Steel temperature prediction device, cooling control device, method, and program

Abstract

To predict a temperature of a steel material in consideration of transformation heating with high accuracy.SOLUTION: In a hot rolling facility including a rolling machine 2, a ROT cooling device 3 and a winding device 4, a cooling actual data extraction part 102 extracts cooling actual data of a steel plate of the same steel kind as an object steel plate 1 from database 107, on the basis of steel kind information on the object steel plate 1. A TTT curve generation part 103 generates a TTT curve for an object steel plate 1 using the cooling actual data extracted by the cooling actual data extraction part 102, and a temperature drop amount prediction model of reflecting a prediction result of a transformation heating amount based on a TTT curve and predicting a winding temperature. The TTT curve generation part 103 newly generates a curve indicating transformation start constituting the TTT curve, and a curve indicating transformation end, and each of the curves give a plurality of nodes on a two-dimensional plane of a temperature and time, and the nodes are interpolated using a curve represented by a polynomial and generated.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present invention relates to a temperature prediction device, a cooling control device, a method, and a program for steel materials.

Background Art

[0002] In hot rolling facilities, the steel plate after finish rolling is cooled by a cooling device and wound into a coil by a winding device. In order to maintain high productivity and obtain good product quality, it is important to control the coiling temperature, which is the temperature of the steel plate before the winding device. For example, when medium carbon steel and high carbon steel are targeted, if the cooling in the cooling device becomes excessive and the coiling temperature becomes low, a hard layer is generated and cracks are likely to occur. Therefore, the quality and yield decrease. On the other hand, if the cooling in the cooling device is insufficient and the coiling temperature becomes high, the phase transformation (transformation) from austenite to ferrite proceeds in the state of the coil after winding, and winding looseness may occur, resulting in coil collapse (deformation into an ellipse and the need for rewinding). Therefore, a correction process such as rewinding is required, and productivity decreases. In order to prevent such a decrease in quality, yield, and productivity, it is necessary to preset the target temperature of the coiling temperature and operate the cooling device to control the coiling temperature to the target temperature.

[0003] However, when the steel plate is cooled to progress the transformation, transformation heat generation occurs. Particularly in medium carbon steel and high carbon steel with a high carbon content, the amount of transformation heat generation is large, and in some cases, a heat amount sufficient to raise the temperature of the steel plate by 100°C or more may be generated from the start to the end of the transformation. Since such transformation heat generation occurs, it is difficult to accurately predict the coiling temperature, and the accuracy of the control to make the coiling temperature the target temperature may deteriorate.

[0004] Patent Document 1 discloses a method for manufacturing a steel material manufactured through a process of cooling a steel plate processed by a finish rolling mill with a cooling device, calculating the amount of transformation heat generation of the steel plate using a TTT curve (also called a constant temperature transformation curve or an isothermal transformation curve), and predicting the coiling temperature of the steel plate in consideration of the amount of transformation heat generation.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Patent Document 2

Patent Document 3

Summary of the Invention

Problems to be Solved by the Invention

[0006] However, when the composition of alloying elements such as carbon in steel is different, differences also occur in the transformation behavior of steel. Therefore, even if using an existing TTT curve for a specific alloying element composition, it is impossible to accurately predict the transformation heat generation amount. Thus, when manufacturing multiple steel grades, a large number of different TTT curves for each steel grade must be newly prepared, which consumes time and cost for that purpose. Therefore, in Patent Document 1, for the sake of time saving and cost reduction, an existing TTT curve diagram is deformed according to the composition of the steel plate cooled by a cooling device (specifically, the existing TTT curve diagram is simply translated parallel in the temperature axis direction and the time axis direction on the temperature - time plane) and used. Therefore, if it is a case targeting a steel grade with a small difference in alloying element composition compared to the steel grade of the existing TTT curve diagram, it is possible to improve the prediction accuracy. However, when targeting a steel grade with a large difference in alloying element composition compared to the steel grade of the existing TTT curve diagram, the effect of improving the prediction accuracy becomes small. Also, generally, a TTT curve is generated using the measurement results of the transformation behavior when a test material is heated and cooled by a test device. However, it is difficult for a test device to completely reproduce the transformation heat generation phenomenon in an actual machine, and there may be cases where the coiling temperature cannot be predicted with high accuracy.

[0007] The present invention has been made in view of the above points, and an object thereof is to enable highly accurate prediction of the temperature of a steel material, regardless of the steel type, taking into account the transformation heat generation, without preparing in advance a large number of TTT curves different for each steel type using a test apparatus or the like.

