Cizhou Kiln Process Digital Twin System and Its Construction Method, Firing Curve Dynamic Optimization Method
By using the Cizhou kiln process digital twin system, a digital twin is generated using some parameters inside the kiln and corrected in real time, which solves the problem of high cost in existing technologies and realizes efficient and high-quality guidance for porcelain production.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-19
- Publication Date
- 2026-03-10
AI Technical Summary
When existing technologies use digital twins to guide the production of porcelain, they require a large amount of data and computing resources, resulting in high manufacturing costs and poor economic efficiency.
A digital twin system for Cizhou kiln technology is adopted. By acquiring some parameters inside the kiln, a digital twin is generated. Based on the generated digital twin, the state of the embryo is predicted, and the digital twin is corrected in real time to reduce the amount of data and computing resources required.
This reduces the computational resource investment required for building and predicting digital twins, improves economic efficiency and model generation efficiency, and ensures the overall efficiency and quality of porcelain production.
Smart Images
Figure CN121118446B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the technical field of Cizhou kiln process preparation, specifically to a digital twin system for Cizhou kiln process and its construction method, and a dynamic optimization method for firing curves. Background Technology
[0002] The production of porcelain requires a great deal of experience to assist in obtaining porcelain with a higher yield and better quality. When using traditional methods, the main basis for judgment is human experience. This not only requires the operators to be experienced, but also to have a high level of understanding of the specific production environment, which is quite challenging.
[0003] With the gradual development of technology, digital twins have been introduced to solve the above problems. This involves building a digital twin based on the real-time environment and then simulating it to guide actual production. However, the current method of generating digital twins is to acquire as many environmental parameters as possible. This method requires a large amount of data, and both building and predicting based on the digital twin require a lot of computing resources. This can lead to high manufacturing costs and poor economic efficiency. Summary of the Invention
[0004] This application provides a digital twin system for Cizhou kiln process and its construction method, as well as a dynamic optimization method for firing curves, which can reduce the final manufacturing cost and improve economic efficiency.
[0005] The specific technical solution of this embodiment is as follows:
[0006] On one hand, this application provides a digital twin system for Cizhou kiln process, including a data acquisition module, a data classification module, a digital twin module, and a judgment module. The data acquisition module is used to acquire multiple parameters in the kiln. The data classification module is used to classify the multiple parameters acquired by the data acquisition module to obtain multiple classification data groups. Each classification data group includes some parameters from the multiple parameters. The digital twin module is used to acquire a classification data group and generate a digital twin based on the acquired classification data group. Then, it performs embryo state prediction based on the generated digital twin. The judgment module is used to determine whether there is a difference between the embryo state predicted by the digital twin module and the actual embryo state of the corresponding process. When the judgment module determines that there is a difference, it sends a first signal to the digital twin module. The digital twin module is also used to acquire a new classification data group after receiving the first signal and generate a digital twin again based on the acquired classification data group.
[0007] In some embodiments, the data acquisition module is configured to acquire the state of the blank and is also configured to acquire at least some parameters of the overall temperature, local temperature, atmosphere, pressure, airflow and humidity inside the kiln.
[0008] In some embodiments, the data classification module is configured to classify at least the following data groups: a first data group including the overall temperature inside the kiln; a second data group including multiple local temperatures inside the kiln; a third data group including humidity; a fourth data group including pressure and airflow; and a fifth data group including atmosphere.
[0009] In some embodiments, the digital twin module is configured to first generate a digital twin based on a first category data group, and then sequentially overlay a second to a fifth category data group after each first signal is received.
[0010] On the other hand, embodiments of this application provide a method for constructing a digital twin of the Cizhou kiln process, including the following steps:
[0011] S10. Obtain the first category data group;
[0012] S20. Generate a digital twin based on the first category of data;
[0013] S30. Based on the generated digital twin and the preset firing curve, predict the embryo's predicted state beyond the current time.
