Rolling material determination device, rolling material determination method, and method for manufacturing plate product

Abstract

To provide a rolling material determination device, a rolling material determination method, and a method for manufacturing a plate product capable of accurately predicting a rolling allowance length, the required slab weight, and slab yield.SOLUTION: A rolling material determination device is used for predicting a rolling allowance length when an assembled product, composed of one or more plate products combined together, is extracted from a large plate obtained by rolling a slab, and includes a prediction unit that predicts the rolling allowance length by inputting a set dimension of the large plate, a calculated dimension of the slab calculated from the set dimension of the large plate, and rolling conditions of the large plate into a prediction model that has been trained using, as explanatory variables, past actual values of the set dimension of the large plate, the dimension of the slab, and the rolling conditions of the large plate, and, as an objective variable, past actual values of the rolling allowance length thereto.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a rolling material determination device, a rolling material determination method, and a method for manufacturing a plate product. [Background technology]

[0002] In manufacturing industries such as the steel industry, when planning the production of plate products, it is common to allocate an organized product made by combining one or more plate products to one large plate rolled from a slab, and to plan the production so that one or more plate products can be obtained from one large plate.

[0003] For example, Patent Document 1 listed below discloses a cutting determination device that groups (organizes) products with the same or similar optimal rolled plate thickness, and determines the cutting size, rolling dimensions, and required slab dimensions and weight of the thick steel plate to be rolled based on the dimensions of the grouped products. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2006-281252 Summary of the Invention [Problem to be solved by the invention]

[0005] Here, the large plates to which the assembled products are assigned require a certain amount of excess length relative to the dimensions of the assembled products. Conventionally, a yield master is set in advance, taking into account the accuracy of the finished plate thickness, the accuracy of the finished dimensions, longitudinal variations, scale peeling in the heating furnace, etc. Then, a method has been adopted in which the corresponding master values ​​are read from the dimensions of the assembled large plate and slab, and the yield is calculated based on the master values, and the required slab weight is calculated.

[0006] For this reason, when evaluating a single large plate, the excess length may not be the appropriate length, and if the excess length is too short or insufficient, the reject rate of the plate product increases, while if the excess length is too long, the yield decreases.

[0007] The present invention has been made in consideration of the above, and aims to provide a rolling material determination device, a rolling material determination method, and a method for manufacturing a plate product that can accurately predict the rolling excess length, the required slab weight, and the slab yield. [Means for solving the problem]

[0008] In order to solve the above-mentioned problems and achieve the object, the rolling material determination device of the present invention is a rolling material determination device that predicts a rolling excess length when an organized product is obtained by combining one or more plate material products from one large plate obtained by rolling a slab, and is equipped with a prediction unit that predicts the rolling excess length by inputting the set dimensions of the large plate, the calculated dimensions of the slab calculated from the set dimensions of the large plate, and the rolling conditions of the large plate into a prediction model that has been trained using the set dimensions of the large plate, the dimensions of the slab, and past actual values ​​of the rolling conditions of the large plate as explanatory variables, and the corresponding past actual values ​​of the rolling excess length as objective variables.

[0009] In addition, the rolling material determination device according to the present invention, in the above invention, further includes a determination unit that determines the required weight of the slab based on the difference between the rolling excess length predicted by the prediction unit and a preset target excess length.

[0010] Furthermore, in the rolling material determination device according to the present invention, in the above invention, the determination unit calculates a yield correction amount for correcting the yield of the slab based on the difference between the rolling excess length and the target excess length and the set dimensions of the large plate, and determines the required weight of the slab based on the calculated yield correction amount.

[0011] Moreover, in the rolling material determination device according to the present invention, in the above invention, the determination unit determines whether or not the rolling surplus length predicted by the prediction unit is within a range of a preset threshold value, and if the rolling surplus length is within the range of the threshold value, the rolling surplus length predicted by the prediction unit is used as is to calculate a difference from the target surplus length, and if the rolling surplus length is outside the range of the threshold value, the numerical value closest to the rolling surplus length predicted by the prediction unit, among numerical values ​​within the range of the threshold value, is regarded as the rolling surplus length, and the difference from the target surplus length is calculated using the threshold value.

