A method and system for sintering control of a manganese-zinc ferrite magnetic ring

Abstract

The application provides a manganese-zinc ferrite magnetic ring sintering control method and system, relates to the technical field of magnetic ceramic material preparation, and the method comprises the following steps: collecting real-time detection data and historical detection data of each collection area in a sintering furnace, constructing a temperature-oxygen partial pressure coupling model, training the temperature-oxygen partial pressure coupling model by using the historical detection data, inputting the real-time detection data into the trained temperature-oxygen partial pressure coupling model, and outputting a prediction result; judging whether the prediction result is consistent with the data range in a preset temperature-oxygen partial pressure reference table; if yes, no processing is performed; and if no, the prediction result is used for regulation and control. The temperature and oxygen partial pressure of each area in the sintering furnace are predicted by constructing the temperature-oxygen partial pressure coupling model, and dynamic adjustment is performed in a timely manner, so that the stability in the magnetic ring sintering process is improved, and the manganese-zinc ferrite magnetic ring defective rate is reduced.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of preparation of magnetic ceramic materials, in particular to a manganese-zinc ferrite magnetic ring sintering control method and system. BACKGROUND

[0002] Under the background of rapid development of modern electronic information technology, manganese-zinc ferrite magnetic rings, as key components in electronic devices, are widely used in power transformers, inductors and electromagnetic interference filters, etc., and play an important role in improving the performance, stability and efficiency of the devices. The manganese-zinc ferrite magnetic ring is made of Fe2O3, MnO and ZnO as main raw materials, and is sintered by high-temperature solid-phase reaction. The formation of its magnetic properties is closely related to the microstructure, and the sintering process determines the grain growth, pore removal and element distribution uniformity, which is a key link affecting the performance of the product.

[0003] In the prior art, the sintering of manganese-zinc ferrite magnetic rings usually adopts a temperature and oxygen partial pressure independent control method: the temperature is realized by PID control of heating power to achieve programmed temperature rise, and the oxygen partial pressure is adjusted by fixed gas proportion to adjust the atmosphere. Although the temperature and atmosphere can be controlled macroscopically, there is a lack of synergistic regulation mechanism. In the sintering process, temperature changes will affect the oxidation-reduction balance and the effective distribution of oxygen partial pressure, while oxygen partial pressure fluctuations will react on the material reaction rate and local thermal field, making it difficult for the independent control system to maintain stable balance. SUMMARY

[0004] In order to improve the accuracy of temperature and oxygen partial pressure synergistic control in the sintering process of manganese-zinc ferrite magnetic rings, the application provides a manganese-zinc ferrite magnetic ring sintering control method and system.

[0005] In the first aspect, the application provides a manganese-zinc ferrite magnetic ring sintering control method, which adopts the following technical scheme:

[0006] A manganese-zinc ferrite magnetic ring sintering control method, the method comprising:

[0007] Collecting data: temperature sensors and oxygen partial pressure probes are arranged in each collection area in the sintering furnace, and real-time detection data and historical detection data of each collection area are collected, the detection data including temperature, oxygen partial pressure and process control data; the process control data including heating current, heating voltage, gas inlet flow and exhaust flow;

[0008] Modeling: a temperature-oxygen partial pressure coupling model is constructed based on the heat balance equation and the gas conservation equation;

[0009] Training the model: According to the preset temperature zone setting, set the temperature zone label, including low temperature label, medium temperature label, high temperature label and cooling label, label the temperature zone label for the historical detection data, input the labeled historical detection data into the temperature oxygen partial pressure coupling model for model training, and obtain the trained temperature oxygen partial pressure coupling model;

[0010] Predicting temperature and oxygen partial pressure: labeling the real-time detection data with the temperature zone label and inputting the trained temperature oxygen partial pressure coupling model, and outputting the prediction result of each collection area;

[0011] Judging the result: according to the preset temperature oxygen partial pressure reference table of each collection area, judging whether the prediction result is consistent with the data range in the corresponding temperature oxygen partial pressure reference table:

[0012] If yes, no processing is performed;

[0013] If not, adjust the temperature and oxygen partial pressure of the sintering furnace according to the prediction result.

[0014] By adopting the above technical scheme, the temperature oxygen partial pressure coupling model is constructed to predict the temperature and oxygen partial pressure changes of each area of the sintering furnace and to judge whether it deviates from the reference table. When the prediction result deviates from the reference table, dynamic adjustment is performed in time, the accurate and cooperative control of the temperature and oxygen partial pressure in the magnetic ring sintering process is realized, the consistency of the magnetic ring density and magnetic performance is improved, the stability in the magnetic ring sintering process is improved, and the rate of defective products of the manganese-zinc ferrite magnetic ring is reduced.

[0015] Optionally, when the step of collecting data is performed, before the temperature sensor and the oxygen partial pressure probe are arranged in each collection area in the sintering furnace, the method further comprises:

[0016] Obtaining initial data: based on the volume in the sintering furnace, the sintering furnace is divided into a plurality of unit areas in equal volume, and initial data of each unit area is collected, the initial data including initial temperature data and initial oxygen partial pressure data;

[0017] Obtaining temperature oxygen partial pressure sensitivity distribution map: according to the geometric structure of the sintering furnace and the initial data of each unit area, a temperature oxygen partial pressure sensitivity distribution map is established;

[0018] Unit area merging: based on the temperature oxygen partial pressure sensitivity distribution map, the sensitivity difference of adjacent unit areas is calculated, and it is judged whether the sensitivity difference between adjacent unit areas is less than a preset sensitivity threshold:

[0019] If yes, the two unit areas corresponding to the sensitivity difference are merged, and the merged unit area is recorded as a collection area of the sintering furnace;

[0020] If not, both of the two unit areas corresponding to the sensitivity difference are recorded as the collection area of the sintering furnace.