[0014] S40. Real-time acquisition of the actual state of the green body inside the kiln;
[0015] S50. Compare the actual state of the embryo inside the kiln with the predicted state of the embryo at the same process.
[0016] S60. When the actual state of the embryo inside the kiln is the same as the predicted state of the embryo at the same process, the process ends.
[0017] S70. When the actual state of the embryo inside the kiln is different from the predicted state of the embryo at the same process, the generated digital twin is corrected.
[0018] In some embodiments, S70, when the actual state of the embryo inside the kiln is different from the predicted state of the embryo at the same process, correcting the generated digital twin includes the following steps:
[0019] S701, Sequentially increase the acquisition of data groups from the second category to the fifth category;
[0020] S702. Generate a digital twin based on the acquired classification data set;
[0021] S703. Repeat steps S30-S70 until it is determined that the actual state of the obtained blank in the kiln is the same as the predicted state of the blank in the same process, then end; or if all classification data groups have been added, but the actual state of the obtained blank in the kiln is still different from the predicted state of the blank in the same process, then end.
[0022] S704. When it is determined that the actual state of the obtained blank in the kiln is the same as the predicted state of the blank in the same process, the corrected digital twin is output.
[0023] S705. When all classification data groups have been added, but the actual state of the embryo inside the kiln is still different from the predicted state of the embryo in the same process, the final corrected digital twin is output.
[0024] In some embodiments, when adding a second category of data, the following steps are also included:
[0025] K10. First, a digital twin is generated based on the temperature of one local area inside the kiln as the second category data group.
[0026] K20, the digital twin obtained based on K10, performs step S703;
[0027] K30. When it is determined that the actual state of the embryo inside the kiln is the same as the predicted state of the embryo in the same process, the corrected digital twin is output.
[0028] K40. When it is determined that the actual state of the obtained green body in the kiln is different from the predicted state of the green body in the same process, an additional local temperature is added as the second classification data group, and steps K10-K30 are repeated until it is determined that the actual state of the obtained green body in the kiln is the same as the predicted state of the green body in the same process, or all local temperatures are used as the second classification data group, and the actual state of the obtained green body in the kiln is still different from the predicted state of the green body in the same process, then steps S701-S705 are continued.
[0029] In some embodiments, after S705, when the actual state of the green body inside the kiln is still different from the predicted state of the green body at the same process, and after outputting the finally corrected digital twin, the following steps are also included:
[0030] S706. Based on the final corrected digital twin, adjust the preset firing curve until the predicted state of the embryo and the actual state of the embryo are consistent.
[0031] This application also provides a method for dynamic optimization of firing profiles, including the following steps:
[0032] T10. Based on the final digital twin and the preset firing curve, the predicted state of the embryo is obtained;
[0033] T20. Compare the predicted state of the embryo with the actual state of the embryo under the same process.
[0034] T30. When the predicted state of the embryo is determined to be different from the actual state of the embryo under the same process, the preset firing curve is optimized until the predicted state of the embryo is determined to be the same as the actual state of the embryo under the same process, and the optimized firing curve is output.
[0035] Compared with the prior art, the embodiments of this application have the following beneficial effects:
[0036] The Cizhou ware digital twin system provided in this application generates a digital twin by acquiring some parameters from the kiln, and predicts the state of the unglazed pieces based on the generated digital twin. This reduces the amount of data acquired and the amount of data required for model construction, thereby reducing the computational resources needed for building the digital twin and making predictions based on it. This reduces the investment of computational resources, ultimately lowering manufacturing costs and improving economic efficiency. It also improves the efficiency of model generation and prediction, further enhancing the overall efficiency of porcelain production. In a specific example, when multiple unglazed pieces exist in the kiln, the actual state of multiple pieces can be acquired simultaneously and compared with the predicted states. If the actual state of one piece differs from the predicted state, the digital twin is corrected in real time, and prediction is performed again. This improves the final production quality of each unglazed piece while maintaining a basic yield rate, thus improving overall quality. This system has significant guiding value for the manufacture of porcelain with multiple batches and high-quality requirements. Attached Figure Description
[0037] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0038] Figure 1 This is a schematic diagram of the process structure of the method for constructing a digital twin of the Cizhou kiln process provided in some embodiments of this application;
[0039] Figure 2 This is a schematic diagram of the process structure for modifying the generated digital twin provided in some embodiments of this application. Detailed Implementation
[0040] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0041] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.