[0012] In addition, in the rolling material determination device of the present invention, in the above invention, the determination unit determines the required weight of the slab, and then determines whether the determined weight of the slab satisfies the equipment constraints, and if the equipment constraints are not satisfied, recalculates the weight of the slab that satisfies the equipment constraints.

[0013] In order to solve the above-mentioned problems and achieve the object, the rolling material determination device of the present invention is a rolling material determination device that predicts the weight of the slab or the yield of the slab, which is required when obtaining an organized product that is made by combining one or more plate material products from a single large plate obtained by rolling a slab, and is equipped with a prediction unit that predicts the weight of the slab or the yield of the slab by inputting the set dimensions of the large plate, the calculated dimensions of the slab calculated from the set dimensions of the large plate, and the rolling conditions of the large plate into a prediction model that has been trained using the set dimensions of the large plate, the dimensions of the slab, and past actual values ​​of the rolling conditions of the large plate as explanatory variables, and the corresponding past actual values ​​of the weight of the slab or the yield of the slab as objective variables.

[0014] In order to solve the above-mentioned problems and achieve the object, the rolling stock determination method of the present invention is a rolling stock determination method for predicting a rolling excess length when an organized product is obtained by combining one or more plate material products from a single large plate obtained by rolling a slab, and includes a prediction step for predicting the rolling excess length by inputting the set dimensions of the large plate, the calculated dimensions of the slab calculated from the set dimensions of the large plate, and the rolling conditions of the large plate into a prediction model that has been trained using the set dimensions of the large plate, the dimensions of the slab, and past actual values ​​of the rolling conditions of the large plate as explanatory variables, and the past actual values ​​of the corresponding rolling excess length as objective variables.

[0015] In order to solve the above-mentioned problems and achieve the objectives, the method for manufacturing a plate material product of the present invention predicts a rolling excess length using the above-mentioned rolling material determination method, rolls a slab whose weight is determined based on the predicted rolling excess length to produce a large plate, and extracts one or more plate material products from the large plate. [Effects of the Invention]

[0016] According to the rolling material determination device, rolling material determination method, and plate product manufacturing method of the present invention, by using a prediction model to predict the rolling excess length, the required slab weight, and the slab yield, it is possible to accurately predict the rolling excess length, the required slab weight, and the slab yield. [Brief explanation of the drawings]

[0017] [Figure 1] FIG. 1 is a diagram showing how knitted products are allocated to large boards. [Figure 2] FIG. 2 is a diagram showing an example of an information processing device for realizing a rolling material determination device according to an embodiment of the present invention. [Figure 3] FIG. 3 is a diagram showing an example of variables to be learned when constructing a prediction model in a rolling material determination device, a rolling material determination method, and a plate product manufacturing method according to an embodiment of the present invention, and the influence of each variable on prediction accuracy. DETAILED DESCRIPTION OF THE INVENTION

[0018] An embodiment of a rolling material determination device, a rolling material determination method, and a plate product manufacturing method according to the present invention will be described with reference to the drawings. The present invention relates to a rolling material determination device, a rolling material determination method, and a plate product manufacturing method when manufacturing plate products such as thick steel plates, and is particularly intended to improve and optimize the yield when manufacturing plate products.

[0019] <Allocation of knitted products to large boards> In this embodiment, a case will be described in which a plurality of plate products (for example, products A to E) are obtained from a single large plate (rolled large plate) obtained by rolling a slab, which is a rolling material, as shown in Fig. 1. The products A to E are pre-combined by manual assembly, in which the plate products are assembled by hand, or by machine assembly, in which the plate products are mechanically assembled, and the assembled product obtained by combining the products A to E is assigned to one large plate.

[0020] In this case, the difference between the effective length of the large plate and the total product length (effective length of large plate - assembled product length) is called "rolling excess length (excess length)." If the rolling excess length is too long, the product yield will decrease, and if it is too short or insufficient, the product reject rate will increase. Therefore, it is necessary to accurately predict the excess length and optimize it to the target excess length. Here, "yield" refers to the percentage of the large plate that can be produced as a product. Therefore, in this embodiment, when performing rolling stock calculations, the rolling excess length is predicted using a rolling stock determination device, and the size (dimensions, weight) of the slab (rolling stock) is determined based on the predicted rolling excess length.

