Operation condition determining device, operation condition determining method and program
The operating condition determining device and method address the challenge of high-mix, low-volume steel production by setting same component values in the refining process, ensuring quality and productivity through inverse analysis, thus integrating steel types effectively.
Patent Information
- Application Number
- JP2023082364
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-05-18
- Publication Date
- 2024-11-28
- Estimated Expiration
- 2043-05-18
AI Technical Summary
The steel industry faces challenges in improving productivity and quality due to high-mix, low-volume production, where setting different steel types for each customer's specifications increases the number of steel types, making it difficult to maintain consecutive steel types with the same composition, leading to inefficiencies in manufacturing processes.
An operating condition determining device and method that uses a material property value prediction model to set component values the same in the refining process, allowing for inverse analysis to determine appropriate operating conditions in subsequent processes, ensuring the probability of meeting material property value specifications exceeds a threshold, thereby integrating multiple steel materials into the same type.
This approach enhances productivity by increasing the number of consecutive steel types with the same composition, improving quality and reducing manufacturing costs by aggregating steel types, even under conditions with limited operational history or high variation.
Smart Images

Figure 2024165832000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to an operation condition determination device, an operation condition determination method, and a program for causing a computer to function as an operation condition determination device, which determine operation conditions in a manufacturing process for a plurality of steel products having different manufacturing specifications. In particular, the present invention relates to an operation condition determination device, an operation condition determination method, and a program for determining appropriate operation conditions capable of improving the productivity and quality of steel products. [Background technology]
[0002] Steel materials such as steel plates are manufactured through manufacturing processes such as a refining process, a continuous casting process, a rolling process (hot rolling process, cold rolling process), a continuous annealing process, and a continuous hot-dip galvanizing process.
[0003] In the converter, which performs the refining process, a batch operation is performed, and the molten steel produced in one refining run (called a charge) is tapped into a ladle. Then, in the continuous casting process, several charges of molten steel produced in the converter and tapped into a ladle are cast continuously. The number of charges cast continuously is called the number of consecutive runs, and from the viewpoint of improving the productivity of steel materials, the larger the number of consecutive runs, the better. On the other hand, when charges of different steel types (types of steel with different component values of molten steel in the refining process) are cast continuously, the components at the joints between the charges will mix, so from the viewpoint of improving the quality of the steel, it is preferable to have a smaller number of consecutive charges for different steel types. Therefore, from the viewpoint of improving both the productivity and quality of steel, it is preferable to increase the number of runs of the same steel type (type of steel having the same component values of molten steel in the refining process) in the continuous casting process.
[0004] However, in the steel industry, high-mix low-volume production is common. Therefore, if the steel grades of steel to be manufactured are set for each customer's different manufacturing specifications, the number of steel grades increases, making it difficult to increase the number of consecutive productions of the same steel grade. Therefore, from the viewpoint of improving both the productivity and quality of steel, it is important to determine the operating conditions in the steel manufacturing process so that the number of steel grades to be set is as small as possible. In order to reduce the number of steel types to be set, it is preferable to set the same steel type for the different manufacturing specifications of each customer, while changing the operating conditions of the processes subsequent to the refining process (e.g., continuous casting process, rolling process (hot rolling process, cold rolling process), continuous annealing process, continuous hot-dip galvanizing process, etc.) according to the manufacturing specifications so as to satisfy the manufacturing specifications. In other words, it is preferable to determine the operating conditions according to the manufacturing specifications so that multiple steel materials that were previously set to different steel types can be consolidated into the same steel type (this is called steel type consolidation).
[0005] As technologies related to steel type consolidation, for example, the devices described in Patent Documents 1 and 2 have been proposed.
[0006] The device described in Patent Document 1 is a device that consolidates steel types by bundling lower specification (low quality) steel types with higher specification (high quality) steel types in order to reduce the risk of not being able to meet customer manufacturing specifications (Claim 1 of Patent Document 1, etc.). The device described in Patent Document 1 has the problem that lower-grade steel products are produced from charges containing excessive amounts of alloy, just like higher-grade steel products, resulting in increased production costs.
[0007] The device described in Patent Document 2 is a device that estimates the mechanical properties of steel products using a database-type material prediction system equipped with a manufacturing information storage means that accumulates, as examples, the actual values of chemical components and operating conditions, and the actual values of mechanical properties for each steel product manufactured in the past, and a material estimation means that utilizes the data accumulated in the manufacturing information storage means to estimate the mechanical properties that will be obtained when the steel product is manufactured based on input specified values of the chemical components and operating conditions.The device confirms that the estimated values satisfy the specifications of the orders, and consolidates multiple orders of different specifications into one production lot (i.e., a steel type consolidation device) (Claim 3 of Patent Document 2, etc.). The device described in Patent Document 2 does not take into consideration the uncertainty of predictions (estimations) due to the absence of actual values of operating conditions, etc. corresponding to the specifications of the order in the past data stored in the manufacturing information storage means. Therefore, even if steel materials are manufactured using a manufacturing lot organized with the device described in Patent Document 2, there is a high risk that the material characteristic values of the manufactured steel materials (mechanical properties in Patent Document 2) will not satisfy the specifications of the order (the material characteristic values of the steel materials will be unacceptable). In addition, since the chemical components and the indicated values of operating conditions to be input to the material estimation means must be manually examined, there is a problem in that appropriate steel type consolidation is not always possible.
[0008] Non-Patent Documents 1 and 2 describe heteroscedastic Gaussian processes (Gaussian process regression taking heteroscedasticity into account). [Prior art documents] [Patent documents]
[0009] [Patent Document 1] JP 2006-159274 A [Patent Document 2] JP 2009-020807 A [Non-patent literature]
[0010] [Non-Patent Document 1] Alan D. Saul, James Hensman, Aki Vehtari, Neil D. Lawrence, "Chained Gaussian Processes", Proceedings of the 19th International Conference on Artificial Intelligence and Statistics, PMLR 51:1431-1440, 2016 [Non-Patent Document 2] "Heteroskedastic Likelihood and Multi-Latent GP", [online], [Retrieved April 21, 2023], Internet<URL:https: / / gpflow.github.io / GPflow / develop / notebooks / advanced / heteroskedastic.html> Summary of the Invention [Problem to be solved by the invention]
[0011] The present invention has been made to solve the problems of the conventional technology described above, and has an object to provide an operating condition determination device, an operating condition determination method, and a program that can determine appropriate operating conditions in a manufacturing process for a plurality of steel products having different manufacturing specifications, which can improve the productivity and quality of the steel products. [Means for solving the problem]
[0012] In order to solve the above problem, the present invention provides an operating condition determination device that determines operating conditions in the manufacturing process for a plurality of steel materials having different manufacturing specifications, which are manufactured through a manufacturing process having a refining process and a subsequent process subsequent to the refining process, under the condition that the component values in the refining process of the plurality of steel materials are set to be identical to each other. The operating condition determination device is provided with a material property value prediction model that inputs the operating conditions in the manufacturing process of a target steel material among the plurality of steel materials to be subjected to reverse analysis and outputs a probability distribution of the material property values of the target steel material, and determines the operating conditions in at least the subsequent process of the manufacturing process of the target steel material by reverse analyzing the material property value prediction model so that the probability of passing the material property values of the target steel material, which is calculated based on the probability distribution of the material property values of the target steel material output when the operating conditions in the manufacturing process of the target steel material, including the component values in the refining process that are set to be identical, and the specifications of the material property values included in the manufacturing specifications of the target steel material, is equal to or greater than a predetermined threshold value.