[0042] The use of "applies to" or "configured to" in this application implies open and inclusive language, which does not exclude the applicability to or configuration to devices performing additional tasks or steps. Additionally, the use of "based on" implies openness and inclusivity, because processes, steps, calculations, or other actions "based on" one or more of the stated conditions or values may in practice be based on additional conditions or values beyond those stated.
[0043] In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.
[0044] On one hand, this application provides a digital twin system for Cizhou kiln process, including a data acquisition module, a data classification module, a digital twin module, and a judgment module. The data acquisition module is used to acquire multiple parameters in the kiln. The data classification module is used to classify the multiple parameters acquired by the data acquisition module to obtain multiple classification data groups. Each classification data group includes some parameters from the multiple parameters. The digital twin module is used to acquire a classification data group and generate a digital twin based on the acquired classification data group. Then, it performs embryo state prediction based on the generated digital twin. The judgment module is used to determine whether there is a difference between the embryo state predicted by the digital twin module and the actual embryo state of the corresponding process. When the judgment module determines that there is a difference, it sends a first signal to the digital twin module. The digital twin module is also used to acquire a new classification data group after receiving the first signal and generate a digital twin again based on the acquired classification data group.
[0045] The data acquisition module can acquire multiple different types of parameters inside the kiln, as well as multiple parameters of the same type in different areas of the kiln. For example, it can acquire the state of the green body, and at least some of the overall temperature, local temperature, atmosphere, pressure, airflow and humidity inside the kiln.
[0046] When classifying multiple parameters, the data classification module can produce a categorized data set that includes multiple data points of different types, or multiple data points of the same type located in different regions, such as temperature and humidity in multiple local areas. The parameters in the categorized data set obtained by the data classification module are only a subset of the multiple parameters acquired by the data acquisition module.
[0047] The digital twin module first acquires a categorized dataset, then generates a digital twin based on this dataset, and finally predicts the embryo state based on the generated digital twin. The steps of generating the digital twin from the categorized dataset and predicting the embryo state based on the generated digital twin can be implemented using existing technologies. In some examples, the embryo state can be predicted based on the generated digital twin and a preset firing curve, yielding theoretical embryo state data.
[0048] The judgment module is used to judge the embryo state predicted by the digital twin module and the actual state of the corresponding process. The corresponding process refers to the state where the predicted time point and the actual time point are the same when the embryo state is predicted based on the input parameters. For example, after the digital twin is generated with the input parameters at the first time, the digital twin predicts the embryo state at the second time. When judging whether there is a difference between the two, the embryo state at the second time under the actual time is also selected.
[0049] When refining the generated digital twin, after acquiring a new classification data set, both the original classification data set and the newly acquired classification data set are used as input parameters for generating the digital twin model to produce a new digital twin. In subsequent steps, each time a new classification data set is added, the previously acquired classification data set will be used as input parameters to obtain a new digital twin.
[0050] In the above embodiments, by acquiring partial parameters from the kiln to generate a digital twin, and then using this digital twin to predict the state of the porcelain blanks, the amount of data acquired and the amount of data required for model construction can be reduced. This reduces the computational resources needed for constructing the digital twin and making predictions based on it, thereby reducing the investment in computational resources and ultimately lowering manufacturing costs. This effectively improves economic efficiency and also increases the efficiency of model generation and prediction, further enhancing the overall efficiency of porcelain production. In a specific example, when multiple blanks exist in the kiln, the actual state of multiple blanks can be acquired simultaneously and compared with the predicted states of multiple blanks. If the actual state of one blank differs from the predicted state, the digital twin is corrected in real time, and prediction is performed again. This can improve the final production quality of each blank while ensuring the basic yield rate, thereby improving the overall quality. This has significant guiding significance for the manufacture of porcelain with multiple batches and high-quality requirements.