[0021] Optionally, after the step of performing unit region merging, further comprising:

[0022] Constructing a feature vector: obtaining the temperature change curve and the oxygen partial pressure change curve of each collection region, and constructing a feature vector of each collection region;

[0023] Region clustering: calculating the Euclidean distance between each collection region using the feature vector of each collection region, clustering based on the calculated Euclidean distance using the K-means clustering method, obtaining collection regions labeled with cluster tags, and merging collection regions belonging to the same cluster and adjacent to each other, and taking the merged collection regions as new collection regions.

[0024] By adopting the above technical solution, a temperature-oxygen partial pressure sensitivity distribution map is established based on the sintering furnace geometric structure and initial temperature and oxygen partial pressure data, and the unit regions are merged based on the sensitivity difference, realizing self-adaptive division of the internal regions of the sintering furnace, making the temperature and oxygen partial pressure change characteristics in each collection region more consistent, and the region boundary more consistent with the actual physical field distribution in the furnace. In addition, according to the temperature and oxygen partial pressure dynamic change characteristics, the collection regions are clustered and refined and merged, reducing the calculation amount and data dimension, avoiding repeated calculation, improving the calculation efficiency of region modeling and control parameter adjustment, significantly improving the accuracy and prediction stability of model training, and helping to more quickly and stably perform dynamic adjustment response.

[0025] Optionally, the temperature-oxygen partial pressure coupling model is:

[0026] ;

[0027] wherein, represents the volume-weighted density, represents the volume-weighted specific heat capacity, represents the temperature, represents the time, represents the effective thermal conductivity, represents the chemical reaction heat source term, , represents the reaction enthalpy, represents the volume reaction rate, represents the external heat source, represents the porosity, represents the oxygen molar concentration, represents the gas phase velocity, represents the effective oxygen diffusion coefficient, represents the gas phase consumption rate, , represents the stoichiometric coefficient, i.e., the number of moles of oxygen consumed per mole of reaction, represents the gas phase inlet / discharge source term, represents the reaction progress, represents the molar concentration of the reactant, represents the specific surface area per unit volume of the reactant, represents the reaction rate at the stage, represents the oxygen partial pressure, represents the gas constant.

[0028] By adopting the above technical solution, the energy conservation and gas phase conservation equation are introduced, and factors such as heat conduction, reaction heat release, atmosphere diffusion and chemical reaction kinetics are comprehensively considered, so that the dynamic coupling relationship between the temperature field and the oxygen partial pressure field in the sintering process can be accurately described. In addition, the mechanism of the mutual influence of heat conduction and redox reaction is quantified, and the local temperature change, oxygen partial pressure distribution and reaction rate are accurately predicted, which provides a scientific basis for subsequent control regulation, and significantly improves the stability and controllability of the magnetic ring sintering process.

[0029] Optionally, the sintering stage in the sintering furnace includes a low-temperature pre-sintering stage, a medium-temperature stage, a high-temperature sintering stage and a cooling stage, and the stage reaction rate includes a low-temperature pre-sintering and cooling reaction rate, a medium-temperature stage rate and a high-temperature sintering reaction rate;

[0030]

[0031] wherein, represents the pre-exponential factor of the low-temperature pre-sintering stage and the cooling stage, represents the pre-exponential factor of the medium-temperature stage, represents the pre-exponential factor of the high-temperature sintering stage, represents the activation energy of the low-temperature pre-sintering stage and the cooling stage, represents the activation energy of the medium-temperature stage, represents the activation energy of the high-temperature sintering stage, represents the oxygen partial pressure reaction order of the low-temperature pre-sintering stage and the cooling stage, represents the oxygen partial pressure reaction order of the medium-temperature stage, represents the oxygen partial pressure reaction order of the high-temperature sintering stage, represents the dependence index of the reaction rate and the reaction progress of the medium-temperature stage, represents the dependence index of the reaction rate and the reaction progress of the high-temperature sintering stage.

[0032] By adopting the technical scheme, the sintering process is divided into a low-temperature pre-sintering stage, a medium-temperature stage, a high-temperature sintering stage and a cooling stage, and pre-exponential factors, activation energies and oxygen partial pressure reaction order numbers are set for the stages respectively, so that the model can accurately reflect different reaction mechanisms and rate characteristics of the stages. In addition, the dependence index parameters of the reaction rate and the reaction progress are introduced in the medium-temperature stage and the high-temperature stage, so that the model not only considers the influence of temperature and oxygen partial pressure, but also reflects the feedback effect of the reaction process on the rate, thereby helping to more truly simulate the kinetic law of the phase change process and improving the accuracy and stability of the temperature and oxygen partial pressure control mechanism.

[0033] Optionally, the prediction result includes a predicted temperature and a predicted oxygen partial pressure, and when the result judging step is performed, the temperature and the oxygen partial pressure of the sintering furnace are adjusted according to the prediction result, including:

[0034] Setting a target trajectory: obtaining a temperature target trajectory and an oxygen partial pressure target trajectory according to a pre-set temperature and oxygen partial pressure reference table;

[0035] Calculating a prediction error: calculating a difference between a predicted temperature of each collection area and a corresponding temperature target trajectory to obtain a temperature error of each collection area; and calculating a difference between a predicted oxygen partial pressure of each collection area and a corresponding oxygen partial pressure target trajectory to obtain an oxygen partial pressure error of each collection area;

[0036] Constructing a cost function: performing weighted summation on the temperature error and the oxygen partial pressure error of each collection area to obtain a weighted error of each collection area, and accumulating the weighted errors of the collection areas to obtain the cost function;

[0037] Variable solving: solving the cost function by using an optimization algorithm to obtain target process control data, and adjusting the temperature and the oxygen partial pressure of the sintering furnace according to the obtained target process control data.