[0013] In the present invention, the "manufacturing process" refers to a process up to a process for manufacturing a steel material into a product by changing the thickness, surface condition, and internal metal structure of the steel material (a rolling process (hot rolling process, cold rolling process), a continuous annealing process, a continuous hot-dip galvanizing process, etc.), and may not include a refinement process in which only cutting and inspection of the steel material is performed. In the present invention, the "manufacturing process" for which the operating conditions are determined by reverse analysis of the material characteristic value prediction model can include, for example, one or more processes among a refining process, a continuous casting process, a rolling process (hot rolling process, cold rolling process), a continuous annealing process, and a continuous hot-dip galvanizing process. However, the manufacturing process for which the operating conditions are determined by reverse analysis of the material characteristic value prediction model includes at least one process among the downstream processes (the continuous casting process, which is a downstream process after the refining process, the rolling process (hot rolling process, cold rolling process), the continuous annealing process, the continuous hot-dip galvanizing process, etc.). In addition, in the present invention, "multiple steel materials" does not mean actual steel materials, but means steel materials that are manufactured (i.e., that will be manufactured in the future) according to the operational conditions in the manufacturing process that are determined by the operational condition determination device according to the present invention. In addition, "steel material" in the present invention does not necessarily mean one steel product, and a group of steel products consisting of multiple steel products may be collectively referred to as one steel material. In this case, "multiple steel materials" means that there are multiple groups of steel products. In the present invention, "manufacturing specifications" refer to specifications relating to steel that are determined in response to customer orders, such as the size of steel and material property values. Manufacturing specifications are not necessarily determined for each customer order, but may be determined for each order from multiple customers (one manufacturing specification may be determined for multiple customer orders). In the present invention, "different manufacturing specifications" is not limited to cases where all specifications of the elements that make up the manufacturing specifications (such as the size of steel and material property values) are different, but also includes cases where some of the specifications are different. In addition, in the present invention, "setting the component values in the refining process to be identical to each other" is not limited to the case where the component values of the steel (molten steel) in the refining process are set to be completely identical, but is a concept that includes the case where the component values are set within the same predetermined range. In the present invention, the "operational conditions" refer to conditions that affect the quality of the steel material, such as its material property values, during the manufacturing process. Examples of such conditions include the component values of carbon (C), manganese (Mn), silicon (Si), etc. during the refining process, and the heating temperature, rolling temperature, and cooling temperature during the rolling process. In addition, in the present invention, "material characteristic value" means the mechanical properties of a steel material, and examples thereof include tensile strength (TS), yield point (YP), elongation (EL), surface hardness, bending strength, toughness value, and fatigue life. Furthermore, in the present invention, "specifications of material property values of steel materials included in the manufacturing specifications" can include, for example, a required upper limit value or a required lower limit value of a material property value, or both.
[0014] The operating condition determination device of the present invention is a device that determines operating conditions in a manufacturing process for multiple steel products having different manufacturing specifications (such as the size of the steel product and material property values) that are manufactured through a manufacturing process including a refining process and a post-process, under conditions that set the component values of the multiple steel products in the refining process to be identical (in other words, conditions that set the steel types of the multiple steel products to be the same). The operating condition determination device according to the present invention is equipped with a material property value prediction model that inputs operating conditions in a manufacturing process of a target steel material to be subjected to reverse analysis among a plurality of steel materials, and outputs a probability distribution of material property values of the target steel material. The material property value prediction model outputs the probability distribution of material property values of the target steel material, in other words, the expected value and variation of material property values, as a prediction result, so that it is expected that the material property values (especially the variation) can be predicted with high accuracy even for operating conditions of the input target steel material that have no past operating history. In addition, it is expected that the prediction accuracy of material property values can be improved even for operating conditions of the target steel material where it is difficult to control the material property values and variation in the material property values is likely to occur. In the operating condition determination device according to the present invention, the pass probability of the material property value of the target steel is calculated based on the probability distribution of the material property value of the target steel output when the operating conditions (including the component values in the refining process set to the same) in the manufacturing process of the target steel are input to the material property value prediction model, and the specifications of the material property value included in the manufacturing specifications of the target steel. For example, if the specifications of the material property value of the target steel included in the manufacturing specifications are the required upper limit and required lower limit of the material property value, the ratio of the area within the range from the required lower limit to the required upper limit to the area of the entire probability distribution of the material property value is calculated as the pass probability. Therefore, by reverse-analyzing this material property value prediction model so that the probability of passing the material property value of the target steel material is above a predetermined threshold value, the operating conditions for at least the later processes in the manufacturing process of the target steel material can be appropriately determined in accordance with the manufacturing specifications. According to the operating condition determination device of the present invention, the operating conditions determined for a plurality of steel materials having different manufacturing specifications are appropriate in accordance with the manufacturing specifications, and the component values in the refining processes of the plurality of steel materials are set to be identical to each other, so that the number of consecutive steel materials of the same type can be increased, thereby improving the productivity and quality of the steel materials.
[0015] The material property value prediction model of the present invention is constructed by using known data in which, for example, the operational results in the manufacturing process of steel material manufactured in the past are input, and the material property values measured on test pieces obtained from the steel material manufactured in the past are output. For example, the model may be constructed by machine learning using such known data as teacher data. In the present invention, the "operational results" used as known data are actual values of operating conditions, and are data that affect the quality of the material property values of the steel material. The operational results may include measured values and set values obtained in the manufacturing process of the steel material, as well as data obtained by processing these values using physical formulas, etc. In addition, the material property value prediction model may be constructed by the operating condition determination device of the present invention, or a material property value prediction model constructed by another device may be stored in the material property prediction device of the present invention.
[0016] In a first preferred configuration of the operating condition determination device of the present invention, the component values in the refining process that are set to be the same are component values of existing steel types that are set in steel materials other than the target steel material among the plurality of steel materials, and by reverse analyzing the material property value prediction model, it is possible to determine the operating conditions in the subsequent process among the manufacturing processes of the target steel material.