Means for Solving the Problems

[0008] The apparatus for predicting the temperature of a steel material of the present invention is an apparatus for predicting the temperature of a steel material to be cooled, and includes cooling performance data for each steel type and including at least a term representing the temperature drop of the steel material due to cooling and a term representing the temperature change of the steel material due to transformation heat generation, TTT curve generation means for generating a TTT curve for a target steel material, using a prediction model that predicts the temperature of the steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve, and temperature prediction means for predicting the temperature of the target steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve generated by the TTT curve generation means in the prediction model , the TTT curve generation means generates the TTT curve for the target steel material by interpolating between a plurality of nodes represented by time and temperature using a curve represented by a polynomial, and based on an evaluation function including the difference between the actual temperature value included in the cooling performance data for a steel material of the same or similar steel type as the target steel material and the temperature predicted by the prediction model using the cooling performance data, optimize the time and temperature of the plurality of nodes characterized in that. The cooling control apparatus of the present invention is a cooling control apparatus for controlling the temperature of a steel material cooled by a cooling apparatus, and includes control means for operating the cooling apparatus so that the temperature of the target steel material predicted by the apparatus for predicting the temperature of a steel material of the present invention becomes a preset target temperature. The method for predicting the temperature of a steel material of the present invention is a method for predicting the temperature of a steel material to be cooled, and includes cooling performance data for each steel type and including at least a term representing the temperature drop of the steel material due to cooling and a term representing the temperature change of the steel material due to transformation heat generation, a TTT curve generation step of generating a TTT curve for a target steel material, using a prediction model that predicts the temperature of the steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve, and a temperature prediction step of predicting the temperature of the target steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve generated in the TTT curve generation step in the prediction model , and in the TTT curve generation step, the TTT curve for the target steel material is generated by interpolating between a plurality of nodes represented by time and temperature using a curve represented by a polynomial, and based on an evaluation function including the difference between the actual temperature value included in the cooling performance data for a steel material of the same or similar steel type as the target steel material and the temperature predicted by the prediction model using the cooling performance data, optimize the time and temperature of the plurality of nodes characterized in that. The cooling control method of the present invention is a cooling control method for controlling the temperature of a steel material cooled by a cooling apparatus, and includes cooling performance data for each steel type and including at least a term representing the temperature drop of the steel material due to cooling and a term representing the temperature change of the steel material due to transformation heat generation, A TTT curve generation step of generating a TTT curve for a target steel material using a prediction model that predicts the temperature of the steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve; a temperature prediction step of predicting the temperature of the target steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve generated in the TTT curve generation step in the prediction model; and a step of operating the cooling device so that the temperature of the target steel material predicted in the temperature prediction step becomes a preset target temperature. , and in the TTT curve generation step, the TTT curve for the target steel material is generated by interpolating between a plurality of nodes represented by time and temperature using a curve represented by a polynomial, and based on an evaluation function including the difference between the actual temperature value included in the cooling performance data for a steel material of the same or similar steel type as the target steel material and the temperature predicted by the prediction model using the cooling performance data, optimize the time and temperature of the plurality of nodes It is characterized by this. The program of the present invention is a program for predicting the temperature of a steel material to be cooled, and uses cooling performance data for each steel type and including at least a term representing the temperature drop of the steel material due to cooling and a term representing the temperature change of the steel material due to transformation heat generation, a TTT curve generation means for generating a TTT curve for a target steel material using a prediction model that predicts the temperature of the steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve, and a temperature prediction means for predicting the temperature of the target steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve generated by the TTT curve generation means in the prediction model, causing a computer to function , the TTT curve generation means generates the TTT curve for the target steel material by interpolating between a plurality of nodes represented by time and temperature using a curve represented by a polynomial, and based on an evaluation function including the difference between the actual temperature value included in the cooling performance data for a steel material of the same or similar steel type as the target steel material and the temperature predicted by the prediction model using the cooling performance data, optimizes the time and temperature of the plurality of nodes, characterized in that . The program of the present invention is a program for controlling the temperature of a steel material to be cooled by a cooling device, and uses cooling performance data for each steel type and including at least a term representing the temperature drop of the steel material due to cooling and a term representing the temperature change of the steel material due to transformation heat generation, a TTT curve generation means for generating a TTT curve for a target steel material using a prediction model that predicts the temperature of the steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve, a temperature prediction means for predicting the temperature of the target steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve generated by the TTT curve generation means in the prediction model, and a control means for operating the cooling device so that the temperature of the target steel material predicted by the temperature prediction means becomes a preset target temperature, causing a computer to function , the TTT curve generation means generates the TTT curve for the target steel material by interpolating between a plurality of nodes represented by time and temperature using a curve represented by a polynomial, and based on an evaluation function including the difference between the actual temperature value included in the cooling performance data regarding a steel material of the same or similar steel type as the target steel material and the temperature predicted by the prediction model using the cooling performance data, optimizes the time and temperature of the plurality of nodes. .

Effect of the Invention

[0009] According to the present invention, since the TTT curve for the target steel material optimized for each steel type is generated using the cooling performance data for each steel type, it is possible to accurately predict the temperature of the steel material considering the transformation heat generation with high accuracy regardless of the steel type without preparing a large number of different TTT curves in advance for each steel type using a test device or the like.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Embodiments for Carrying Out the Invention

[0011] Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. First, with reference to FIG. 1, the schematic configuration of the hot rolling facility will be described. FIG. 1 is a diagram showing the schematic configuration of the hot rolling facility. The hot rolling facility includes a rolling mill 2, a ROT (Run out table) cooling device 3, and a coiling device 4. The rolling mill 2 finishes rolling the steel plate 1. The steel plate 1 after finish rolling by the rolling mill 2 is conveyed on a conveying table and passes through the ROT cooling device 3. The ROT cooling device 3 cools the steel plate 1 by spraying cooling water onto the upper surface side and the lower surface side of the steel plate 1 from a plurality of cooling headers (not shown) having a nozzle group. The coiling device 4 winds up the steel plate 1 cooled by the ROT cooling device 3 into a coil shape.