[0038] Optionally, after the step of constructing the cost function is performed, before the step of variable solving is performed, further including:

[0039] Setting a constraint condition: the constraint condition includes a numerical upper and lower limit constraint and a change rate constraint; the numerical upper and lower limit constraint is that the target process control data is within a pre-set control data range; and the change rate constraint is that a change rate of the target process control data within a unit time cannot exceed a pre-set change rate threshold.

[0040] Optionally, when the step of variable solving is performed, the cost function is solved by using a particle swarm algorithm to obtain the target process control data, including:

[0041] Initialize the particle swarm by corresponding each particle to a set of process control data; perform iterative calculation to solve the cost function and update the particle position and velocity until the maximum iteration number is reached, and output a set of process control data, which is taken as the target process control data.

[0042] By adopting the above technical solutions, the target trajectory of temperature and oxygen partial pressure and the cost function construction mechanism are introduced, and the temperature prediction error and the oxygen partial pressure prediction error of each collection area in the sintering process are comprehensively weighted, thereby achieving the technical effect of dynamic balance optimization of temperature and oxygen partial pressure. In addition, the optimization algorithm of numerical constraint and change rate constraint is adopted to solve the cost function, on the basis of ensuring that the change process of the process control data is in a safe and controllable range, the temperature and oxygen partial pressure are cooperatively and adaptively adjusted, so that the temperature and oxygen partial pressure control of the sintering process is more accurate and faster in response speed, the stability and control precision of the sintering furnace operation are improved, and the sintering uniformity and performance consistency of the magnetic ring product are improved.

[0043] In a second aspect, the application provides a manganese-zinc ferrite magnetic ring sintering control system, which is suitable for the control method of any one of the first aspect.

[0044] A manganese-zinc ferrite magnetic ring sintering control system, the system comprising:

[0045] The data acquisition module is used to set temperature sensors and oxygen partial pressure probes in each collection area in the sintering furnace, and to collect real-time detection data and historical detection data of each collection area, wherein the detection data includes temperature, oxygen partial pressure, and process control data; the process control data includes heating current, heating voltage, air inlet flow rate, and air outlet flow rate.

[0046] The model construction module is used to construct a temperature-oxygen partial pressure coupling model based on a heat balance equation and a gas conservation equation.

[0047] The model training module is used to set temperature zone labels according to a preset temperature zone, including a low-temperature label, a medium-temperature label, a high-temperature label, and a cooling label, to label the temperature zone labels for the historical detection data, to input the labeled historical detection data into the temperature-oxygen partial pressure coupling model for model training, and to obtain a trained temperature-oxygen partial pressure coupling model.

[0048] The predicted temperature and oxygen partial pressure module is used to label the temperature zone labels for the real-time detection data and input the labeled real-time detection data into the trained temperature-oxygen partial pressure coupling model, and to output the prediction results of each collection area.

[0049] The result judgment module is used to judge whether the prediction results are consistent with the data range in the corresponding temperature-oxygen partial pressure reference table according to a preset temperature-oxygen partial pressure reference table of each collection area.

[0050] If the prediction result is consistent with the data range in the corresponding temperature-oxygen partial pressure reference table, no processing is performed;

[0051] If the prediction result is inconsistent with the data range in the corresponding temperature-oxygen partial pressure reference table, the temperature and oxygen partial pressure of the sintering furnace are adjusted according to the prediction result.

[0052] In summary, the present application includes at least one of the following beneficial technical effects:

[0053] 1. The present application predicts the temperature and oxygen partial pressure changes in each region of the sintering furnace by constructing a temperature-oxygen partial pressure coupling model and determines whether it deviates from the reference table. When the prediction result deviates from the reference table, timely dynamic adjustment is performed, realizing accurate and coordinated control of the temperature and oxygen partial pressure in the magnetic ring sintering process, improving the consistency of the density and magnetic properties of the magnetic ring, and improving the stability in the magnetic ring sintering process, while reducing the scrap rate of manganese-zinc ferrite magnetic rings.

[0054] 2. The present application establishes a temperature-oxygen partial pressure sensitivity distribution map based on the geometric structure of the sintering furnace and the initial temperature and oxygen partial pressure data, and merges the unit regions based on the sensitivity difference, realizing adaptive division of the internal regions of the sintering furnace, making the temperature and oxygen partial pressure change characteristics inside each collection region more consistent, and the region boundary more consistent with the actual physical field distribution in the furnace.

[0055] 3. The present application introduces the energy conservation and gas phase conservation equation, comprehensively considers factors such as heat conduction, reaction heat release, atmosphere diffusion and chemical reaction kinetics, and can accurately describe the dynamic coupling relationship between the temperature field and the oxygen partial pressure field in the sintering process. In addition, the model quantifies the mechanism of the mutual influence of heat conduction and oxidation-reduction reaction, realizes accurate prediction of local temperature change, oxygen partial pressure distribution and reaction rate, provides a scientific basis for subsequent control and regulation, and significantly improves the stability and controllability of the magnetic ring sintering process.