[0017] According to the first preferred configuration, for example, among a plurality of steel products, a steel product whose production volume per unit time is equal to or less than a predetermined threshold value (a steel product whose production volume per unit time is small) is set as the target steel product, and the component values of existing steel types set for steel products other than the target steel product (in other words, steel products whose production volume per unit time is relatively large) are set as the component values in the refining process that are set to be the same. In this case, the operating conditions are determined to consolidate the steel products to the existing steel products whose production volume per unit time is relatively large, making it possible to sufficiently improve the productivity of steel products.
[0018] In the first preferred configuration, the operating conditions for consolidating the steel type to an existing steel type are determined, but the present invention is not limited to this, and it is also possible to determine operating conditions for consolidating the steel type to a steel type not included in the existing steel types. In this case, since the steel type is not included in the existing steel types, the operating condition determination device also determines the component values in the refining process by performing a reverse analysis of the material characteristic value prediction model. In other words, as a second preferred configuration of the operating condition determination device of the present invention, the component values in the refining process that are set to be the same are component values that do not exist in existing steel types, and by reverse analyzing the material property value prediction model, it is also possible to determine the component values in the refining process and the operating conditions in the subsequent process among the manufacturing processes of the target steel.
[0019] According to the second preferred configuration, for example, it is possible to set all of the multiple steel materials as target steel materials. In this case, the component values not present in the existing steel materials are set to the same component values in the refining process, so that the operating conditions are determined to consolidate all of the multiple steel materials into a new steel material not present in the existing steel materials, and the productivity of the new steel material material can be sufficiently improved.
[0020] In the operation condition determination device according to the present invention, the material characteristic value prediction model is, for example, a model using Bayesian estimation. Also, the material characteristic value prediction model is, for example, a model using a heteroscedastic Gaussian process.
[0021] In order to solve the above problem, the present invention provides an operating condition determination method for determining operating conditions in the manufacturing process for a plurality of steel materials having different manufacturing specifications, which are manufactured through a manufacturing process having a refining process and a subsequent process subsequent to the refining process, under the condition that the component values in the refining process of the plurality of steel materials are set to be the same for each other. The operating conditions in the manufacturing process of the target steel material to be subjected to reverse analysis are input, and a material property value prediction model is used that outputs a probability distribution of the material property values of the target steel material. The operating conditions in the manufacturing process of the target steel material, including the component values in the refining process that are set to be the same, are input to the material property value prediction model. The operating condition in the manufacturing process of the target steel material, which is calculated based on the probability distribution of the material property values of the target steel material output when the operating conditions in the manufacturing process of the target steel material including the component values in the refining process that are set to be the same, and the specifications of the material property values included in the manufacturing specifications of the target steel material, are reverse-analyzed to determine the operating conditions in at least the subsequent process of the manufacturing process of the target steel material, so that the pass probability of the material property values of the target steel material is equal to or greater than a predetermined threshold value.
[0022] Furthermore, in order to achieve the above object, the present invention also provides a program for causing a computer to function as the operation condition determination device. Effect of the Invention
[0023] According to the present invention, it is possible to determine appropriate operating conditions in a manufacturing process for a plurality of steel products having different manufacturing specifications, which can improve the productivity and quality of the steel products. [Brief description of the drawings]
[0024] [Figure 1] 1 is a diagram for illustrating an overview of an operation condition determination device according to an embodiment of the present invention; [Diagram 2] 1 is a block diagram showing a schematic configuration of an operation condition determination device according to a first embodiment. [Diagram 3] 3 is a flowchart showing an outline of a procedure of an operation condition determining method executed by using the operation condition determining device shown in FIG. 2. [Figure 4] 3 is a diagram showing an example of order information acquired by the order information acquiring unit 1 shown in FIG. 2. FIG. [Diagram 5] FIG. 1 is a diagram showing an example of known data used to construct a material characteristic value prediction model. [Figure 6] FIG. 2 is a diagram for explaining a construction example of a material characteristic value prediction model according to the first embodiment in comparison with a construction example of a conventional machine learning model. [Figure 7] FIG. 3 is a diagram showing an example of the contents output from an operation condition output unit 6 shown in FIG. [Figure 8] FIG. 3 is a diagram showing an example of the contents output from an operation condition output unit 6 shown in FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0025] [Outline of the operation condition determination device] Hereinafter, an operation condition determination device according to an embodiment (first embodiment and second embodiment) of the present invention will be described with reference to the accompanying drawings as appropriate, taking a case where the steel material is a steel plate as an example. First, an overview of the operation condition determination device according to the present embodiment will be described.
[0026] FIG. 1 is a diagram for explaining an outline of an operation condition determination device according to the present embodiment. The operating condition determination device according to the present embodiment is a device that determines operating conditions in a manufacturing process for a plurality of steel materials, which are manufactured through a manufacturing process having a refining process and a post-process, under conditions in which the component values in the refining process of the plurality of steel materials are set to be the same (in other words, conditions in which the steel types of the plurality of steel materials are set to be the same). Specifically, as shown in FIG. 1(a), the operating condition determination device according to the present embodiment is provided with a material property value prediction model that inputs operating conditions in a manufacturing process of a target steel material to be subjected to reverse analysis among the plurality of steel materials, and outputs a probability distribution of the material property value of the target steel material. Therefore, based on the probability distribution of the material property value of the target steel material output from the material property value prediction model and the specifications of the material property value included in the manufacturing specifications of the target steel material (for example, the required upper limit value and the required lower limit value of the material property value), the ratio of the area within the range from the required lower limit value to the required upper limit value (the area of the hatched region shown in FIG. 1(a)) to the area of the entire probability distribution of the material property value is calculated as the pass probability. As shown in Figure 1(b), the operating condition determination device of this embodiment determines the operating conditions for at least the later process in the manufacturing process of the target steel by inversely analyzing the material property value prediction model so that the pass probability of the material property value of the target steel (the area of the hatched area shown in Figure 1(b) relative to the area of the entire probability distribution) calculated based on the probability distribution of the material property values of the target steel output when the operating conditions (including the component values in the refining process which are set to the same) in the manufacturing process of the target steel are input into the material property value prediction model and the specifications of the material property values included in the manufacturing specifications of the target steel (for example, the required upper and lower limits of the material property values) is equal to or greater than a predetermined threshold value. The multiple steel materials (e.g., steel materials K01, K02, K03, and K04 described below) for which operating conditions are determined by the operating condition determination device of this embodiment are determined, for example, by an operator or automatically by the operating condition determination device, by selecting steel materials that can be consolidated into steel types in accordance with predetermined criteria related to the steel material's use (including workability and weldability) and quality (material characteristic values, shape, defects, etc.).
[0027] In this embodiment, the manufacturing processes for which operating conditions are determined by reverse analysis of a material characteristic value prediction model will be described as examples of a downstream rolling process (first embodiment), and a refining process and a downstream rolling process (second embodiment). Hereinafter, a first embodiment and a second embodiment, which are different from each other in the manufacturing process and the like for which the operating conditions are determined by inverse analysis, will be described in order.