[0021] <Rolling material determination device> 2 shows an example of the configuration of an information processing device 1 for realizing the rolling material determination device according to this embodiment. This information processing device 1 is realized by a general-purpose computer such as a workstation or a personal computer. The information processing device 1 also includes an input unit 10, a storage unit 20, a calculation unit 30, and an output unit 40.

[0022] The input unit 10 is an input means for the calculation unit 30, and is realized by an input device such as a keyboard, a mouse pointer, a numeric keypad, etc. The input unit 10 inputs information necessary for various calculations in the calculation unit 30.

[0023] The storage unit 20 is realized by a recording medium such as an EPROM (Erasable Programmable ROM), a hard disk drive (HDD), or a removable medium. Examples of the removable medium include a disk recording medium such as a USB (Universal Serial Bus) memory, a CD (Compact Disc), a DVD (Digital Versatile Disc), or a BD (Blu-ray (registered trademark) Disc).

[0024] The storage unit 20 can store an operating system (OS), various programs, various tables, various databases, etc. The storage unit 20 stores an operation DB 21 and a prediction model 22.

[0025] Operation data acquired (collected) during past operations is stored in the operation DB 21. Examples of this operation data include the set dimensions of large plates, slab dimensions, rolling conditions for large plates, excess rolling length, etc.

[0026] The prediction model 22 is constructed by the learning unit 31 and stored in the storage unit 20 .

[0027] The calculation unit 30 is realized by a processor such as a CPU (Central Processing Unit) and a memory (main storage unit) such as a RAM (Random Access Memory) or a ROM (Read Only Memory).

[0028] The calculation unit 30 loads a program into the working area of ​​the main memory and executes it, and controls each component through the execution of the program to realize functions that meet a predetermined purpose. Through the execution of the program, the calculation unit 30 functions as a learning unit 31, a prediction unit 32, and a determination unit 33. Note that while FIG. 2 shows an example in which the functions of each unit of the calculation unit 30 are realized by a single computer, the method for realizing the functions of each unit is not particularly limited, and for example, the functions of each unit may be realized by multiple computers. Also, some functions of the calculation unit 30 may be provided on the cloud.

[0029] The learning unit 31 constructs the prediction model 22 through machine learning using operation data acquired from the operation DB 21. The learning unit 31 constructs the prediction model 22 by learning past actual values ​​of the set dimensions of the large plate, the slab dimensions, the rolling conditions of the large plate, and the rolling excess length, with the set dimensions of the large plate, the slab dimensions, and the rolling conditions of the large plate as explanatory variables (input variables) and the corresponding rolling excess length as a target variable (output variable). The learning unit 31 then stores the constructed prediction model 22 in the storage unit 20. Examples of machine learning techniques implemented by the learning unit 31 include linear regression, neural networks, decision trees, random forests, support vector regression, and gradient boosting.

[0030] The prediction unit 32 predicts the rolling excess length (effective length of large plate - knitted product length) by inputting the set dimensions of the large plate, the calculated dimensions of the slab, and the rolling conditions of the large plate into the prediction model 22 constructed by the learning unit 31.

[0031] The determination unit 33 determines the necessary weight (dimension) of the slab based on the difference between the rolling excess length predicted by the prediction unit 32 and a preset target excess length.

[0032] The determination unit 33 may calculate a yield correction amount for correcting the yield of the slab based on the difference between the rolling excess length predicted by the prediction unit 32 and a predetermined target excess length, and the set dimensions of the large plate, and may determine the required weight of the slab based on the calculated yield correction amount.

[0033] Furthermore, the determination unit 33 may determine whether the rolling excess length predicted by the prediction unit 32 is within a preset threshold range or not, and when it is determined that the rolling excess length is within the threshold range, the rolling excess length predicted by the prediction unit 32 may be used as is to calculate the difference from the target excess length. When it is determined that the rolling excess length predicted by the prediction unit 32 is outside the threshold range, the determination unit 33 may regard the numerical value within the threshold range that is closest to the rolling excess length predicted by the prediction unit 32 as the rolling excess length, and may calculate the difference from the target excess length using the threshold. Note that "the numerical value within the threshold range that is closest to the rolling excess length predicted by the prediction unit 32" refers to the upper limit or lower limit of the range of the threshold values.