[0051] In some embodiments, the data acquisition module is configured to acquire the state of the blank and is also configured to acquire at least some of the parameters of overall temperature, local temperature, atmosphere, pressure, airflow and humidity inside the kiln.
[0052] The overall temperature inside the kiln refers to the temperature value that best reflects the overall environmental temperature inside the kiln. In addition to this temperature value, multiple local temperatures are also set to make a more precise judgment on the environment inside the kiln. The nature of the atmosphere in different parts of the kiln can be classified as oxidizing, neutral, or reducing.
[0053] Through the settings of the above embodiments, the data classification module can selectively classify among the multiple parameters obtained above, and generate digital twins based on different classification data groups obtained by the digital twin module. Then, when it is necessary to optimize the generated digital twins, different classification data groups can be added to generate new digital twins.
[0054] In some embodiments, the data classification module is configured to classify at least the following data groups, including a first data group, a second data group, a third data group, a fourth data group, and a fifth data group, wherein the first data group includes the overall temperature inside the kiln, the second data group includes multiple local temperatures inside the kiln, the third data group includes humidity, the fourth data group includes pressure and airflow, and the fifth data group includes atmosphere.
[0055] The second category of data includes multiple local temperatures, which can be a subset of all local temperature parameters or all local temperature parameters. The fourth category of data includes pressure and airflow. The pressure value inside the kiln can be determined based on the input airflow value and the airflow value detected in the detection area, or the pressure value inside the kiln can be obtained directly.
[0056] Through the settings of the above embodiments, the digital twin module can selectively acquire a first category data group to generate a digital twin, or acquire a first category data group and a second category data group to generate a digital twin, or acquire a first category data group, a second category data group, and a third category data group to generate a digital twin, or acquire data from a first category data group, a second category data group, a third category data group, and a fourth category data group to generate a digital twin, or acquire a first category data group, a second category data group, a third category data group, a fourth category data group, and a fifth category data group to generate a digital twin.
[0057] In other embodiments, the first category data group may include the overall temperature and multiple local temperatures within the kiln, the second category data group may include multiple local temperatures and humidity, the third category data group may include humidity, pressure and airflow, and the fourth category data group may include pressure, airflow and atmosphere.
[0058] In other embodiments, each classification data group may include not only one of overall temperature, local temperature, atmosphere, pressure, airflow and humidity, but also multiple of overall temperature, local temperature, atmosphere, pressure, airflow and humidity, and multiple classification data groups may include the same parameters, and more classification data groups may be set.
[0059] In some embodiments, the digital twin module is configured to first generate a digital twin based on a first category data group, and then sequentially overlay a second to a fifth category data group after each first signal is received.
[0060] When the newly generated digital twin needs to be processed, the judgment module sends a first signal to the digital twin module, and the digital twin module continues to acquire the second category data group, that is, it acquires the first category data group and the second category data group simultaneously, until finally it acquires the first category data group, the second category data group, the third category data group, the fourth category data group and the fifth category data group simultaneously.
[0061] By using the settings described in the above embodiments, the digital twin can be adjusted multiple times to ensure that the generated digital twin can achieve better prediction results.
[0062] On the other hand, please see Figure 1 This application provides a method for constructing a digital twin of the Cizhou kiln process, including the following steps:
[0063] S10. Obtain the first category data group.
[0064] S20. Generate a digital twin based on the first category of data.
[0065] In S20, a digital twin can be generated based on the overall temperature inside the kiln.
[0066] S30. Based on the generated digital twin and the preset firing curve, the predicted state of the embryo beyond the current time is obtained.