[0028] [First embodiment] Fig. 2 is a block diagram showing a schematic configuration of an operation condition determination device according to the first embodiment. Fig. 3 is a flow chart showing a schematic procedure of an operation condition determination method executed by using the operation condition determination device shown in Fig. 2. The operating condition determination device 100 of the first embodiment is configured such that the component values in the refining process, which are set to the same value, are the component values of existing steel types that are set for steel materials other than the target steel material among multiple steel materials, and the operating conditions for the subsequent process (rolling process) in the manufacturing process of the target steel material are determined by performing inverse analysis of a material property value prediction model. As shown in FIG. 2, the operating condition determination device 100 according to the first embodiment includes an order information acquisition unit 1, a material characteristic value prediction model storage unit 2, a material characteristic value prediction model acquisition unit 3, a constraint condition setting unit 4, an operating condition determination unit 5, and an operating condition output unit 6.
[0029] The operating condition determination device 100 is composed of a computer having one or more hardware processors such as a CPU (Central Processing Unit), one or more memories such as a RAM (Random Access Memory) and a ROM (Read Only Memory), and one or more programs stored in the memory are executed by the one or more hardware processors to perform various calculations. As a result, the operating condition determination device 100 functions as an order information acquisition unit 1, a material characteristic value prediction model storage unit 2, a material characteristic value prediction model acquisition unit 3, a constraint condition setting unit 4, an operating condition determination unit 5, and an operating condition output unit 6. The operating condition determination device 100 may be a PLC (Programmable Logic Controller) or may be realized by dedicated hardware such as an ASIC (Application Specific Integrated Circuit). Below, each of the components 1 to 6 included in the operating condition determination device 100 will be described in order.
[0030] <Order Information Acquisition Unit 1> The order information acquiring unit 1 executes step ST11 shown in Fig. 3. Specifically, the order information acquiring unit 1 acquires the order information from a predetermined database (not shown). Fig. 4 is a diagram showing an example of order information acquired by the order information acquisition unit 1. As shown in Fig. 4, the order information includes, for example, manufacturing specifications, operating conditions in the refining process, operating conditions in the post-process, and monthly production volume of each type of steel, all linked to a steel material ID corresponding to each steel material ordered by a customer.
[0031] Manufacturing specifications are specifications for steel that are determined in response to customer orders, and may include the size of the steel (plate thickness in the example shown in Fig. 4) and material property values (yield point (YP) and tensile strength (TS) in the example shown in Fig. 4). In the example shown in Fig. 4, the specifications for the material property values include both the required upper and lower limit values for YP and TS, but are not limited to this, and there are also cases where only one of the required upper and lower limit values is included. The operating conditions in the refining process include the component values of the steel (molten steel) (in the example shown in Fig. 4, the component values of carbon (C), silicon (Si), and manganese (Mn)), and the steel grades (in the example shown in Fig. 4, steel grades A, B, and C) are determined according to each component value. These steel grades A, B, and C are existing steel grades. In the example shown in FIG. 4, the operating conditions in the post-processing step are the rolling temperature and the cooling temperature in the rolling step.
[0032] <Material characteristic value prediction model storage section 2> The material characteristic value prediction model storage unit 2 stores (memorizes) a material characteristic value prediction model constructed by a material characteristic value prediction model construction unit (not shown) included in the operation condition determination device 100 or by another device.
[0033] <Material characteristic value prediction model acquisition unit 3> The material characteristic value prediction model acquisition unit 3 executes step ST12 shown in Fig. 4. Specifically, the material characteristic value prediction model acquisition unit 3 acquires the material characteristic value prediction model stored in the material characteristic value prediction model storage unit 2.
[0034] Hereinafter, a method for constructing the material characteristic value prediction model stored in the material characteristic value prediction model storage unit 2 and acquired by the material characteristic value prediction model acquisition unit 3 will be described. The material property value prediction model is constructed by using known data in which the operational results of a manufacturing process of a steel material manufactured in the past (also referred to as a "reference steel material" in this specification) are used as input and the material property values measured on a test piece obtained from the reference steel material are used as output. For example, the material property value prediction model may be constructed by machine learning using the known data as training data. FIG. 5 is a diagram showing an example of known data used to construct a material characteristic value prediction model. The known data shown in FIG. 5 includes, for each reference steel, component values of carbon (C), silicon (Si), and manganese (Mn) as operational results in the refining process. The known data also includes, for each reference steel, rolling temperatures and cooling temperatures in the rolling process as operational results in the subsequent process. Furthermore, the known data also includes, for each reference steel, measured values of YP and TS as material characteristic values. The material characteristic value prediction model is constructed by using the operational results (operational results in the refining process and operation results in the subsequent process) of the known data shown in FIG. 5 as inputs of the material characteristic value prediction model and using the material characteristic values as outputs of the material characteristic value prediction model.
[0035] In the first embodiment, in order to quantify the uncertainty of prediction due to the lack of past operating records, a material property value prediction model is constructed using Gaussian process regression, which is a type of Bayesian estimation, but the present invention is not limited to this. For example, a material property value prediction model may be constructed using a statistical method such as quantile regression using known data, or a material property value prediction model may be constructed by other machine learning such as a Bayesian neural network. Below, we will explain an overview of the Gaussian process regression used in the first embodiment. Note that in the following equations, variables written in bold represent vectors or matrices.
[0036] The input x expressed by the following formula (1) i When the output vector f, which is expressed by the following equation (3) and corresponds to the input x, which is expressed by the following equation (2), follows a Gaussian distribution N(0,K) with a mean of 0 and a covariance matrix K, which is expressed by the following equation (4), the output f is said to follow a Gaussian process, and can be written as shown in the following equation (5).
number
number
number
[0037] Although it is possible to use a normal Gaussian process regression as shown in the above formula (7) as a material characteristic value prediction model, the normal Gaussian process regression cannot take into account the change in the variation in the material characteristic value caused by the difference in the operational performance. For this reason, in the first embodiment, a Gaussian process gP(0,K 1 ), gP(0,K 2 ) and uses a heteroscedastic Gaussian process (a Gaussian process that takes heteroscedasticity into account) in which the final prediction result is a Gaussian distribution with the output of each Gaussian process as a parameter.
number
[0038] By developing the above method, it is possible to estimate the covariance matrix Σ(x) between output variables, for example, as described below. Let D be the number of explanatory variables (input variables to the material property value prediction model), M be the number of objective variables (output variables from the material property value prediction model), and N be the number of records of actual data. For example, the past output actual data Y expressed by the following formula (13) is decorrelated based on formula (15) using the load matrix W expressed by the following formula (14) to obtain T expressed by the following formula (16).
number
number
[0039] That is, the output mean value vector μ corresponding to the new input x is expressed by the following equation (19). g (x) and the output standard deviation vector σ gBased on these prediction vectors, the output mean vector μ(x) and covariance matrix Σ(x) can be calculated by the following equations (21) and (22), respectively.
number
[0040] The material property value prediction model of the first embodiment is constructed by the method described above. The material property value prediction model is a model that takes the operating conditions in the manufacturing process of the target steel material as input and outputs the probability distribution of the material property values of the target steel material. When the correlation between output variables is not taken into consideration, the material property value prediction model is the model expressed by the above-mentioned formula (12). When the correlation between output variables is taken into consideration, the material property value prediction model is the model expressed by the above-mentioned formulas (18), (21), and (22).