[0012] On the outlet side of the rolling mill 2 and on the inlet side of the ROT cooling device 3, a thickness gauge 5 for measuring the thickness of the steel sheet 1 and a thermometer 6 for measuring the temperature of the steel sheet 1 (referred to as the finishing outlet temperature) are installed. Also, on the outlet side of the ROT cooling device 3 and on the inlet side of the coiling device 4, a thermometer 7 for measuring the coiling temperature, which is the temperature of the steel sheet 1 before the coiling device 4, is installed. Further, a conveying speed meter 8 for measuring the conveying speed of the steel sheet 1 is installed.

[0013] In such a hot rolling facility, the cooling control device 100 uses a prediction model to predict the coiling temperature, which is the temperature of the target steel sheet 1 after cooling by the ROT cooling device 3, and operates the ROT cooling device 3 for control to make the predicted coiling temperature the target temperature.

[0014] Hereinafter, the cooling control device 100 will be described in detail. Before that, the prediction model for predicting the coiling temperature will be described. As a prediction model for predicting the coiling temperature, a temperature drop amount prediction model for predicting the temperature drop amount of the steel sheet is used. Equations (1) to (3) show the temperature drop amount prediction model for calculating the temperature drop amount of the steel sheet from the outlet side of the rolling mill 2 to before the coiling device 4. The section from the thermometer 6 to the thermometer 7 is divided into a plurality of zones, and the heat transfer coefficients of water cooling, radiation, and convection are given to each zone based on the operating state of the ROT cooling device 3 to calculate the temperature drop amount Δθ(i) of the steel sheet. In Equation (1), the first term on the right side represents the temperature change of the steel sheet due to water cooling, and the second term on the right side represents the temperature change of the steel sheet due to heat radiation, the temperature change of the steel sheet due to contact with air, and the temperature change of the steel sheet due to transformation heat generation. By subtracting the temperature drop amount Δθ(i) of the steel sheet in each zone from the finishing outlet temperature measured by the thermometer 6, the temperature of the steel sheet in each zone can be calculated. i: Subscript indicating the target zone Δθ(i): Temperature drop amount of the steel sheet (°C) θ(i): Temperature of the steel sheet at the inlet side of the zone (referred to as the zone inlet temperature) (°C) θ W : Water temperature of the cooling water (°C) θ A : Air temperature (°C) Δt(i): Zone passing time (hr) α U (i), α L (i): Upper and lower heat transfer rates (kcal / m 2 / hr / °C) C P (i): Zone specific heat (kcal / kg / °C) ρ: Density of the steel sheet (kg / m 3 ) h: Thickness of the steel sheet (m) ε: Emissivity (-) σ: Stefan-Boltzmann constant θ K : 273.15 °C α A : Convective heat transfer rate (kcal / m 2 / hr / °C) q H (i): Zone transformation heat generation amount (kcal / m 3 / hr) W(i): Zone water density (m 3 / m 2 / min) a 1-4 : Constant

[0015]

Number

[0016] As shown in Equations (1) to (3), the temperature drop prediction model includes the zone transformation heat generation amount q H (i). Therefore, the coiling temperature is predicted by reflecting the prediction results of the transformation heat generation amount based on the TTT curve in the temperature drop prediction model of Equations (1) to (3). Figure 2 is a diagram showing an example of a TTT curve. As shown in Figure 2, the TTT curve shows the change in the structure of steel due to temperature and time by taking temperature on the vertical axis and time on the horizontal axis. The TTT curve is composed of a curve indicating the start of transformation (a curve showing the relationship between the time and temperature at which isothermal transformation starts) and a curve indicating the end of transformation (a curve showing the relationship between the time and temperature at which isothermal transformation ends), and represents the transformation start time and transformation end time when the steel is held at a certain temperature. At the temperature θ(i) on the inlet side of the i zone, the transformation starts at time t s (i) and ends at time tf (i) indicates the completion of transformation. The transformation start time t derived from the TTT curve in this way s (i) and the transformation end time t f (i) are used to calculate the zone transformation heat generation amount q H (i) of the i zone according to formulas (4) to (10). i: Subscript indicating the target zone θ(i): Temperature of the steel plate on the inlet side of the zone (°C) t s (i): Transformation start time (s) at θ(i) (derived from the TTT curve) t f (i): Transformation end time (s) at θ(i) (derived from the TTT curve) ξ S : Transformation rate at the start of transformation (=0.01) ξ F : Transformation rate at the end of transformation (=0.99) ξ(i), ξ(i + 1): Transformation rates on the inlet and outlet sides of the zone t(i): Elapsed time (s) of the transformation rate equation required to match ξ(i) Δt(i): Zone passing time (s) q H (i): Zone transformation heat generation amount (kcal / m 3 / hr) α H : Latent heat of transformation (kcal / kg)

[0017]

Equation

[0018] As described above, the transformation start time t s (i) and the transformation end time t f (i) at the inlet temperature θ(i) of each zone are derived from the TTT curve, and the zone transformation heat generation amount q H (i) can be calculated. Then, if the calculation of the temperature drop amount Δθ(i) of the steel plate for each zone and the zone transformation heat generation amount q H (i) is advanced for each zone, it becomes possible to predict the coiling temperature considering the transformation heat generation.