[0056] 4. The present application divides the sintering process into a low-temperature pre-sintering stage, a medium-temperature stage, a high-temperature sintering stage and a cooling stage, and sets pre-exponential factors, activation energies and oxygen partial pressure reaction orders for each stage, so that the model can accurately reflect the different reaction mechanisms and rate characteristics of each stage. In addition, the dependence index parameters of reaction rate and reaction progress are introduced in the medium-temperature stage and the high-temperature stage, so that the model not only considers the influence of temperature and oxygen partial pressure, but also reflects the feedback effect of reaction progress on rate, thereby helping to more realistically simulate the kinetic laws of the phase change process, and improving the accuracy and stability of the temperature and oxygen partial pressure control mechanism. BRIEF DESCRIPTION OF DRAWINGS

[0057] Fig. 1 is a flowchart of embodiment 1 of the present application;

[0058] Fig. 2This is a flowchart of S11 region division in Embodiment 1 of this application;

[0059] Fig. 3 This is a flowchart of parameter adjustment in S51 of Embodiment 1 of this application. Detailed Implementation

[0060] The following combination Figs. 1 to 3 This application will be described in further detail.

[0061] Example 1: This example discloses a method for controlling the sintering of manganese-zinc ferrite magnetic rings, such as... Fig. 1 As shown, the method includes: collecting real-time and historical detection data from various collection areas within the sintering furnace; constructing a temperature-oxygen partial pressure coupling model; training the temperature-oxygen partial pressure coupling model using historical detection data; inputting real-time detection data into the trained temperature-oxygen partial pressure coupling model; and outputting prediction results. The method then determines whether the prediction results match the data range in a preset temperature-oxygen partial pressure reference table. If yes, no processing is performed; otherwise, adjustments are made based on the prediction results. This embodiment includes the following steps:

[0062] S1 Data Acquisition: This includes S11 region division, S12 real-time detection data acquisition, and S13 historical detection data acquisition. S11 region division includes S111 initial data acquisition, S112 acquiring temperature and oxygen partial pressure sensitivity distribution maps, S113 unit region merging, S114 feature vector construction, and S115 region clustering. Fig. 2 As shown.

[0063] S111 Acquisition of Initial Data: Based on the volume inside the sintering furnace, the sintering furnace is divided into several unit regions, and the initial data of each unit region is collected. The initial data includes initial temperature data and initial oxygen partial pressure data.

[0064] Based on the volume inside the sintering furnace, the sintering furnace is divided into several unit regions. A temperature sensor and an oxygen partial pressure probe are set at the center point of each unit region. Temperature data of each unit region within a preset initial time window are collected as the initial temperature data of each unit region. Oxygen partial pressure data of each unit region within a preset initial time window are collected as the initial oxygen partial pressure data of each unit region.

[0065] S112 Obtaining Temperature and Oxygen Partial Pressure Sensitivity Distribution Map: Based on the geometry of the sintering furnace and the initial data of each unit region, a temperature and oxygen partial pressure sensitivity distribution map is established.

[0066] A coordinate system is constructed based on the geometry of the sintering furnace. Then, based on the initial data of each unit region, the temperature sensitivity and oxygen partial pressure sensitivity of each unit region are calculated using the temperature sensitivity calculation formula and the oxygen partial pressure sensitivity calculation formula, respectively. The comprehensive sensitivity of each unit region is calculated using the comprehensive sensitivity formula. The position of each unit region is mapped to the spatial coordinates of the sintering furnace and the comprehensive sensitivity of each unit region is marked to obtain the temperature and oxygen partial pressure sensitivity distribution map.

[0067] The formula for calculating temperature sensitivity is:

[0068] ;

[0069] in, Indicates the current cell region. This indicates the number of stages in the sintering process. Indicates the number is The sintering stage, Indicates the number is The stage weights of the sintering stage. Indicates coordinates as Temperature gradient mode , The gradient at all time points within a certain stage is taken as the i-th In this embodiment, the percentile is... =90%. In this embodiment, the sintering process of manganese-zinc ferrite magnetic ring is mainly divided into four stages, including low-temperature pre-sintering stage, medium-temperature stage, high-temperature sintering stage and cooling stage. The stage weight of the low-temperature pre-sintering stage is 0.1, the stage weight of the medium-temperature stage is 0.25, the stage weight of the low-temperature and high-temperature sintering stage is 0.5, and the stage weight of the cooling stage is 0.15.

[0070] The formula for calculating oxygen partial pressure sensitivity is:

[0071] ;

[0072] in, Indicates coordinates as oxygen partial pressure gradient mode, .

[0073] The formula for calculating temperature sensitivity is:

[0074] G ;

[0075] in, Indicates the weight of temperature sensitivity. In this embodiment, the oxygen partial pressure sensitivity weight is represented. .

[0076] S113 unit region merging: based on the temperature oxygen partial pressure sensitivity distribution map, the sensitivity difference of adjacent unit regions is calculated, and it is judged whether the sensitivity difference between adjacent unit regions is less than a preset sensitivity threshold:

[0077] If yes, the two unit regions corresponding to the sensitivity difference are merged, and the merged unit region is recorded as a collection region of the sintering furnace;

[0078] If no, both of the two unit regions corresponding to the sensitivity difference are recorded as the collection region of the sintering furnace.

[0079] S114 constructing feature vector: a temperature sensor and an oxygen partial pressure probe are arranged at the center of each collection region, and the temperature data and oxygen partial pressure data of each collection region in a preset time window are collected, so as to obtain the temperature change curve and oxygen partial pressure change curve of each collection region in the preset time window (for example, 5 min), extract the representative features of each sintering stage of each collection region, including temperature features (temperature mean, temperature variance, temperature change rate), oxygen partial pressure features (oxygen partial pressure mean, oxygen partial pressure variance, oxygen partial pressure change rate) and collection region center coordinates, combine the representative features according to the sintering stage, wherein 6 representative features (including temperature mean, temperature variance, temperature change rate, oxygen partial pressure mean, oxygen partial pressure variance and oxygen partial pressure change rate) are extracted for each sintering stage, and 3 dimensions of the collection region center coordinates are added to construct a 27-dimensional feature vector of each collection region. Then, the 27-dimensional feature vector of each collection region is normalized to obtain the feature vector of each collection region.