[0067] In S30, the preset firing curve can be preset based on previous porcelain of the same type, or it can be preset based on the operator's experience.
[0068] S40. Real-time acquisition of the actual state of the blanks inside the kiln.
[0069] In S40, the actual state of the green body inside the kiln can be obtained based on the camera module. For example, a high-temperature industrial camera can be used to record the deformation process of the green body during firing.
[0070] S50. Compare the actual state of the embryo inside the kiln with the predicted state of the embryo at the same process.
[0071] In S50, the actual state of the embryonic body acquired by the camera module can be modeled and then compared with the predicted state of the embryonic body at the same stage to determine the morphological differences between the two. In a specific example, a threshold can be set when determining the morphological differences between the two, for example, only when the morphological difference exceeds a certain proportion is it determined that there is a morphological difference between the two.
[0072] S60. When the actual state of the embryo inside the kiln is the same as the predicted state of the embryo at the same process, the process ends.
[0073] In S60, when it is determined that the actual state of the embryo in the kiln is the same as the predicted state of the embryo in the same process, that is, the currently generated digital twin can meet the prediction requirements, then the digital twin is used for prediction.
[0074] S70. When the actual state of the embryo inside the kiln is different from the predicted state of the embryo at the same process, the generated digital twin is corrected.
[0075] In the above embodiments, by acquiring partial parameters from the kiln to generate a digital twin, and then using this digital twin to predict the state of the porcelain blanks, the amount of data acquired and the amount of data required for input can be reduced. This reduces the computational resources needed to construct the digital twin and make predictions based on it, thereby reducing the investment in computational resources and ultimately lowering manufacturing costs. This effectively improves economic efficiency and also increases the efficiency of model generation and prediction, further enhancing the overall efficiency of porcelain production. In a specific example, when multiple blanks exist in the kiln, the actual states of multiple blanks are acquired simultaneously and compared with the predicted states of multiple blanks. If the actual state of one blank differs from the predicted state, the digital twin is corrected in real time and prediction is performed again. This can improve the final production quality of each blank while ensuring the basic yield rate, thereby improving the overall quality. This has significant guiding significance for the production of high-quality porcelain in multiple batches.
[0076] In some of these embodiments, please refer to Figure 2 In step S70, when the actual state of the embryo inside the kiln differs from the predicted state of the embryo at the same stage, the correction of the generated digital twin includes the following steps:
[0077] S701, sequentially acquire data from the second category to the fifth category.
[0078] In S701, a new classification data group is added each time the digital twin is modified.
[0079] S702. Generate a digital twin based on the acquired classification data set.
[0080] S703. Repeat steps S30-S70 until it is determined that the actual state of the obtained blank in the kiln is the same as the predicted state of the blank in the same process, then end; or if all classification data groups have been added, but the actual state of the obtained blank in the kiln is still different from the predicted state of the blank in the same process, then end.
[0081] In steps S701-S703, when correcting the generated digital twin, firstly, a second category data group is acquired, and a digital twin is generated based on the previous first category data group and the added second category data group. Then, steps S30-S70 are repeated. In this step, it is possible that the actual state of the acquired blank in the kiln is the same as the predicted state of the blank in the same process, in which case the process ends. It is also possible that the actual state of the acquired blank in the kiln is different from the predicted state of the blank in the same process, in which case a third category data group is added, until all category data groups have been added, but the actual state of the acquired blank in the kiln is still different from the predicted state of the blank in the same process, then the process ends.
[0082] S704. When it is determined that the actual state of the embryo inside the kiln is the same as the predicted state of the embryo in the same process, the corrected digital twin is output.
[0083] S705. When all classification data groups have been added, but the actual state of the embryo inside the kiln is still different from the predicted state of the embryo in the same process, the final corrected digital twin is output. In a specific example, it can be a digital twin generated based on the first classification data group, the second classification data group, the third classification data group, the fourth classification data group, and the fifth classification data group.