[0041] FIG. 6 is a diagram for explaining an example of construction of a material characteristic value prediction model of the first embodiment in comparison with an example of construction of a conventional machine learning model. Specifically, for convenience of illustration, FIG. 6 shows an example of construction of a model that predicts tensile strength (TS) as a material characteristic value by plotting only the cooling temperature of the steel material in the rolling process among multiple parameters constituting the operating conditions. FIG. 6(a) shows an example of construction of a conventional machine learning model, and FIG. 6(b) shows an example of construction of a material characteristic value prediction model of the first embodiment. Data plotted with "●" in FIG. 6 is known data. In the conventional machine learning model shown in Fig. 6(a), the average value of the material characteristic value (TS) is output as the prediction result for the input operating condition (cooling temperature). Therefore, in the area surrounded by the dashed line in Fig. 6(a) where the material characteristic value varies widely even under the same operating condition, there is a problem that the prediction accuracy of the material characteristic value deteriorates. In contrast, the material characteristic value prediction model of the first embodiment shown in FIG. 6(b) outputs a probability distribution of the material characteristic value as a prediction result. Specifically, the material characteristic value prediction model of the present embodiment is a model using a non-uniform variance Gaussian process, and outputs a mean value and a standard deviation as a prediction result. The hatched area shown in FIG. 6(b) is a 95% prediction range (95% confidence interval) of a normal distribution obtained from the mean value and the standard deviation. As shown in FIG. 6(b), in the material characteristic value prediction model, an area with a large variation in the material characteristic value (an area corresponding to the area surrounded by a dashed line in FIG. 6(a)) is included in the probability distribution (95% prediction range) and is reflected as an output. In addition, in the material characteristic value prediction model, a probability distribution (95% prediction range) is output even in an area with no operating record (an area corresponding to the hatched area in FIG. 6(a)), so that the uncertainty of the prediction is taken into consideration.
[0042] <Constraint Condition Setting Section 4> The constraint condition setting unit 4 executes step ST13 shown in Fig. 3. Specifically, the constraint condition setting unit 4 sets constraint conditions for the back analysis executed by the operation condition determination unit 5 described below as necessary. The constraint condition setting unit 4 in the first embodiment sets upper and / or lower limit values of parameters constituting the operation conditions in the post-process to be determined (in the first embodiment, the rolling temperature and cooling temperature in the rolling process) as necessary. The constraint condition setting unit 4 may also set constraint conditions between parameters. In the first embodiment, from the viewpoint of operational feasibility, the constraint condition setting unit 4 sets, for example, a constraint condition between parameters constituting the operational conditions in the post-process, as shown in the following formula (23). Rolling temperature -400℃ ≦ Cooling temperature ≦ Rolling temperature (23)
[0043] <Operation condition determination section 5> The operating condition determination unit 5 first executes step ST14 shown in Fig. 3. Specifically, the operating condition determination unit 5 determines a target steel product to be subjected to back analysis from among a plurality of steel products. The operation condition determination unit 5 of the first embodiment, for example, as described below, determines, from among a plurality of steel materials, steel materials whose production volume per unit time (specifically, monthly production volume) is equal to or less than a predetermined threshold value as target steel materials from the viewpoint of sufficiently improving productivity of steel materials. In the example of the order information shown in Fig. 4, for example, if the predetermined threshold value is a monthly production volume of 500 tons, then, from among a plurality of steel materials corresponding to order IDs K01, K02, K03, and K04 (hereinafter, the respective steel materials will be referred to as steel material K01, steel material K02, steel material K03, and steel material K04), steel materials K01 and K02 whose monthly production volume is 200 tons will be determined as target steel materials. In the above example, the case where the operating condition determination unit 5 automatically determines the target steel material has been described as an example, but the present invention is not limited to this, and it is also possible to adopt a mode in which an operator determines the target steel material and inputs the determined target steel material to the operating condition determination unit 5. In this case, it is possible to determine the target steel material by making a comprehensive judgment including not only the monthly production volume but also the application and manufacturing cost of each steel material.
[0044] Next, the operating condition determination unit 5 executes step ST15 shown in Fig. 3. Specifically, as outlined with reference to Fig. 1, the operating condition determination unit 5 determines at least the operating conditions in the post-process (in the first embodiment, the rolling temperature and the cooling temperature in the rolling process, which is the post-process) in the manufacturing process of the target steel materials K01 and K02 by performing a reverse analysis of the material property value prediction model acquired by the material property value prediction model acquisition unit 3 so that the pass probability of the material property value of the target steel materials K01 and K02 shown in Fig. 4 is equal to or higher than a predetermined threshold value. At this time, the operating condition determination unit 5 sets the component values in the refining process of the target steel materials K01 and K02 to the component values of the existing steel type B set in the steel material K03 other than the target steel materials K01 and K02, or the component values of the existing steel type C set in the steel material K04 other than the target steel materials K01 and K02, and reverse analyzes the material property value prediction model. In the first embodiment, since the steel materials K01 and K02 are the target steel materials, the component values in the refining process of the target steel materials K01 and K02 are set to the component values of the existing steel type B set in the steel material K03 other than the target steel materials K01 and K02, or the component values of the existing steel type C set in the steel material K04 other than the target steel materials K01 and K02, and the material characteristic value prediction model is reverse analyzed, but the present invention is not limited to this. For example, it is also possible to adopt a mode in which the steel materials K03 and K04 are the target steel materials, the component values in the refining process of the target steel materials K03 and K04 are set to the component values of the existing steel type A set in the steel materials K01 and K02 other than the target steel materials K03 and K04, and the material characteristic value prediction model is reverse analyzed. The contents of the back analysis executed by the operating condition determination unit 5 will be specifically described below.