[0034] In addition, after determining the weight of the required slab, the determination unit 33 determines whether the determined weight of the slab satisfies the equipment constraints, and if it determines that the equipment constraints are not satisfied, it may recalculate the weight of the slab that satisfies the equipment constraints.

[0035] The output unit 40 is an output means for outputting the calculation results by the calculation unit 30. This output unit 40 is realized by an input device such as a display, a printer, etc. The output unit 40 also outputs, for example, the rolling excess length predicted by the prediction unit 32, the weight of the slab determined by the determination unit 33, etc.

[0036] <Rolling material selection method> The rolling material determination method according to this embodiment includes a learning step, a prediction step, and a determination step. Each step will be described in detail below. The learning step of the rolling material determination method may be performed in advance at a timing different from that of the prediction step and the determination step.

[0037] (Learning Steps) In the learning step, the learning unit 31 constructs a prediction model 22 by learning the past actual values ​​of the set dimensions of the large plate, the slab dimensions, the rolling conditions of the large plate, and the rolling excess length, using the set dimensions of the large plate, the slab dimensions, and the rolling conditions of the large plate as explanatory variables and the corresponding rolling excess length as a target variable.

[0038] In the learning step, the following variables are learned, as shown in Figure 3. This figure also shows some of the variables learned when building the prediction model 22 and the influence of these variables on the prediction accuracy.

[0039] [Set dimensions of large board] (1) Knitting dimension_length value: length of the knitted large board (2) Representative nominal thickness: Representative product thickness of knitted product (3) Knitted dimension_thickness value: ⇒ Thickness of the knitted large plate

[0040] Slab Dimensions (1) Width and length reversal sign: Whether the slab has been rotated 90 degrees and the width and length of the slab have been swapped. (2) Slab weight value: Weight of the slab (3) Slab dimension_length value: length of the slab (4) Hot / cold slab classification: Classification of whether the slab is a hot slab (charged directly into the heating furnace as cast) or not

[0041] [Large plate rolling conditions] (1) Lengthwise reduction ratio: (slab thickness × slab width) / (plate thickness × plate width) (2) Rolling method: How many mm of reduction is required for slab rolling? (3) Finishing temperature: The predicted temperature of the plate at the end of rolling. (4) Unheated tensile strength classification: Classification indicating the tensile strength (tensile strength) of the large plate (5) Requested steel type: Slab composition (6) Temperature control classification (CB): Whether or not temperature control is performed by strictly adhering to temperature control conditions during rolling. (7) Temperature control category (CXB): Whether or not temperature control is performed to strictly adhere to the finishing temperature during rolling. (8) Different thickness and width code: Whether or not large plates are rolled to have width deviation in the longitudinal direction. (9) Tensile strength after heat treatment: Tensile strength of the large plate after heat treatment (10) Unheated tensile strength classification: tensile strength of the large plate before heat treatment (11) Heating furnace - heating temperature: Conditions for heating the slab in the heating furnace (12) Cooling section: Whether or not to cool (accelerated cooling) after rolling is completed (13) Reference yield - corrected value: The yield tentatively determined by referring to the yield master.

[0042] (Prediction step) In the prediction step, the prediction unit 32 predicts the rolling excess length by inputting the set dimensions of the large plate, the calculated dimensions of the slab, and the rolling conditions of the large plate for each slab as explanatory variables into the prediction model 22.

[0043] Here, the "calculated slab dimensions" input as explanatory variables are calculated values ​​(provisional values) calculated from the set dimensions of the large plate. Specifically, a separately prepared yield master is referenced, and the yield determined based on the set dimensions of the large plate is used to provisionally determine the slab dimensions and slab weight to be input as explanatory variables. In this way, by using the provisionally determined calculated slab dimensions as explanatory variables, it is possible to predict the rolling excess length for the provisionally determined slab dimensions for a newly organized large plate. Then, by using the predicted rolling excess length to fine-tune the slab dimensions (weight), it is possible to calculate the final required slab dimensions (weight).