[0084] By setting up the above embodiments, the final output digital twin can better match the actual embryonic state and corresponding environment, thereby enabling better prediction of the embryonic morphology and obtaining more accurate prediction results.
[0085] In some embodiments, when adding a second category of data, the following steps are also included:
[0086] K10. First, a digital twin is generated based on the temperature of one local area in the kiln as the second category data group.
[0087] K20, based on the digital twin obtained from K10, performs step S703.
[0088] K30. When it is determined that the actual state of the embryo inside the kiln is the same as the predicted state of the embryo in the same process, the corrected digital twin is output.
[0089] K40. When it is determined that the actual state of the obtained green body in the kiln is different from the predicted state of the green body in the same process, then add another local temperature as the second classification data group, and repeat steps K10-K30 until it is determined that the actual state of the obtained green body in the kiln is the same as the predicted state of the green body in the same process, or all local temperatures are used as the second classification data group, and the actual state of the obtained green body in the kiln is still different from the predicted state of the green body in the same process, then continue to steps S701-S705.
[0090] In the above embodiments, by sequentially increasing the acquired local temperatures as the second classification data group, an accurate digital twin can be obtained while acquiring less data. This reduces the amount of data acquired and the amount of data required for input, thereby reducing the computational resources needed to construct the digital twin and make predictions based on it. This reduces the investment in computational resources, thereby reducing the final manufacturing cost and effectively improving economic efficiency. It also improves the efficiency of model generation and prediction, further enhancing the overall efficiency of porcelain production.
[0091] In some embodiments, after outputting the final corrected digital twin in step S705 when the actual state of the green body inside the kiln is still different from the predicted state of the green body at the same process, the following steps are also included:
[0092] S706. Based on the final corrected digital twin, adjust the preset firing curve until the predicted state of the embryo and the actual state of the embryo are consistent.
[0093] Through the settings of the above embodiments, the preset firing curve is adjusted based on the final corrected digital twin until the predicted state of the embryo and the actual state of the embryo are consistent. By adjusting the firing curve to better fit the current embryo condition, better prediction results are achieved.
[0094] This application also provides a method for dynamic optimization of firing profiles, including the following steps:
[0095] T10. Based on the final digital twin and the preset firing curve, the predicted state of the embryo is obtained;
[0096] T20. Compare the predicted state of the embryo with the actual state of the embryo under the same process.
[0097] T30. When the predicted state of the embryo is determined to be different from the actual state of the embryo under the same process, the preset firing curve is optimized until the predicted state of the embryo is determined to be the same as the actual state of the embryo under the same process, and the optimized firing curve is output.
[0098] By setting the above embodiments, the firing curve is dynamically optimized to better fit the current embryo condition, thereby achieving better prediction results.
[0099] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.