[0045] As described above, by using the material property value prediction model, when the operating condition x expressed by the following formula (24) is given, the material property value y (tensile strength y ts yield point y yp ) can be output (using equation (12) or using equations (18), (21) and (22)).
number
number
[0046] In the above formula (26), y S upper [k], y S lower [k] is the upper and lower limits of the tensile strength (TS) of the target steel material k (k = K01 or K02) determined by the manufacturing specifications when S = ts. Also, y S upper [k], y S lower [k] is the upper and lower limit values of the yield point (YP) of the target steel material k, determined by the manufacturing specifications, when S = yp. If the manufacturing specifications do not include either the upper limit or the lower limit, it is sufficient to set the upper limit to a sufficiently large value and the lower limit to a sufficiently small value. In addition, in the above formula (26), μ S (x[k]), σ S (x[k]) is the material property value y of the target steel material k output from the material property value prediction model. S Probability distribution p(yS |x[k]) are the mean vector and standard deviation vector of In the above formulas (27) and (29), u rt [k] is the rolling temperature of the target steel material k. In the above formulas (27) and (29), u ct [k] is the cooling temperature of the target steel material k. In the above formula (28) and formula (31), v t [k] is the plate thickness of the target steel material k in the manufacturing specifications. In the above formula (30), v C is the carbon content of existing steel grades B and C, ν Si is the silicon content of existing steel grades B and C, and ν Mn is the manganese content of existing steel grades B and C. The ν in the above formula (28) is a parameter group set by fixing the component values of the existing steel type B or C, as shown in the above formula (30). The ν in the above formula (28) t [k] is a group of parameters that are fixed to the plate thickness of the manufacturing specification of the target steel material k, as shown in the above formula (31). t [k] is treated as a constant. The above formula (27) is the constraint condition set by the constraint condition setting unit 4.
[0047] In the above formula (26), an example is given where there are two target steel materials, K01 and K02, so the exponent located at the rightmost position in formula (26) is "1 / 2", but generally, when there are Q target steel materials, the exponent is "1 / Q". In addition, in the above formula (26), an example is given where there are two material characteristic values, TS and YP, so the exponent located second from the right in formula (26) is "1 / 2", but generally, when there are R material characteristic values, the exponent is "1 / R". In addition, in the above description, the constrained trust region method is used as the optimization method, but this is not limited to this, and other optimization methods such as differential evolution method and particle swarm optimization can also be used.
[0048] <Operation condition output section 6> The operating condition output unit 6 executes step ST16 shown in Fig. 3. Specifically, the operating condition output unit 6 outputs the operating conditions (rolling temperature and cooling temperature in the rolling process) in the downstream process of the target steel materials K01, K02 determined by the operating condition determination unit 5. The operating condition output unit 6 not only outputs the operating conditions in the downstream process of the determined target steel materials K01, K02, but also preferably outputs the probability distribution (average value, standard deviation) of the material characteristic value output from the material characteristic value prediction model when the manufacturing specifications of the target steel materials K01, K02 and the operating conditions (including the component values in the refining process) in the manufacturing process of the determined target steel materials K01, K02 are input to the material characteristic value prediction model, and the pass probability of the material characteristic value of the target steel materials K01, K02 calculated from this probability distribution and the manufacturing specifications. In addition, the operating condition output unit 6 may also output the operating conditions in the downstream processes of the target steel materials K01, K02 where the pass probability P(u) calculated by the operating condition determination unit 5 by inverse analysis of the material characteristic value prediction model is not greater than a predetermined threshold value (is less than the threshold value).
[0049] FIG. 7 is a diagram showing an example of the contents output from the operating condition output unit 6. The numerical values described in "Post-process operating conditions", "YP" and "TS" in the column P1 shown in FIG. 7 are values obtained when the component values in the refining process of the target steel materials K01 and K02 are set to the component values of the existing steel type B set in the steel material K03 and the material characteristic value prediction model is reverse-analyzed, and the numerical values described in "Post-process operating conditions", "YP" and "TS" in the column P2 are values obtained when the component values in the refining process of the target steel materials K01 and K02 are set to the component values of the existing steel type C set in the steel material K04 and the material characteristic value prediction model is reverse-analyzed. The numerical value described in "Pass probability" shown in FIG. 7 is the geometric mean of the maximum pass probability of each material characteristic value (YP, TS) calculated for each of the target steel materials K01 and K02. For example, if the aforementioned threshold value (threshold value to be compared with the pass probability P(u)) is 0.98, the geometric mean of the pass probability (0.945, 0.967) described in the "Pass Probability" column P1 shown in FIG. 7 for all the target steel materials K01 and K02 is less than 0.98, which is less than the above threshold value. On the other hand, the geometric mean of the pass probability (0.997, 0.998) described in the "Pass Probability" column P2 shown in FIG. 7 for all the target steel materials K01 and K02 is 0.98 or more, which is greater than the above threshold value. Therefore, the numerical value described in the "Post-process operating conditions" column P2 shown in FIG. 7 becomes the operating conditions in the post-process of the target steel materials determined by the operating condition determination unit 5, and the operating conditions for consolidating the steel types of the target steel materials K01 and K02 into the existing steel type C set for the steel material K04 are determined.
[0050] As described above, according to the operating condition determination device 100 of the first embodiment, for multiple steel materials K01, K02, and K04 having different manufacturing specifications, the operating conditions determined are appropriate in accordance with the manufacturing specifications (the pass probability of the material characteristic values of the target steel materials K01, K02 is equal to or greater than a predetermined threshold value), and the component values in the refining process of the multiple steel materials K01, K02, and K04 are set to be identical to each other (set to the component values of the existing steel type C), so that the number of consecutive pieces of the same steel type C can be increased, thereby enabling improvement in the productivity and quality of the steel materials. In the first embodiment, as shown in FIG. 4, steel materials K01 and K02, which have a combined monthly production volume of only 200 tons, are treated as the target steel materials, and these steel types are consolidated into steel material K04, steel type C, which has a monthly production volume of 1,800 tons. As a result, the monthly production volume of the same steel type C becomes 2,000 tons, making it possible to sufficiently improve the productivity of steel materials.
[0051] [Second embodiment] The operation condition determination device according to the second embodiment has the same configuration as the operation condition determination device 100 according to the first embodiment, except for the calculation contents of the operation condition determination unit 5. For this reason, Fig. 2 and Fig. 3 are also used for the operation condition determination device according to the second embodiment, and the reference numerals given to the operation condition determination device and its components in Fig. 2 are also used. The operating condition determination device 100 according to the above-described first embodiment is configured to determine operating conditions for consolidating steel grades into existing steel grades. In contrast, the operating condition determination device 100 according to the second embodiment is configured to determine operating conditions for consolidating steel types to steel types not present in the existing steel types. That is, in the operating condition determination device 100 according to the second embodiment, the component values in the refining process that are set to be the same are component values not present in the existing steel types, and the component values in the refining process and the operating conditions in the subsequent process (rolling process) among the manufacturing processes of the target steel material are determined by performing reverse analysis of the material characteristic value prediction model. Hereinafter, each component of the operating condition determination device 100 according to the second embodiment will be described. However, explanations of the order information acquisition unit 1, the material characteristic value prediction model storage unit 2, the material characteristic value prediction model acquisition unit 3 and the constraint condition setting unit 4, which are the same as those of the operating condition determination device 100 according to the first embodiment, will be omitted. Instead, an operating condition determination unit 5 and an operating condition output unit 6, which have calculation contents different from those of the operating condition determination device 100 according to the first embodiment, will be described.