[0019] Next, with reference to FIG. 3, the cooling control device 100 will be described in detail. FIG. 3 is a diagram showing the functional configuration of the cooling control device 100. The cooling control device 100 includes an input unit 101, a cooling performance data extraction unit 102, a TTT curve generation unit 103, a storage unit 104, a temperature prediction unit 105, and a control unit 106. The cooling control device 100 is also connected to a database 107 that stores cooling performance data for each steel type. The cooling control device 100 configured in this way can be realized by a computer device including, for example, a CPU, a ROM, a RAM, etc. The CPU reads a program stored in the ROM, for example, and by executing this program, the functions of each of the units 101 to 106 are realized. In this embodiment, it is assumed that the process computer that controls the production process in the steelworks functions as the cooling control device 100.

[0020] The input unit 101 inputs steel type information and the target temperature of the coiling temperature for the target steel plate 1 that is rolled, cooled, and coiled in the hot rolling facility. The input unit 101 also inputs the operating state of the ROT cooling device 3 (including the water volume of the cooling headers in each zone), the plate thickness of the target steel plate 1 measured by the plate thickness gauge 5, the finish exit side temperature of the target steel plate 1 measured by the thermometer 6, the coiling temperature of the target steel plate 1 measured by the thermometer 7, the conveyance speed measured by the conveyance speed meter 8, the water temperature of the cooling water, and the air temperature. The input unit 101 also inputs information such as various constants necessary for formulas (1) to (10), etc. However, for fixed values, they may be held within the cooling control device 100. The input unit 101 may input this information, for example, directly from an external device or via a network, or by manual input by the user, and the input method is not limited. Note that if the fluctuation range of the water temperature and air temperature of the cooling water is limited, they may be treated as fixed values.

[0021] The cooling performance data extraction unit 102 extracts cooling performance data regarding steel plates of the same or similar steel grades as the target steel plate 1 from the database 107 based on the steel grade information of the target steel plate 1 input by the input unit 101. Here, similarity means allowing a variation of a predetermined percentage (varying depending on the steel grade) in the added element components. The cooling performance data for each steel grade stored in the database 107 includes measured values (actual values) of the finish side temperature, measured values (actual values) of the coiling temperature, actual values of the passing speed in each zone, actual values of the water volume of the cooling headers in each zone, and actual values of the water temperature and air temperature of the cooling water, sampled in the longitudinal direction of the steel plates manufactured in the past. Note that if the variation range of the water temperature and air temperature of the cooling water is limited, it may be treated as a fixed value.

[0022] The TTT curve generation unit 103 generates a TTT curve for the target steel plate 1 using the cooling performance data extracted by the cooling performance data extraction unit 102 and the temperature drop prediction models of formulas (1) to (3). The TTT curve generation unit 103 newly generates two curves (a curve indicating the start of transformation and a curve indicating the end of transformation) that constitute the TTT curve as the TTT curve for the target steel plate 1. Each curve is given a plurality of nodes on the two-dimensional plane of temperature and time, and is generated by interpolating using a curve represented by a polynomial between the nodes. At that time, based on an evaluation function including the difference between the measured value of the coiling temperature included in the cooling performance data regarding the steel plates of the same or similar steel grades as the target steel plate 1 extracted by the cooling performance data extraction unit 102 and the predicted value of the coiling temperature predicted by the temperature drop prediction models of formulas (1) to (3) using the cooling performance data, the optimization of the TTT curve is performed.

[0023] Hereinafter, with reference to FIGS. 4 and 5, a method for generating a TTT curve will be described in detail. FIGS. 4 and 5 are diagrams for explaining the generation of the TTT curve. As shown in FIG. 4, a plurality of nodes F1 to F6 of a curve t = f(θ) indicating the start of transformation and a plurality of nodes G1 to G6 of a curve t = g(θ) indicating the end of transformation are given on the two-dimensional plane of temperature and time, and the optimization of time and temperature is performed. In the example of FIG. 4, an example of setting six nodes F1 to F6 and G1 to G6 for each curve is shown, but the number of nodes may be other than six.