[0044] (Decision step) In the determination step, the determination unit 33 determines the weight (dimension) of the slab required for a fabricated product newly formed by combining one or more sheet material products (for example, thick steel plate products). In the determination step, first, the determination unit 33 calculates the difference between the rolling elongation predicted by the prediction unit 32 and the preset target elongation. Here, in the determination step, it is determined whether the rolling elongation predicted by the prediction unit 32 is within the range of the preset threshold value. When it is determined that the rolling elongation is within the range of the threshold value, the difference from the target elongation is calculated using the rolling elongation predicted by the prediction unit 32 as it is. On the other hand, when it is determined that the rolling elongation is outside the range of the threshold value, among the numerical values within the range of the threshold value, the numerical value closest to the rolling elongation predicted by the prediction unit 32 is regarded as the "rolling elongation", and the difference from the target elongation is calculated using this numerical value.

[0045] Specifically, for example, when the threshold value is ±100, when the rolling elongation is 30, the numerical value 30 is used as it is, and when the rolling elongation is 200, the numerical value 100 is used as the rolling elongation. Then, in the determination step, based on the obtained difference and the dimension (fabricated length) of the large plate, as shown in the following formulas (1) to (3), the yield correction allowance N' for correcting the yield is calculated using the preset gain (coefficient).

[0046]

Number

[0047] Note that the above formula (1) is a formula for calculating the yield correction allowance N' when the rolling elongation x is smaller than the lower limit a0 of the threshold value (x < a0), and (a0 - 3σ0 - p0) is the difference between the rolling elongation and the target elongation. Also, the above formula (2) is a formula for calculating the yield correction allowance N' when the rolling elongation x is greater than or equal to the lower limit a0 of the threshold value and less than or equal to the upper limit b0 of the threshold value (a0 ≤ x ≤ b0), and (x - 3σ0 - p0) is the difference between the rolling elongation and the target elongation. Also, the above formula (3) is a formula for calculating the yield correction allowance N' when the rolling elongation x is greater than the upper limit b0 of the threshold value (b0 < x), and (b0 - 3σ0 - p0) is the difference between the rolling elongation and the target elongation.

[0048] The final yield for the dimensions of the large plate is determined by adding the calculated yield correction amount N' to the yield determined by referring to the yield master, and the required slab weight is determined based on the determined yield. The slab dimensions are determined by back-calculating from the determined slab weight.

[0049] In the determination step, after determining the required slab weight (dimension), it is determined whether the determined slab weight (dimension) satisfies the equipment constraints. If it is determined that the slab weight (dimension) does not satisfy the equipment constraints, the slab weight (dimension) that satisfies the equipment constraints is recalculated.

[0050] <Other embodiments> In the above embodiment, the rolling excess length is predicted using the prediction model 22, the final yield is determined from the predicted rolling excess length, and the required slab weight (dimensions) is determined based on the determined yield. However, a configuration may also be adopted in which the prediction model 22 is made to learn the slab weight and slab yield as objective variables, and the prediction model 22 is used to directly predict the slab weight and slab yield.

[0051] <Manufacturing method for plate products> When predicting the rolling excess length of a slab and manufacturing a plate product, first, one or more plate products are assembled by hand or machine, and then calculations for the rolling material are performed. At this time, the rolling excess length is predicted using the prediction model 22, and the weight (dimensions) of the rolling material, i.e., the slab, is determined based on the predicted rolling excess length. After that, the slab is rolled according to the rolling conditions input as explanatory variables, and plate products are obtained from the large plate after rolling.

[0052] According to the rolling material determination device, rolling material determination method, and plate product manufacturing method of the above-described embodiments, the rolling excess length, the required slab weight, and the slab yield can be predicted with high accuracy by using the prediction model 22 to predict the rolling excess length, the required slab weight, and the slab yield. Therefore, for example, by determining the weight (dimensions) of the rolling material (slab) by adjusting the yield so that the predicted rolling excess length becomes the target excess length, in the manufacture of plate products at a steelworks or the like, if the predicted rolling excess length is longer than the target excess length, it is possible to achieve an improvement in yield. Furthermore, if the predicted rolling excess length is shorter than the target excess length, it is possible to prevent the plate product from being rejected due to a lack of longitudinal allowance.