[0052] <Operation condition determination section 5> As in the first embodiment, the operating condition determination unit 5 first executes step ST14 shown in Fig. 3. Specifically, the operating condition determination unit 5 determines a target steel material to be subjected to back analysis from among a plurality of steel materials. The operating condition determination unit 5 of the second embodiment, unlike the first embodiment, determines operating conditions for consolidating steel types to steel types not included in the existing steel types, and therefore determines all of the multiple steel products for which operating conditions are to be determined as target steel products. In the example of the order information shown in Fig. 4, all of the steel products K01, K02, K03, and K04 are determined as target steel products.
[0053] Next, the operating condition determination unit 5 executes step ST15 shown in Fig. 3. Specifically, as outlined with reference to Fig. 1, the operating condition determination unit 5 determines the component values in the refining process and the operating conditions in the subsequent process (the component values in the refining process and the rolling temperature and the cooling temperature in the rolling process, which is the subsequent process) in the manufacturing process of the target steel materials K01, K02, K03, and K04 by performing a reverse analysis of the material characteristic value prediction model acquired by the material characteristic value prediction model acquisition unit 3 so that the pass probability of the material characteristic values of the target steel materials K01, K02, K03, and K04 shown in Fig. 4 is equal to or higher than a predetermined threshold value. The contents of the back analysis executed by the operating condition determination unit 5 will be specifically described below.
[0054] The operating condition determination unit 5 sets the manufacturing specifications (plate thickness, required upper and lower limits of YP, required upper and lower limits of TS) as the manufacturing specifications of the target steel materials K01, K02, K03, and K04, and then searches for and determines the operating conditions x that give an acceptance probability equal to or greater than a predetermined threshold value. That is, the operating condition determination unit 5 determines the component value ν in the refining process and the operating conditions u in the subsequent process (rolling process) by solving the optimization problem P2 formulated, for example, as in the following equations (32) to (37). In other words, the operating condition determination unit 5 determines the component value v (common to all target steel materials k) in the refining process and the operating condition u (specifically, the operating condition u[k] for each target steel material k) in the post-process (rolling process) that makes the pass probability P(u, v) (the geometric mean of the pass probability of each material characteristic value for each target steel material k, the geometric mean of the geometric mean for all target steel materials k) expressed by the following formula (32) equal to or greater than a predetermined threshold value. In the second embodiment, unlike the first embodiment, not only the operating condition u[k] in the post-process (rolling process) shown in formula (35) is used as a decision variable, but also the component value v in the refining process shown in formula (36) is used as a decision variable (in the first embodiment, the component value v in the refining process shown in formula (30) is a constant determined by the existing steel type), so that the operating condition for consolidating the steel type to a steel type not included in the existing steel types is determined.
number
[0055] In the above formula (32), y S upper [k], y S lower [k] is the upper and lower limits of the tensile strength (TS) of the target steel material k (k = K01, K02, K03 or K04) determined by the manufacturing specifications when S = ts. Also, y S upper [k], y S lower [k] is the upper and lower limit values of the yield point (YP) of the target steel material k, determined by the manufacturing specifications, when S = yp. If the manufacturing specifications do not include either the upper limit or the lower limit, it is sufficient to set the upper limit to a sufficiently large value and the lower limit to a sufficiently small value. In addition, in the above formula (32), μ S (x[k]), σ S (x[k]) is the material property value y of the target steel material k output from the material property value prediction model. S Probability distribution p(y S |x[k]) are the mean vector and standard deviation vector of In the above formulas (33) and (35), u rt [k] is the rolling temperature of the target steel material k. In the above formulas (33) and (35), u ct [k] is the cooling temperature of the target steel material k. In the above formula (34) and formula (37), v t [k] is the plate thickness of the target steel material k in the manufacturing specifications. In the above formula (36), v C is the carbon content, ν Si is the component value of silicon, ν Mn is the manganese content. ν shown in the above formula (34) t [k] is a group of parameters that are fixed to the plate thickness of the manufacturing specification of the target steel material k, as shown in the above formula (37). t [k] is treated as a constant. The above formula (33) is the constraint condition set by the constraint condition setting unit 4.
[0056] In the above formula (32), an example is given where there are four target steel materials, K01, K02, K03, and K04, so the exponent located at the rightmost position in formula (32) is "1 / 4", but generally, when there are Q target steel materials, the exponent is "1 / Q". In addition, in the above formula (32), an example is given where there are two material characteristic values, TS and YP, so the exponent located second from the right in formula (32) is "1 / 2", but generally, when there are R material characteristic values, the exponent is "1 / R". In addition, in the above description, the constrained trust region method is used as the optimization method, but this is not limited to this, and other optimization methods such as differential evolution method and particle swarm optimization can also be used.
[0057] <Operation condition output section 6> The operating condition output unit 6 executes step ST16 shown in Fig. 3, similarly to the first embodiment. Specifically, the operating condition output unit 6 outputs the component values in the refining process of the target steel materials K01, K02, K03, and K04, and the operating conditions in the post-process (rolling temperature and cooling temperature in the rolling process) determined by the operating condition determination unit 5. It is preferable that the operating condition output unit 6 not only outputs the component values and operating conditions in the subsequent processes in the refining process of the determined target steel materials K01, K02, K03, K04, but also outputs the manufacturing specifications of the target steel materials K01, K02, K03, K04, and the probability distribution (average value, standard deviation) of the material characteristic values output from the material characteristic value prediction model when the operating conditions in the manufacturing process of the determined target steel materials K01, K02, K03, K04 (component values in the refining process and operating conditions in the subsequent processes) are input into the material characteristic value prediction model, and the probability of passing the material characteristic values of the target steel materials K01, K02, K03, K04 calculated from this probability distribution and the manufacturing specifications. In addition, the operating condition output unit 6 may also output the component values in the refining process and the operating conditions in the subsequent processes of the target steel materials K01, K02, K03, and K04 whose pass probability P(u, ν) is not equal to or greater than a predetermined threshold value, as calculated by the operating condition determination unit 5 by performing inverse analysis of the material characteristic value prediction model.
[0058] Fig. 8 is a diagram showing an example of the contents output from the operating condition output unit 6. The numerical value described in the "Pass probability" shown in Fig. 8 is the geometric mean of the maximum pass probability of each material characteristic value (YP, TS) calculated for each of the target steel materials K01, K02, K03, and K04. For example, if the aforementioned threshold value (threshold value to be compared with the pass probability P(u,ν)) is 0.98, the geometric mean of the pass probabilities (0.982, 0.980, 0.981, 0.981) described in the "Pass Probability" shown in Figure 8 for all the target steel materials K01, K02, K03, and K04 is 0.98 or more, which is greater than the above threshold value. Therefore, the numerical values described in the "Refining Process Operation Conditions" and "Post-Process Operation Conditions" shown in Figure 8 become the operation conditions in the post-process of the target steel material determined by the operation condition determination unit 5, and the operation conditions for consolidating the steel materials into a new steel material D having component values of C component: 10, Si component: 19, and Mn component: 125, which are not present in the existing steel materials, are determined.