[0080] S115 region clustering: based on the feature vector of each collection region, the Euclidean distance between each collection region is calculated, and K-means clustering method is used for clustering to obtain collection regions labeled with cluster labels. The collection regions belonging to the same cluster and adjacent to each other are merged, and the merged collection regions are taken as new collection regions.

[0081] S12 collecting real-time detection data: including collecting real-time temperature, real-time oxygen partial pressure and real-time process control data. Based on the collection regions obtained in S11 region division, temperature sensors and oxygen partial pressure probes are arranged at the center positions of each collection region in the sintering furnace, and the real-time temperature and real-time oxygen partial pressure of each collection region are collected by the temperature sensors and oxygen partial pressure probes. Real-time process control data includes real-time heating current, real-time heating voltage, real-time air flow and real-time exhaust flow. The real-time heating current, real-time heating voltage, real-time air flow and real-time exhaust flow collected at the same time are taken as a group of real-time process control data.

[0082] S13 collects historical detection data: including collecting historical temperature, historical oxygen partial pressure and historical process control data. The historical process control data includes historical heating current, historical heating voltage, historical inlet gas flow and historical exhaust gas flow. The historical heating current, historical heating voltage, historical inlet gas flow and historical exhaust gas flow collected at the same time are taken as a set of historical process control data.

[0083] S2 constructs a model: a temperature-oxygen partial pressure coupling model is constructed based on a heat balance equation and a gas phase conservation equation. Specifically, the temperature-oxygen partial pressure coupling model is:

[0084]

[0085] wherein, represents the volume-weighted density, represents the volume-weighted specific heat capacity, represents the temperature, represents the time, represents the effective thermal conductivity, represents the chemical reaction heat source term, , represents the reaction enthalpy, represents the volume reaction rate, represents the external heat source, represents the porosity, represents the oxygen molar concentration, represents the gas phase velocity, represents the effective oxygen diffusion coefficient, represents the gas phase consumption rate, , represents the stoichiometric coefficient, i.e. the number of moles of oxygen consumed per mole of reaction, represents the gas phase inlet / discharge source term, represents the reaction progress, represents the reactant molar concentration, represents the specific surface area of the reactant per unit volume, represents the stage reaction rate, represents the oxygen partial pressure, represents the gas constant, .

[0086] The sintering stage in the sintering furnace includes a low-temperature pre-sintering stage, a medium-temperature stage, a high-temperature sintering stage and a cooling stage, and the stage reaction rate includes low-temperature pre-sintering and cooling reaction rates, medium-temperature stage rates and high-temperature sintering reaction rates.

[0087] The stage reaction rate formula is:

[0088]

[0089] in, The pre-index factor represents the difference between the low-temperature preheating stage and the cooling stage. The pre-index factor represents the temperature range during the intermediate temperature phase. The pre-index factor represents the high-temperature sintering stage. This indicates the activation energy during the low-temperature pre-firing stage and the cooling stage. This represents the activation energy at the mesophilic stage. This represents the activation energy during the high-temperature sintering stage. This indicates the oxygen partial pressure reaction order during the low-temperature pre-calcination stage and the cooling stage. This indicates the oxygen partial pressure reaction order in the mesophilic phase. This indicates the oxygen partial pressure reaction order during the high-temperature sintering stage. This indicates the reaction rate and reaction extent during the mesophilic phase. Dependence index, Indicates the reaction rate and reaction progress during the high-temperature sintering stage. Dependence index.

[0090] In this embodiment, temperature is collected by a set temperature sensor. Oxygen partial pressure is collected by a set oxygen partial pressure probe. Gas phase velocity can be obtained through an anemometer or flow meter. The intake air flow rate is collected by a flow meter. and outlet gas flow rate Obtain the oxygen partial pressure at the air inlet. Oxygen partial pressure in the sampling area where the exhaust port is located and the volume of the sintering furnace According to the formula Calculate the gas phase inlet / emission source terms. Estimate reactant porosity using mercury pressure, CT scans, or image processing. The material density of the reactants was obtained by consulting the materials. Molar mass of reactants Specific heat capacity of reactants Gas constant Oxygen molar concentration Oxygen specific heat capacity Two-body diffusion coefficient of oxygen in free space Using the formula calculate Using the formula Calculate the molar concentration of reactants The specific surface area of ​​the reactants was obtained using the BET nitrogen adsorption test. Using the formula Calculate the specific surface area per unit volume of the reactants. Using formulas Calculating oxygen effective diffusion coefficient wherein, denotes the topological coefficient, , The effective thermal conductivity coefficient is obtained by laser transient method or steady-state thermal conductivity meter The reaction enthalpy is obtained by differential scanning calorimetry or from thermochemical table or by DSC / TGA coupled measurement The number of moles of oxygen consumed per mole of reaction is determined according to the reaction equation during sintering The above various parameters obtained and calculated according to the query of relevant information are taken as known parameters.

[0091] S3 training model: according to the preset temperature zone division, set the temperature zone label, including low temperature label, medium temperature label, high temperature label and cooling label, label the temperature zone label for the historical detection data, input the labeled historical detection data into the temperature oxygen partial pressure coupling model for model training, and obtain the trained temperature oxygen partial pressure coupling model.