Claims
1. A digital twin system for Cizhou kiln technology, characterized in that, The application relates to a kiln process prediction method and system. The application comprises: a data acquisition module that acquires multiple parameters in a kiln; a data classification module that classifies the multiple parameters acquired by the data acquisition module to obtain multiple classification data sets, each of which comprises part of the multiple parameters; a digital twin module that acquires a classification data set and generates a digital twin based on the acquired classification data set, and then performs embryo state prediction based on the generated digital twin; a judgment module that judges whether there is a difference between the embryo state predicted by the digital twin module and the actual state of the corresponding process, and sends a first signal to the digital twin module when there is a difference; the digital twin module is further configured to, after receiving the first signal, acquire a new classification data set and generate a digital twin again based on the acquired classification data set; the data acquisition module is configured to acquire the embryo state and at least part of the parameters of the overall temperature, local temperature, atmosphere, pressure, airflow and humidity in the kiln; the data classification module is configured to at least classify the following classification data sets: a first classification data set comprising the overall temperature in the kiln; a second classification data set comprising multiple local temperatures in the kiln; a third classification data set comprising humidity; a fourth classification data set comprising pressure and airflow; a fifth classification data set comprising the atmosphere; 2. A method for constructing a digital twin of a Cizhou kiln process, based on the digital twin system of a Cizhou kiln process according to claim 1, characterized in that, the digital twin module is configured to first generate a digital twin based on the first classification data set, and sequentially superimpose the second classification data set to the fifth classification data set after receiving the first signal each time. The application comprises the following steps: S10, acquiring a first classification data set; S20, generating a digital twin based on the first classification data set; S30, predicting the embryo prediction state beyond the current time based on the generated digital twin and a preset firing curve; S40, acquiring the actual state of the embryo in the kiln in real time; S50, comparing the acquired actual state of the embryo in the kiln with the predicted embryo prediction state of the same process; S60, when the acquired actual state of the embryo in the kiln is the same as the predicted embryo prediction state of the same process, the process ends; 3. The Cizhou kiln process digital twin construction method of claim 2, wherein, S70, when the acquired actual state of the embryo in the kiln is different from the predicted embryo prediction state of the same process, the generated digital twin is modified. S70, when the acquired actual state of the embryo in the kiln is different from the predicted embryo prediction state of the same process, the generated digital twin is modified, which comprises the following steps: S701, sequentially increasing the acquisition of the second classification data set to the fifth classification data set; S702, generating a digital twin based on the acquired classification data set; S703, repeating the steps S30-S70 until it is determined that the acquired actual state of the embryo in the kiln is the same as the predicted embryo prediction state of the same process, and then the process ends; or all classification data sets have been increased, but the acquired actual state of the embryo in the kiln is still different from the predicted embryo prediction state of the same process, and then the process ends. S704, when the actual state of the green body in the kiln obtained is determined to be the same as the predicted state of the green body of the same process, output the modified digital twin; S705, when all the classification data sets have been added, but the actual state of the green body in the kiln obtained is still not the same as the predicted state of the green body of the same process, output the last modified digital twin.
4. The Cizhou kiln process digital twin construction method of claim 3, wherein, When the second classification data set is added, the following steps are further included: K10, first take one of the local temperatures in the kiln as the second classification data set, and generate a digital twin based on the added second classification data set; K20, perform the S703 step based on the digital twin obtained in K10; K30, when the actual state of the green body in the kiln obtained is determined to be the same as the predicted state of the green body of the same process, output the modified digital twin; K40, when the actual state of the green body in the kiln obtained is determined to be not the same as the predicted state of the green body of the same process, then add another local temperature as the second classification data set, and repeat the K10-K30 steps until the actual state of the green body in the kiln obtained is determined to be the same as the predicted state of the green body of the same process, or all local temperatures are taken as the second classification data set, and the actual state of the green body in the kiln obtained is still not the same as the predicted state of the green body of the same process, then continue to perform the S701-S705 steps. S705, when the actual state of the green body in the kiln obtained is still not the same as the predicted state of the green body of the same process, after outputting the last modified digital twin, the following steps are further included:
5. The Cizhou kiln process digital twin construction method of claim 3, wherein, S706, based on the last modified digital twin, adjust the preset firing curve until the green body predicted state and the green body actual state are consistent. The following steps are included:
6. A firing curve dynamic optimization method, based on the digital twin construction method of the Meizhou kiln process according to any one of claims 2-5, wherein the digital twin obtained is characterized by, T10, based on the final digital twin and the preset firing curve, obtain the green body predicted state; T20, compare the green body predicted state with the actual state of the green body of the same process; T30, when the green body predicted state is determined to be not the same as the actual state of the green body of the same process, then optimize the preset firing curve until the green body predicted state is determined to be the same as the actual state of the green body of the same process, and output the optimized firing curve.
Citation Information
Patent Citations
Digital twin system for ceramic sintering process
CN116822310A
Digital twinborn model automatic optimization method
CN119622986A