[0024] Nodes F1(θ1, t1), ···, F6(θ6, t6), G1(θ7, t7), ···, G6(θ 12 , t 12 ) need to define the temperature θ and time t. When setting 6 nodes for each curve, a total of 24 parameters are required. Since it is desirable to have fewer parameters for optimization, the temperature θ1 and time t1 of node F1, which is the end point on the lowest temperature side of the curve t = f(θ), and the temperature θ7 and time t7 of node G1, which is the end point on the lowest temperature side of the curve t = g(θ), are treated as fixed values. Also, the time t6 of node F6, which is the end point on the later time side of the curve t = f(θ), and the time t 12 of node G6, which is the end point on the later time side of the curve t = g(θ), are treated as fixed values. As a result, the number of parameters to be optimized can be reduced to 18, and the load of the optimization calculation can be reduced. For example, when medium carbon steel or high carbon steel is used as the target steel plate 1, it is almost impossible to cool it to 500 °C or lower with the ROT cooling device 3. Therefore, if nodes F1 and G1 are nodes with a temperature of 500 °C or lower, even if the respective temperatures θ1, θ7 and times t1, t7 are fixed, there will be no practical inconvenience in predicting the transformation heat generation amount.

[0025] If the heat transfer coefficients of water cooling, radiation, and convection of the temperature drop prediction model of equations (1) to (3) are appropriately given, the coiling temperature can be treated as a function of the coordinates of the TTT curve. Therefore, as shown in equation (11), the measured value T of the coiling temperature included in the cooling performance data of steel plates of the same steel type as the target steel plate 1 act and the predicted value T of the coiling temperature predicted by the temperature drop prediction model of equations (1) to (3) using the cooling performance data cal are used to optimize the TTT curve so that the evaluation function including the difference between them is minimized. Specifically, a temporary TTT curve is generated by spline interpolation between the coordinates (time and temperature) of a plurality of candidate nodes F1 to F6, G1 to G6, and the predicted value T of the coiling temperature is obtained using the temporary TTT curve and equations (1) to (10) calThe process of calculating and evaluating using the evaluation function of Equation (11) is repeated while changing (moving) the coordinates of nodes F1 to F6 and G1 to G6 until the evaluation function of Equation (11) becomes the minimum. Then, the TTT curve when the evaluation function of Equation (11) is the minimum is set as the optimal TTT curve for the target steel plate 1. By such an optimization calculation, it becomes possible to generate an optimal TTT curve that conforms to the actual transformation heat generation phenomenon. As described above, the temperature θ1 and time t1 of node F1, the temperature θ7 and time t7 of node G1, the time t6 of node F6, and the time t of node G6 12 are set as constants (fixed values).

[0026] [Number]

[0027] As described above, the TTT curve for the target steel plate 1 is generated. For the spline interpolation between multiple nodes, a cubic spline curve is used. This is an interpolation method in which the curve between nodes is represented by a cubic polynomial so as not to be discontinuous at nodes other than the endpoints. By interpolating using a curve represented by a polynomial between nodes in this way, a TTT curve represented by a smooth curve can be generated, and it becomes possible to reproduce it as being close to the actual transformation heat generation phenomenon. However, interpolation may be performed using other curves or straight lines instead of the cubic spline curve.

[0028] Note that the movement amount of the node coordinates and the calculated value T of the coiling temperature calSince the relationship with it has a strong non-linearity, as an optimization method, PSO (Particle Swarm Optimization), which is highly evaluated as an approximate solution method for non-linear optimization problems, was executed. In PSO, it is necessary to set the initial particle generation range for each coordinate. As shown in FIG. 5, the initial particle generation ranges for the coordinates of each node F2 to F6 and G2 to G6 were set as the ranges surrounded by squares, respectively (in FIG. 5, nodes F6 and G6 are not shown because they are outside the display range). This initial particle generation range may be set, for example, with reference to known TTT curves described in Patent Document 1 or the like. Note that the particle coordinates after optimization are often set outside the initial particle range, and it is not necessary to strictly set the initial particle range.

[0029] Returning to the explanation in FIG. 3, the storage unit 104 stores and saves the TTT curve for the target steel plate 1 generated by the TTT curve generation unit 103.

[0030] By using the TTT curve for the target steel plate 1 stored in the storage unit 104, the transformation start time t s (i) and the transformation end time t f (i) are derived, and the zone transformation heat generation amount q H (i), which is the prediction result of the transformation heat generation amount, is calculated according to formulas (4) to (10). Then, the temperature prediction unit 105 reflects the calculated zone transformation heat generation amount q H (i) in the temperature drop prediction model of formulas (1) to (3) to predict the coiling temperature of the target steel plate 1.

[0031] The control unit 106 operates the ROT cooling device 3 so that the coiling temperature of the target steel plate 1 predicted by the temperature prediction unit 105 becomes the target temperature. Specifically, the control unit 106 sets an operation amount for adjusting the water volume of the cooling header of the ROT cooling device 3 so that the coiling temperature of the target steel plate 1 predicted by the temperature prediction unit 105 approaches the target temperature.

[0032] Next, with reference to FIG. 6, the processes executed by the cooling control device 100 will be described. FIG. 6 is a flowchart showing the processes executed by the cooling control device 100. The flowchart shown in FIG. 6 starts before the target steel plate 1 enters the ROT cooling device 3. In step S1, the input unit 101 inputs steel type information and the target temperature of the coiling temperature for the target steel plate 1.