[0059] As described above, according to the operating condition determination device 100 of the second embodiment, for multiple steel materials K01, K02, K03, and K04 having different manufacturing specifications, the operating conditions determined are appropriate in accordance with the manufacturing specifications (the pass probability of the material characteristic values of the target steel materials K01 and K02 is equal to or greater than a predetermined threshold value), and the component values in the refining process of the multiple steel materials K01, K02, K03, and K04 are set to be identical to each other (set to the component values of the existing steel type D), so that the number of consecutive pieces of the same steel type D can be increased, thereby making it possible to improve the productivity and quality of the steel materials. In the second embodiment, as shown in FIG. 4, the target steel materials are steel materials K01 and K02, with a combined monthly production volume of 200 tons, and steel materials K03 and K04, with a combined monthly production volume of 1,800 tons, respectively, and these steel types are all consolidated into steel type D, so that the monthly production volume of the same steel type D becomes 3,800 tons, making it possible to sufficiently improve the productivity of steel materials.
[0060] In addition, if the pass probability P(u,ν) when all of the multiple steel materials K01, K02, K03, and K04 are treated as target steel materials is less than a predetermined threshold value, then the multiple steel materials are appropriately reselected from the multiple steel materials, in this example three or less target steel materials, and the optimization problem P2 is solved again by the operating condition determination unit 5, and component values in the refining process and operating conditions in the subsequent processes are determined so that the pass probability P(u,ν) is equal to or greater than the predetermined threshold value. It is also possible to adopt a configuration in which, from the beginning, the operating condition determination unit 5 solves the optimization problem P2 for all possible combination patterns of target steel materials {K01, K02}, {K01, K03}, {K01, K04}, {K02, K03}, {K02, K04}, {K03, K04}, {K01, K02, K03}, {K01, K02, K04}, {K01, K03, K04}, {K02, K03, K04}, {K01, K02, K03, K04}, and determines the component values in the refining process and the operating conditions in the subsequent process for the combination patterns of target steel materials for which the pass probability P(u, ν) is equal to or greater than a predetermined threshold value.
[0061] The first embodiment described above is a configuration for determining the operating conditions for aggregating steel types to existing steel types, and the second embodiment is a configuration for determining the operating conditions for aggregating steel types to steel types not included in the existing steel types, but the present invention is not limited to either of these configurations. For example, it is also possible to adopt a configuration that combines the first and second embodiments, in which, among multiple steel materials for which operating conditions are determined in a manufacturing process, operating conditions for aggregating steel types to existing steel types are determined for some steel materials, and operating conditions for aggregating steel types to steel types not included in the existing steel types are determined for the remaining steel materials.
[0062] In the embodiment described above, the operation condition determination apparatus 100 may be implemented using a plurality of computers. For example, the operation condition determination apparatus 100 may be implemented using a device such as a cloud server. Furthermore, the hardware processor and memory of the operation condition determination apparatus 100 may be distributed and implemented in a plurality of computers.
[0063] Although one embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to this embodiment, and designs that do not deviate from the gist of the present invention are also included. [Explanation of symbols]
[0064] 1. Order information acquisition unit 2. Material property value prediction model storage section 3. Material property value prediction model acquisition section 4. Constraint Condition Setting Section 5. Operation condition determination section 6. Operation condition output section 100 Operation condition determination device
Claims
1. An operating condition determination device for determining operating conditions in a manufacturing process for a plurality of steel materials having different manufacturing specifications, the manufacturing process being produced through a manufacturing process having a refining process and a post-process subsequent to the refining process, under a condition that component values in the refining process of the plurality of steel materials are set to be the same for the plurality of steel materials, A material characteristic value prediction model is provided that inputs an operating condition in the manufacturing process of a target steel material to be subjected to reverse analysis among the plurality of steel materials, and outputs a probability distribution of material characteristic values of the target steel material, The operating conditions in at least the subsequent process of the manufacturing process of the target steel are determined by reverse analysis of the material property value prediction model so that the pass probability of the material property value of the target steel calculated from the probability distribution of the material property value of the target steel output when the operating conditions in the manufacturing process of the target steel, including the component values in the refining process set to be the same, are input into the material property value prediction model, and the specifications of the material property value included in the manufacturing specifications of the target steel, is equal to or greater than a predetermined threshold value. Operation condition determination device.
2. The component values in the refining process that are set to be the same are component values of existing steel types that are set for steel materials other than the target steel material among the plurality of steel materials, By performing a reverse analysis of the material characteristic value prediction model, an operating condition for the post-process in the manufacturing process of the target steel material is determined. The operating condition determining device according to claim 1 .
3. The component values in the refining process that are set to be the same are component values that do not exist in existing steel types, By performing a reverse analysis of the material characteristic value prediction model, the component values in the refining process and the operating conditions in the subsequent process are determined among the manufacturing processes of the target steel material. The operating condition determining device according to claim 1 .
4. The material characteristic value prediction model is a model using Bayesian estimation.
4. An operating condition determining device according to claim 1.
5. The material characteristic value prediction model is a model using a heteroscedastic Gaussian process.
4. An operating condition determining device according to claim 1.
6. A method for determining operating conditions in a manufacturing process for a plurality of steel materials having different manufacturing specifications, the manufacturing process being produced through a refining process and a post-process subsequent to the refining process, under a condition that component values in the refining process of the plurality of steel materials are set to be the same for the plurality of steel materials, the method comprising: A material property value prediction model is used, which inputs the operating conditions in the manufacturing process of a target steel material to be subjected to the reverse analysis among the plurality of steel materials, and outputs a probability distribution of the material property value of the target steel material; The operating conditions in at least the subsequent process of the manufacturing process of the target steel are determined by reverse analysis of the material property value prediction model so that the pass probability of the material property value of the target steel calculated from the probability distribution of the material property value of the target steel output when the operating conditions in the manufacturing process of the target steel, including the component values in the refining process set to be the same, are input into the material property value prediction model, and the specifications of the material property value included in the manufacturing specifications of the target steel, is equal to or greater than a predetermined threshold value. How to determine operating conditions.
7. A program for causing a computer to function as the operation condition determination device according to claim 1.
Citation Information
Patent Citations
Operation analysis device, operation analysis method, computer program and computer-readable storage medium
JP2011145764A
Determination method for production specification of metallic material, production method of metallic material and determination device for production specification
JP2020114597A
Manufacturing specification determination support device, manufacturing specification determination support method, computer program, and computer readable recording medium
JP2022014878A
Operation condition determination device, operation condition determination method, and program
JP2024151608A
Design assistance method for metal material, prediction model generation method, metal material manufacturing method, and design assistance device
WO2020152993A1