[0092] Specifically, since the temperature ranges of the four stages in the sintering process are different, four preset temperature zones are set according to the temperature ranges of each stage, temperature zone labels are set based on the preset temperature zones, including low temperature label, medium temperature label, high temperature label and cooling label, and the obtained historical detection data is labeled with temperature zone label, the labeled historical detection data and the known parameters in the S2 model are input as input, and the temperature oxygen partial pressure coupling model is input for forward solving to obtain a prediction value, the residual error between the prediction value and the true value of the input is calculated, and the parameters are updated using gradient method according to the calculated residual error, and then iterative training is performed until the residual error is less than a preset threshold or the maximum number of iterations is reached, the model training is completed, and the trained temperature oxygen partial pressure coupling model is obtained, which contains 、 、 、 、 、 、 、 、 、 and the parameter values of

[0093] S4 prediction of temperature and oxygen partial pressure: input the real-time detection data of different regions in the sintering furnace and the known parameters in the S2 model into the trained temperature oxygen partial pressure coupling model, and output the prediction results of each collection region, wherein the prediction results include predicted temperature and predicted oxygen partial pressure.

[0094] S5 result judgment: by consulting data, the ideal temperature value range and ideal oxygen partial pressure value range of the manganese zinc ferrite magnetic ring in the sintering process in each collection area of the sintering furnace are obtained, and are presented in table form, to obtain a temperature oxygen partial pressure reference table of each collection area in the sintering furnace. The temperature oxygen partial pressure reference table includes combustion time, combustion stage, ideal temperature value range and ideal oxygen partial pressure value range. The combustion time in the temperature oxygen partial pressure reference table in this embodiment refers to a certain combustion time period, which is a time range interval, not a single time point. According to the pre-set temperature oxygen partial pressure reference table of each collection area, it is judged whether the obtained prediction results of each collection area are consistent with the data range in the temperature oxygen partial pressure reference table of the corresponding area, that is, whether the predicted temperature and predicted oxygen partial pressure of each collection area are within the ideal temperature value range and ideal oxygen partial pressure value range of the corresponding combustion time:

[0095] If yes, no processing is performed;

[0096] If no, S51 parameter adjustment is performed.

[0097] S51 parameter adjustment: adjusting the temperature and oxygen partial pressure of the sintering furnace according to the prediction results of each collection area, including S511 setting target trajectory, S512 calculating prediction error, S513 constructing cost function, S514 setting constraint condition and S515 variable solving, as shown in Fig. 3

[0098] S511 setting target trajectory: according to the pre-set temperature oxygen partial pressure reference table, taking the middle value of the ideal temperature value range and the middle value of the ideal oxygen partial pressure value range of each combustion time as the target temperature and target oxygen partial pressure of each combustion time, collecting the target temperature and target oxygen partial pressure of all combustion times to obtain the temperature target trajectory and oxygen partial pressure target trajectory.

[0099] S512 calculation of prediction error: calculating and collecting the difference between the predicted temperature of each collection area and the target temperature of the corresponding combustion time to obtain the temperature error of each collection area; calculating and collecting the difference between the predicted oxygen partial pressure of each collection area and the target oxygen partial pressure of the corresponding combustion time to obtain the oxygen partial pressure error of each collection area.

[0100] S513 construction of cost function: weighted sum of temperature error and oxygen partial pressure error of each collection area to obtain weighted error of each collection area, and cumulative weighted error of each collection area to obtain cost function.

[0101] The temperature error calculation formula is:

[0102] ;

[0103] The oxygen partial pressure error calculation formula is: ​

[0104] ;

[0105] The temperature and oxygen partial pressure errors in each data collection area were statistically analyzed, and the statistical formulas are as follows:

[0106] ; ;

[0107] The cost function formula is:

[0108] ; .

[0109] in, express Time of the first Predicted temperature for each data collection area express Time of the first The target temperature in the temperature target trajectory of each acquisition area , express Time of the first Predicted oxygen partial pressure in each collection area express Time of the first The target oxygen partial pressure in the oxygen partial pressure target trajectory of each collection area. Indicates the first Weighted error of each data collection area Indicates the temperature error weight. This indicates the weight of the oxygen partial pressure error. Indicates the number of data collection areas.

[0110] S514 sets constraints: Constraints include upper and lower limit constraints on numerical values ​​and rate of change constraints. The upper and lower limit constraints on numerical values ​​ensure that the target process control data remains within the preset control data range. , , , Because the sintering process in the sintering furnace includes a low-temperature pre-sintering stage, a medium-temperature stage, a high-temperature sintering stage, and a cooling stage, the preset control data range for the target process control data will also change at different sintering stages. The rate of change constraint is that the rate of change of the target process control data per unit time must not exceed the preset rate of change threshold, i.e. , , , .

[0111] in, Indicates the target heating current. Indicates the target heating voltage. Indicates the target intake airflow. Indicates the target exhaust flow rate. This represents the minimum value of the heating current. This indicates the maximum value of the heating current. This indicates the minimum value of the heating voltage. This indicates the maximum value of the heating voltage. This represents the minimum intake airflow. This indicates the maximum intake airflow rate. This represents the minimum exhaust flow rate. This indicates the maximum exhaust flow rate. This represents the change in target heating current per unit time. This indicates the change in target heating voltage per unit time. This indicates the change in target airflow rate per unit time. This represents the change in target exhaust flow rate per unit time. This represents the maximum rate of change of heating current per unit time. This represents the maximum rate of change of the target heating voltage per unit time. This represents the maximum rate of change of the target intake airflow per unit time. This represents the maximum rate of change of the target exhaust flow rate per unit time.

[0112] S515 Variable Solving: The cost function is solved using a particle swarm optimization algorithm. The specific steps are as follows: Each particle corresponds to a set of process control data, including heating current, heating voltage, inlet flow rate, and exhaust flow rate. The particle swarm is initialized; the cost function is calculated in each iteration. It updates the particle position and velocity until the maximum number of iterations is reached, outputting the optimal solution, which is a set of process control data. The output process control data is used as the target process control data, and adjustments are made based on the obtained target process control data, such as increasing the heating current to the target heating current.