[0033] In step S2, based on the steel type information of the target steel plate 1 input in step S1, the cooling performance data extraction unit 102 extracts cooling performance data regarding steel plates of the same or similar steel types as the target steel plate 1 from the database 107.

[0034] In step S3, the TTT curve generation unit 103 optimizes the TTT curve so that the evaluation function (Equation (11)) including the difference between the measured coiling temperature T act in the cooling performance data of steel plates of the same or similar steel types as the target steel plate 1 extracted in step S2 and the coiling temperature T cal predicted by the temperature drop prediction models of Equations (1) to (3) using the cooling performance data is minimized, and generates a TTT curve for the target steel plate 1. In step S4, the TTT curve generation unit 103 stores and saves the TTT curve for the target steel plate 1 generated in step S3 in the storage unit 104.

[0035] When the temperature prediction target part of the target steel plate 1 reaches the position of the thermometer 6, the processes after step S5 are started. In step S5, the input unit 101 inputs the operating state of the ROT cooling device 3 (including the water volume of the cooling headers in each zone), the plate thickness of the target steel plate 1 measured by the plate thickness gauge 5, the finish side temperature of the target steel plate 1 measured by the thermometer 6, the conveyance speed measured by the conveyance speed meter 8, the water temperature of the cooling water, and the air temperature. Among the information described here, those that can be input at the stage of step S1 may be input in step S1.

[0036] In step S6, the temperature prediction unit 105 uses the TTT curve for the target steel plate 1 stored in the storage unit 104 to obtain the transformation start time t s (i) and the transformation end time t f (i), and according to formulas (4) to (10), calculates the zone transformation heat generation amount q H (i) which is the prediction result of the transformation heat generation amount. Then, the temperature prediction unit 105 reflects the calculated zone transformation heat generation amount q H (i) in the temperature drop prediction model of formulas (1) to (3) to predict the coiling temperature of the target steel plate 1.

[0037] In step S7, the control unit 106 operates the ROT cooling device 3 so that the coiling temperature of the target steel plate 1 predicted in step S6 becomes the target temperature. Specifically, the control unit 106 sets an operation amount for adjusting the water volume of the cooling header of the ROT cooling device 3 so as to bring the coiling temperature of the target steel plate 1 predicted in step S7 closer to the target temperature.

[0038] In step S8, the control unit 106 determines whether the end condition has been reached. In this embodiment, the end condition is that the temperature prediction target part of the target steel plate 1 reaches the position of the thermometer 7. If the end condition has not been reached, the process returns to step S5. In step S5, if there is changed information, it is input again. For example, since the operation state of the ROT cooling device 3 changes in the process of step S7, it is input again. The processes of steps S5 to S8 are repeated at a cycle of several hundred milliseconds. On the other hand, if the end condition has been reached, the process proceeds to step S9.

[0039] In step S9, the input unit 101 inputs the coiling temperature of the target steel plate 1 measured by the thermometer 7. In step S10, the input unit 101 stores the cooling performance data of the target steel plate 1 obtained in steps S1, S5, and S9 in the database 107. The cooling performance data of the target steel plate 1 includes the measured values (actual values) of the finish side temperature, the measured values (actual values) of the coiling temperature, the actual values of the passing speeds in each zone, the actual values of the water amounts of the cooling headers in each zone, and the actual values of the water temperature and air temperature of the cooling water, sampled in the longitudinal direction of the target steel plate 1.

[0040] Note that in this embodiment, as shown in the flowchart of FIG. 6, the series of processes from the generation of the TTT curve to the control of the ROT control device 3 are executed online, but the present invention is not limited to this. For example, the processes of steps S2 to S4 may be executed offline, and the TTT curve for the steel grade assumed as the target steel plate 1 may be generated in advance.

[0041] As described above, the TTT curve for the target steel plate 1 is generated using the cooling performance data for each steel grade and the prediction model that predicts the coiling temperature by reflecting the prediction results of the transformation heat generation amount based on the TTT curve. As a result, even when manufacturing multiple steel grades, a new TTT curve suitable for the steel grade can be generated, and the coiling temperature considering the transformation heat generation can be predicted with high accuracy regardless of the steel grade. As a result, high-precision control to set the coiling temperature to the target temperature can be realized, high productivity can be maintained, and good product quality can be obtained.

[0042] [Example] As an example, for structural carbon steel S70C and carbon tool steel SK85, the TTT curves were generated by the method described in the embodiment using the cooling performance data of each steel grade. Further, as a comparative example, for structural carbon steel S70C and carbon tool steel SK85, the TTT curves were deformed by the method described in Patent Document 1. In generating the TTT curves of the examples, as cooling performance data, data for 9 coils of S70C and 16 coils of SK85 were used. The cooling performance data includes the measured values of the finish side temperature, the measured values of the coiling temperature, the actual values of the passing speeds in each zone, the actual values of the water volume of the cooling headers in each zone, and the actual values of the water temperature and air temperature of the cooling water, sampled at 10 m intervals in the longitudinal direction of the steel plate. Note that since the first and last 50 m of the steel plate are unsteady parts, they are excluded from the cooling performance data.