[0053] The rolling material determination device, rolling material determination method, and plate product manufacturing method according to the present invention have been specifically described above using the mode and examples for carrying out the invention, but the gist of the present invention is not limited to these descriptions and must be broadly interpreted based on the claims. Furthermore, it goes without saying that various changes and modifications based on these descriptions are also included in the gist of the present invention. [Explanation of symbols]

[0054] 1. Information processing equipment 10 Input section 20 Memory section 21 Operation DB 22 Predictive Models 30 Arithmetic section 31 Learning Department 32 Prediction Department 33 Decision Section 40 Output section

Claims

1. A rolling material determination device for predicting rolling excess length when a knitted product is obtained by combining one or more plate material products from one large plate obtained by rolling a slab, a prediction unit that predicts the rolling excess length by inputting the set dimensions of the large plate, the calculated dimensions of the slab calculated from the set dimensions of the large plate, and the rolling conditions of the large plate into a prediction model that has been trained using the set dimensions of the large plate, the dimensions of the slab, and past actual values ​​of the rolling conditions of the large plate as explanatory variables, and the past actual values ​​of the rolling excess length corresponding to the set dimensions of the large plate as objective variables; Rolling material determination device.

2. The rolling mill further includes a determination unit that determines a required weight of the slab based on a difference between the rolling excess length predicted by the prediction unit and a preset target excess length.

2. The rolling material determination device according to claim 1.

3. the determination unit calculates a yield correction amount for correcting the yield of the slab based on the difference between the rolling excess length and the target excess length and the set dimensions of the large plate, and determines the required weight of the slab based on the calculated yield correction amount.

3. The rolling material determination device according to claim 2.

4. The determination unit determining whether the rolling excess length predicted by the prediction unit is within a range of a preset threshold value; When the rolling excess length is within the range of the threshold value, the rolling excess length predicted by the prediction unit is used as it is to calculate a difference from the target excess length, When the rolling excess length is outside the range of the threshold value, a numerical value that is closest to the rolling excess length predicted by the prediction unit among numerical values ​​within the range of the threshold value is regarded as the rolling excess length, and a difference from the target excess length is calculated using the threshold value.

4. The rolling material determination device according to claim 2 or 3.

5. After determining the required weight of the slab, the determination unit determines whether the determined weight of the slab satisfies the equipment constraints, and if the equipment constraints are not satisfied, recalculates the weight of the slab that satisfies the equipment constraints.

4. The rolling material determination device according to claim 2 or 3.

6. A rolling material determination device for predicting the weight of a slab or the yield of the slab, which is required when combining one or more plate products to obtain a combined product from a single large plate obtained by rolling a slab, a prediction unit that predicts the weight or yield of the slab by inputting the set dimensions of the large plate, the calculated dimensions of the slab calculated from the set dimensions of the large plate, and the rolling conditions of the large plate into a prediction model that has been trained using the set dimensions of the large plate, the dimensions of the slab, and past actual values ​​of the rolling conditions of the large plate as explanatory variables, and the corresponding past actual values ​​of the weight or yield of the slab as objective variables; Rolling material determination device.

7. A rolling material determination method for predicting rolling excess length when a knitted product is obtained by combining one or more plate material products from one large plate obtained by rolling a slab, comprising: a prediction step of predicting the rolling excess length by inputting the set dimensions of the large plate, the calculated dimensions of the slab calculated from the set dimensions of the large plate, and the rolling conditions of the large plate into a prediction model that has been trained using the set dimensions of the large plate, the dimensions of the slab, and past actual values ​​of the rolling conditions of the large plate as explanatory variables, and the past actual values ​​of the rolling excess length corresponding thereto as objective variables; Method for determining rolling material.

8. A rolling excess length is predicted using the rolling material determination method according to claim 7, a slab whose weight is determined based on the predicted rolling excess length is rolled to produce a large plate, and one or more plate products are extracted from the large plate. Manufacturing method for sheet metal products.

Images