[0113] The particle position update formula and velocity update formula are as follows:

[0114] ;

[0115] ;

[0116] in, Indicates the first The first particle The position of the next iteration. Indicates the first The first particle The speed of each iteration represents an inertia weight, and represents an acceleration coefficient, and The value range of is generally [1.0, 2.5], and in the embodiment, can be 2.0, and represents a random number, and The value range is [0, 1], represents the optimal solution of the particle itself history, represents the global optimal solution.

[0117] Embodiment 2: The embodiment discloses a sintering control system for a manganese-zinc ferrite magnetic ring, and the system comprises:

[0118] A data acquisition module is configured to set temperature sensors and oxygen partial pressure probes in each collection area in a sintering furnace, and to collect real-time detection data and historical detection data of each collection area. The detection data includes temperature, oxygen partial pressure, and process control data. The process control data includes heating current, heating voltage, air inlet flow rate, and exhaust flow rate.

[0119] A model construction module is configured to construct a temperature-oxygen partial pressure coupling model based on a heat balance equation and a gas phase conservation equation.

[0120] A model training module is configured to set temperature zone labels according to a preset temperature zone, including a low-temperature label, a medium-temperature label, a high-temperature label, and a cooling label, to label the temperature zone labels for the historical detection data, to input the labeled historical detection data into the temperature-oxygen partial pressure coupling model for model training, and to obtain a trained temperature-oxygen partial pressure coupling model.

[0121] A predicted temperature and oxygen partial pressure module is configured to label the temperature zone labels for the real-time detection data and to input the real-time detection data into the trained temperature-oxygen partial pressure coupling model, and to output a prediction result of each collection area.

[0122] A result judgment module is configured to judge whether the prediction result is consistent with a data range in a corresponding temperature-oxygen partial pressure reference table according to a preset temperature-oxygen partial pressure reference table of each collection area.

[0123] If the prediction result is consistent with the data range in the corresponding temperature-oxygen partial pressure reference table, no processing is performed.

[0124] If the prediction result is inconsistent with the data range in the corresponding temperature-oxygen partial pressure reference table, the temperature and oxygen partial pressure of the sintering furnace are adjusted according to the prediction result.

[0125] In the embodiment, the temperature and oxygen partial pressure coupling model is constructed to predict the temperature and oxygen partial pressure changes of each area of the sintering furnace and to determine whether the changes deviate from the reference table. When the prediction result deviates from the reference table, dynamic adjustment is performed in time, the accurate cooperative control of the temperature and oxygen partial pressure in the sintering process of the magnetic ring is realized, the consistency of the density and magnetic performance of the magnetic ring is improved, the stability in the sintering process of the magnetic ring is improved, and the substandard product rate of the manganese-zinc ferrite magnetic ring is reduced.

[0126] The above are preferred embodiments of the present application, and do not limit the protection scope of the present application, so: all equivalent changes made according to the structure, shape, principle of the present application should be covered within the protection scope of the present application.

Claims

1. A method of sintering control of a manganese zinc ferrite magnetic ring, characterized by, The method comprises: Collecting data: temperature sensors and oxygen partial pressure probes are arranged in each collection area in the sintering furnace, real-time detection data and historical detection data of each collection area are collected, and the detection data includes temperature, oxygen partial pressure and process control data; the process control data includes heating current, heating voltage, air inlet flow and air outlet flow; Model construction: a temperature-oxygen partial pressure coupling model is constructed based on a heat balance equation and a gas phase conservation equation; Model training: temperature zone labels are set according to a preset temperature zone, including low-temperature labels, medium-temperature labels, high-temperature labels and cooling labels, the historical detection data are labeled with the temperature zone labels, the labeled historical detection data are input into the temperature-oxygen partial pressure coupling model for model training, and a trained temperature-oxygen partial pressure coupling model is obtained; Prediction of temperature and oxygen partial pressure: the real-time detection data are labeled with the temperature zone labels and input into the trained temperature-oxygen partial pressure coupling model, and a prediction result of each collection area is output; Result judgment: whether the prediction result is consistent with the data range in the corresponding temperature-oxygen partial pressure reference table is judged according to a preset temperature-oxygen partial pressure reference table of each collection area: If yes, no processing is performed; If no, the temperature and oxygen partial pressure of the sintering furnace are adjusted according to the prediction result; When the step of collecting data is performed, before temperature sensors and oxygen partial pressure probes are arranged in each collection area in the sintering furnace, the method further comprises: Obtaining initial data: based on the volume of the sintering furnace, the sintering furnace is divided into a plurality of unit regions in equal volume, and initial data of each unit region is collected, the initial data including initial temperature data and initial oxygen partial pressure data; Obtaining a temperature-oxygen partial pressure sensitivity distribution map: a temperature-oxygen partial pressure sensitivity distribution map is established according to the geometric structure of the sintering furnace and the initial data of each unit region; Unit region merging: based on the temperature-oxygen partial pressure sensitivity distribution map, the sensitivity difference of adjacent unit regions is calculated, and whether the sensitivity difference between adjacent unit regions is less than a preset sensitivity threshold is judged: If yes, the two unit regions corresponding to the sensitivity difference are merged, and the merged unit region is recorded as a collection area of the sintering furnace; If no, both of the two unit regions corresponding to the sensitivity difference are recorded as collection areas of the sintering furnace.

2. The sintering control method of a manganese zinc ferrite magnetic ring according to claim 1, characterized by, After the step of unit region merging is performed, the method further comprises: Constructing a feature vector: obtaining a temperature change curve and an oxygen partial pressure change curve of each collection area, and constructing a feature vector of each collection area; Region clustering: the Euclidean distance between each collection area is calculated using the feature vector of each collection area, the K-means clustering method is used for clustering based on the calculated Euclidean distance, and collection areas labeled with cluster labels are obtained, and adjacent collection areas belonging to the same cluster are merged, and the merged collection area is taken as a new collection area.