[0043] Figure 7 is a diagram showing the TTT curves of the examples, where (a) shows the TTT curve of steel grade S70C and (b) shows the TTT curve of steel grade SK85. Figure 8 is a diagram showing the TTT curves of the comparative examples, where (a) shows the TTT curve of steel grade S70C and (b) shows the TTT curve of steel grade SK85. In the comparative example, an existing single TTT curve is deformed by sliding it on the two-dimensional plane of temperature and time. As shown in Figures 8(a) and (b), the shapes of the curves indicating the start of transformation and the curves indicating the end of transformation are the same between different steel grades, but their positions on the two-dimensional plane of temperature and time are different. In contrast, in the example, the TTT curves are newly generated using the cooling performance data of each steel grade, so that the shapes of the curves indicating the start of transformation and the curves indicating the end of transformation are suitable for the steel grade. Note that in the example, for the purpose of reducing the number of nodes to be optimized, the nodes below 500 °C are fixed and the curve shape below 500 °C is intentionally changed. Since it is unlikely to cool medium carbon steel and high carbon steel down to 500 °C or lower, there is no practical inconvenience in predicting the transformation heat generation amount.

[0044] Using the TTT curves obtained in this way, the prediction accuracy of the coiling temperature and the control accuracy of the coiling temperature were evaluated. The results are shown in Table 1. The root mean square error (RMSE) representing the prediction accuracy of the coiling temperature was evaluated. In this evaluation, a coil different from that used to generate the TTT curve was used, and 14 coils of S70C and 33 coils of SK85 were used in both the examples and the comparative examples. As a result, it was confirmed that in the examples, the RMSE was improved by about 5 to 10% compared with the comparative examples. Thus, in the examples, a new TTT curve suitable for the steel grade can be generated, and the coiling temperature considering the transformation heat generation can be predicted with high accuracy.

[0045] In addition, in order to evaluate the control accuracy of the coiling temperature, the hitting rate (coil longitudinal direction target ± 25 °C hitting rate) with respect to the target of the coiling temperature was evaluated. In this evaluation, a coil different from that used to generate the TTT curve was used. In the comparative examples, 28 coils of S70C and 42 coils of SK85 were used, and in the examples, 7 coils of S70C and 12 coils of SK85 were used. As a result, it was confirmed that in the examples, the hitting rate of the target ± 25 °C was about 2 to 4% compared with the comparative examples. Thus, in the examples, since the coiling temperature can be predicted with high accuracy, high-precision control of setting the coiling temperature to the target temperature can be realized.

[0046]

Table 1

[0047] The above-described embodiments merely show examples of the implementation of the present invention, and the technical scope of the present invention should not be construed limitedly by these. That is, the present invention can be implemented in various forms without departing from its technical idea or its main features. In the above-described embodiment, the cooling control device 100 functions as a steel material temperature prediction device to which the present invention is applied and a cooling control device to which the present invention is applied. However, for example, a steel material temperature prediction device to which the present invention is applied and a cooling control device to which the present invention is applied may be configured as separate devices. In addition, in this embodiment, an example of predicting the coiling temperature in hot rolling equipment has been described, but the present invention is not limited thereto. The present invention is applicable to predicting the temperature of a steel material to be cooled, and can be used, for example, to predict the temperature of a steel plate in thick plate cooling equipment or annealing equipment for heating and cooling steel materials. It may also be used to predict the temperature of the steel plate at the position of the in-line thermometer in hot rolling equipment. In this case, for example, the TTT curve may be optimized based on an evaluation function including the difference between the predicted value and the measured value at the position of the in-line thermometer and the difference between the predicted value and the measured value at the position of the coiling thermometer so as to improve the accuracy of both the in-line temperature and the coiling temperature. Note that the function of predicting the temperature of the steel material to which the present invention is applied and the function of cooling control can also be realized by supplying software (program) to a system or device via a network or various storage media and having a computer of the system or device read and execute the program. Further, the function of controlling the blast furnace process to which the present invention is applied may be realized by a PLC (Programmable Logic Controller) or may be realized by dedicated hardware such as an ASIC.

Explanation of reference numerals

[0048] 1: Steel plate, 2: Rolling mill, 3: ROT cooling device, 4: Coiling device, 5: Plate thickness gauge, 6, 7: Thermometers, 8: Conveying speed meter, 100: Cooling control device, 101: Input unit, 102: Cooling performance data extraction unit, 103: TTT curve generation unit, 104: Storage unit, 105: Temperature prediction unit, 106: Control unit, 107: Database

Claims

1. A steel material temperature prediction device for predicting the temperature of a steel material to be cooled, comprising at least cooling performance data by steel type, terms representing the temperature drop amount of the steel material due to cooling, and terms representing the temperature change of the steel material due to transformation heat generation, and using a prediction model for predicting the temperature of the steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve, a TTT curve generation means for generating a TTT curve for the target steel material; temperature prediction means for predicting the temperature of the target steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve generated by the TTT curve generation means in the prediction model; the TTT curve generation means generates the TTT curve for the target steel material by interpolating between a plurality of nodes represented by time and temperature using a curve represented by a polynomial; Based on an evaluation function including the difference between the actual temperature value included in the cooling performance data regarding a steel material of the same or similar steel type as the target steel material and the temperature predicted by the prediction model using the cooling performance data, the time and temperature of a plurality of the nodes are optimized. A steel material temperature prediction device characterized by this.

