical performance parameters; Collect data on the strength of finished sintered ore corresponding to various raw material and fuel ratios during the sintering production process; Based on the raw material and fuel ratio and the corresponding metallurgical performance parameters, calculate the amount of liquid phase generated in the blended ore; Organize the data on the amount of mixed ore liquid phase generated and the corresponding strength data of finished sintered ore to generate the original dataset; A prediction model is built based on the original dataset; The calculation of the liquid phase generation amount of the blended ore based on the raw material and fuel ratio and the corresponding metallurgical performance parameters includes: The amount of liquid phase generated in the blended ore is calculated based on the content of adhering particles, the burn loss rate, and the dry material amount in the raw material and fuel ratio. The calculation of the liquid phase generation of the blended ore based on the content of adhering particles, burn-off rate, and dry material quantity in the raw material and fuel ratio includes: Calculate the amount of liquid phase generated in the homogenized ore according to Formula 1; Formula 1: in, This represents the amount of mineral liquid phase generated during homogenization. This refers to the dry weight of each raw material. The content of 0 to 1 mm particles in the raw materials. The burn loss rate of each raw material in the stated raw material and fuel ratio; The step of building a prediction model based on the original dataset includes: The relationship between liquid phase generation and intensity was fitted using a nonlinear regression model. The nonlinear regression model is trained and validated based on the original dataset. The nonlinear regression model that meets the accuracy requirements after verification is used as the prediction model; The step of calling the prediction model to calculate the range of mixed mineral liquid phase generation required to achieve the target intensity, and generating constraints on the mixed mineral liquid phase generation, includes: The upper and lower limits of the amount of mixed mineral liquid phase required to achieve the target strength are calculated based on the prediction model. Based on the upper limit and the lower limit, Formula 2 is generated as a constraint condition; Formula 2: Where A is the lower limit value and B is the upper limit value. The proportions of each raw material, The amount of liquid phase in each raw material.

2. The method as described in claim 1, characterized in that, Before training and validating the nonlinear regression model based on the original dataset, the method further includes: Perform data cleaning on the original dataset; The cleaned dataset is then standardized.

3. The method as described in claim 1, characterized in that, After generating the optimized sintering ore blending scheme based on the constraints, the method further includes: A cost analysis is performed on the optimized sintering ore blending scheme, and a cost analysis report is generated. The performance of the finished sinter is predicted based on the optimized sintering ore blending scheme, and a performance evaluation report is generated. Output the optimized sintering ore blending scheme, along with the corresponding cost analysis report and performance evaluation report.

4. An apparatus for sintering blending based on the amount of mixed ore liquid phase generated, characterized in that, The method for sintering and blending based on the amount of homogenized liquid phase generated according to claim 1 includes: The model building module is used to establish a predictive model of the amount of mixed ore liquid phase generated and the strength of finished sinter based on the strength data of finished sinter corresponding to various sintering blending schemes. The indicator determination module is used to determine the intensity indicator to be achieved, which serves as the target intensity. The constraint calculation module is used to call the prediction model to calculate the range of mixed mineral liquid phase generation required to achieve the target intensity, and to generate the constraint conditions for the mixed mineral liquid phase generation. The scheme generation module is used to generate an optimized sintering ore blending scheme based on the constraints.

5. A server, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method as described in any one of claims 1 to 3.

Images