3. The sintering control method of a manganese zinc ferrite magnetic ring according to claim 1, characterized by, The temperature-oxygen partial pressure coupling model is: ； wherein, represents the volume-weighted density, represents the volume-weighted specific heat capacity, represents the temperature, represents the time, represents the effective thermal conductivity, represents the chemical reaction heat source term, , represents the reaction enthalpy, represents the volume reaction rate, represents the external heat source, represents the porosity, represents the oxygen molar concentration, represents the gas phase velocity, represents the oxygen effective diffusion coefficient, represents the gas phase consumption rate, , represents the stoichiometric coefficient, i.e. the number of moles of oxygen consumed per mole of reaction, represents the gas phase inlet / exhaust source term, represents the reaction progress, represents the reactant molar concentration, represents the specific surface area per volume of reactant, represents the phase reaction rate, represents the oxygen partial pressure, represents the gas constant.

4. The sintering control method of a manganese zinc ferrite magnetic ring according to claim 3, characterized by, The sintering stage in the sintering furnace includes a low-temperature pre-sintering stage, a medium-temperature stage, a high-temperature sintering stage and a cooling stage, and the stage reaction rate includes a low-temperature pre-sintering and cooling reaction rate, a medium-temperature stage rate and a high-temperature sintering reaction rate; The stage reaction rate is: ； wherein, represents a pre-exponential factor for the low-temperature pre-burning and cooling stages, represents a pre-exponential factor for the medium-temperature stage, represents a pre-exponential factor for the high-temperature sintering stage, represents an activation energy for the low-temperature pre-burning and cooling stages, represents an activation energy for the medium-temperature stage, represents an activation energy for the high-temperature sintering stage, represents an oxygen partial pressure reaction order for the low-temperature pre-burning and cooling stages, represents an oxygen partial pressure reaction order for the medium-temperature stage, represents an oxygen partial pressure reaction order for the high-temperature sintering stage, represents a dependence index of the reaction rate and the reaction progress for the medium-temperature stage, represents a dependence index of the reaction rate and the reaction progress for the high-temperature sintering stage.

5. The sintering control method of a manganese zinc ferrite magnetic ring according to claim 1, characterized by, The prediction result includes a predicted temperature and a predicted oxygen partial pressure, and when the step of result judgment is performed, the temperature and oxygen partial pressure of the sintering furnace are adjusted according to the prediction result, including: Setting target trajectory: obtaining temperature target trajectory and oxygen partial pressure target trajectory according to pre-set temperature oxygen partial pressure reference table; Calculating prediction error: calculating the difference between the predicted temperature and the corresponding temperature target trajectory for each collection area to obtain the temperature error of each collection area; calculating the difference between the predicted oxygen partial pressure and the corresponding oxygen partial pressure target trajectory for each collection area to obtain the oxygen partial pressure error of each collection area; Constructing cost function: weighting and summing the temperature error and the oxygen partial pressure error of each collection area to obtain the weighted error of each collection area, and accumulating the weighted error of each collection area to obtain the cost function; Variable solving: solving the cost function by using an optimization algorithm to obtain the target process control data, and adjusting the temperature and oxygen partial pressure of the sintering furnace according to the obtained target process control data.

6. The sintering control method of a manganese zinc ferrite magnetic ring according to claim 5, characterized by, After performing the step of constructing the cost function and before performing the step of variable solving, further comprising: Setting constraints: the constraints include numerical upper and lower limit constraints and change rate constraints; the numerical upper and lower limit constraints are that the target process control data is within a pre-set control data range; the change rate constraint is that the change rate of the target process control data per unit time cannot exceed a pre-set change rate threshold.

7. The sintering control method of a manganese zinc ferrite magnetic ring according to claim 5, characterized by, When performing the step of variable solving, the particle swarm algorithm is used to solve the cost function to obtain the target process control data, including: Corresponding each particle to a set of process control data, initializing the particle swarm; performing iterative calculation to solve the cost function and update the particle position and velocity until the maximum iteration number is reached, outputting a set of process control data, and outputting the process control data as the target process control data.

8. A manganese zinc ferrite magnetic ring sintering control system, characterized by, The system is suitable for the method of any one of claims 1-7, and the system comprises: Data acquisition module: for setting temperature sensors and oxygen partial pressure probes in each collection area in the sintering furnace, collecting real-time detection data and historical detection data of each collection area, and detecting data including temperature, oxygen partial pressure and process control data; process control data includes heating current, heating voltage, gas flow and exhaust flow; Model construction module: for constructing a temperature-oxygen partial pressure coupling model based on heat balance equation and gas conservation equation; Model training module: for setting temperature zone labels according to pre-set temperature zone, including low temperature label, medium temperature label, high temperature label and cooling label, labeling temperature zone labels for historical detection data, inputting labeled historical detection data into temperature-oxygen partial pressure coupling model for model training, and obtaining trained temperature-oxygen partial pressure coupling model; Temperature and oxygen partial pressure prediction module: for labeling temperature zone labels for real-time detection data and inputting them into the trained temperature-oxygen partial pressure coupling model to output the prediction results of each collection area; Result judgment module: for judging whether the prediction results are consistent with the data range in the corresponding temperature-oxygen partial pressure reference table according to the pre-set temperature-oxygen partial pressure reference table of each collection area: If the prediction results are consistent with the data range in the corresponding temperature-oxygen partial pressure reference table, no processing is performed; If the prediction results are not consistent with the data range in the corresponding temperature-oxygen partial pressure reference table, adjusting the temperature and oxygen partial pressure of the sintering furnace according to the prediction results.

Images