2. The steel material temperature prediction device according to claim 1, wherein the TTT curve generation means executes a particle swarm optimization method as an optimization method for optimizing the time and temperature of the nodes.

3. The steel material temperature prediction device according to claim 1 or 2, wherein the TTT curve generation means treats at least one of the time and temperature of a part of the plurality of nodes as a fixed value.

4. A cooling control device for controlling the temperature of a steel material cooled by a cooling device, characterized by comprising control means for operating the cooling device so that the temperature of the target steel material predicted by the steel material temperature prediction device according to any one of claims 1 to 3 becomes a preset target temperature.

5. A steel material temperature prediction method for predicting the temperature of a steel material to be cooled, comprising at least cooling performance data by steel type, terms representing the temperature drop amount of the steel material due to cooling, and terms representing the temperature change of the steel material due to transformation heat generation, and using a prediction model for predicting the temperature of the steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve, a TTT curve generation step for generating a TTT curve for the target steel material; A temperature prediction step of predicting the temperature of the target steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve generated in the TTT curve generation step into the prediction model, In the TTT curve generation step, the TTT curve for the target steel material is generated by interpolating between a plurality of nodes represented by time and temperature using a curve represented by a polynomial, Based on an evaluation function including the difference between the actual temperature value included in the cooling performance data regarding a steel material of the same or similar steel type as the target steel material and the temperature predicted by the prediction model using the cooling performance data, the time and temperature of a plurality of the nodes are optimized. A method for predicting the temperature of a steel material, characterized by this.

6. A cooling control method for controlling the temperature of a steel material cooled by a cooling device, Including at least steel type-specific cooling performance data, a term representing the temperature drop amount of the steel material due to cooling, and a term representing the temperature change of the steel material due to transformation heat generation, and using a prediction model that predicts the temperature of the steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve, a TTT curve generation step of generating a TTT curve for the target steel material, A temperature prediction step of predicting the temperature of the target steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve generated in the TTT curve generation step into the prediction model, Having a step of operating the cooling device so that the temperature of the target steel material predicted in the temperature prediction step becomes a preset target temperature, In the TTT curve generation step, the TTT curve for the target steel material is generated by interpolating between a plurality of nodes represented by time and temperature using a curve represented by a polynomial, Based on an evaluation function including the difference between the actual temperature value included in the cooling performance data regarding a steel material of the same or similar steel type as the target steel material and the temperature predicted by the prediction model using the cooling performance data, the time and temperature of a plurality of the nodes are optimized. A cooling control method for a steel material, characterized by this.

7. A program for predicting the temperature of a steel material to be cooled, Including at least steel type-specific cooling performance data, a term representing the temperature drop amount of the steel material due to cooling, and a term representing the temperature change of the steel material due to transformation heat generation, and using a prediction model that predicts the temperature of the steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve, a TTT curve generation means for generating a TTT curve for the target steel material, Causing a computer to function as temperature prediction means for predicting the temperature of the target steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve generated by the TTT curve generation means in the prediction model. The TTT curve generation means generates the TTT curve for the target steel material by interpolating between a plurality of nodes represented by time and temperature using a curve represented by a polynomial. A program characterized by optimizing the time and temperature of a plurality of the nodes based on an evaluation function including the difference between the actual temperature value included in the cooling performance data regarding a steel material of the same or similar steel type as the target steel material and the temperature predicted by the prediction model using the cooling performance data.

8. A program for controlling the temperature of a steel material cooled by a cooling device, Using at least steel type-specific cooling performance data, a term representing the temperature drop amount of the steel material due to cooling, and a term representing the temperature change of the steel material due to transformation heat generation, and a prediction model for predicting the temperature of the steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve, a TTT curve generation means for generating the TTT curve for the target steel material, Temperature prediction means for predicting the temperature of the target steel material by reflecting the prediction result of the transformation heat generation amount based on the TTT curve generated by the TTT curve generation means in the prediction model, Causing a computer to function as control means for operating the cooling device so that the temperature of the target steel material predicted by the temperature prediction means becomes a preset target temperature, The TTT curve generation means generates the TTT curve for the target steel material by interpolating between a plurality of nodes represented by time and temperature using a curve represented by a polynomial. A program characterized by optimizing the time and temperature of a plurality of the nodes based on an evaluation function including the difference between the actual temperature value included in the cooling performance data regarding a steel material of the same or similar steel type as the target steel material and the temperature predicted by the prediction model using the cooling performance data.

Images