Rolling control device

Abstract

In the present invention, during rolling, a first actual results collection unit collects first exit-side sheet thickness measured values at a first management point of a rolling material, and rolling load measured values for each stand. During the rolling of the rolling material, a second actual results collection unit collects second exit-side temperature measured values at a second management point, which is farther toward the tail end side than the first management point. After the rolling of the rolling material, a first training unit trains a deformation resistance model, a rolling load model, and a mill elongation model on the basis of the first exit-side sheet thickness measured values and the rolling load measured values. After the rolling of the rolling material, a second training unit trains a temperature model on the basis of the second exit-side temperature measured values. Before rolling of the next material, a setting calculation unit calculates a roll gap setting value, a roll peripheral speed setting value, and a cooling water amount setting value, using the deformation resistance model, the rolling load model, and the mill elongation model that were trained by the first training unit, and the temperature model that was trained by the second training unit.

Description

Rolling control device

[0001] The present disclosure relates to a rolling control device, and more particularly to a rolling control device that controls the temperature and thickness of a rolled material on the delivery side of a finishing mill (finishing rolling mill).

[0002] The quality requirements for rolled material (hereinafter also referred to as "coil") rolled in a hot rolling line have become increasingly stringent in recent years. The thickness of the rolled material at the exit of the finishing mill (hereinafter also referred to as "exit thickness"), which is one of the quality requirements, affects the dimensional accuracy when processed into a final product, so the exit thickness must be precisely controlled. In addition, the temperature of the rolled material at the exit of the finishing mill (hereinafter also referred to as "exit temperature") affects the material properties of the rolled material, such as yield stress, tensile strength, and elongation, so the exit temperature must also be precisely controlled.

[0003] FIG. 7 is a schematic diagram showing the configuration of a conventional rolling control device. This device includes a setting calculation unit (FSU) 110 that performs setting calculations before the rolled material arrives at the finishing mill 4. The setting calculations calculate the roll gap setting value for each stand, the roll peripheral speed setting value (motor speed setting value), and the cooling water flow rate setting value for the inter-stand cooling device (ISC) so that the measured values ​​of the finisher delivery thickness (FDH) and the finisher delivery temperature (FDT) measured by a thickness gauge (hereinafter also referred to as the "outlet thickness gauge") 8 and a thermometer (hereinafter also referred to as the "outlet thermometer") 9, respectively, match their target values. This calculation uses a prediction model consisting of a set of mathematical formulas for predicting the temperature, deformation resistance, rolling load, and mill elongation of the rolled material. The deformation resistance is the stress required to deform the rolled material. The mill elongation is the change in the roll gap due to the elastic deformation of the rolling rolls. The set values ​​calculated in the setting calculation are set for the various pieces of equipment in the hot rolling line.

[0004] When the rolled material heated in the heating furnace passes through the roughing mill and reaches the finishing mill 4, rolling begins in the finishing mill 4. The rolled material is sequentially bitten into each stand of the finishing mill 4, and when the leading edge of the rolled material leaves the finishing mill, measurement of the outlet thickness FDH and outlet temperature FDT begins.

[0005] During rolling by the finishing mill 4, a monitor AGC (Automatic Gage Control) that calculates the amount of change in the roll gap based on the FDH measurement value and a gauge meter AGC (GM-AGC) that estimates the outlet thickness of each stand based on the rolling load and actual roll gap values ​​and keeps the estimated thickness value (gauge meter thickness) constant are executed by the feedback thickness control unit 117, and a feedback temperature control (FB-FDTC) that calculates the amount of change in the roll peripheral speed and cooling water flow rate of each stand based on the outlet temperature measurement value is executed by the feedback temperature control unit 118. As a result, the outlet thickness measurement value and the outlet temperature measurement value are controlled to approach their respective target values. Such feedback control is disclosed, for example, in Patent Document 1 listed below.

[0006] In the finishing mill 4, the leading edge of the rolled material is rolled before feedback control is performed. Therefore, the accuracy of the outlet thickness FDH and outlet temperature FDT at the leading edge of the rolled material depends solely on the respective set values ​​calculated by the setting calculation unit (FSU) 110. The rolling material temperature control device disclosed in the following Patent Document 2 uses a convergence calculation method to improve the accuracy of the temperature model used in the setting calculation. That is, before the rolled material arrives at the finishing mill, the outlet temperature is calculated when a calculation point (control point) on the rolled material reaches a predetermined position on the finishing mill's outlet side. This calculated value is compared with the target outlet temperature value. If the compared value is outside the allowable range, the calculated motor speed value is corrected. Accordingly, the finishing mill's inlet temperature, the heat balance of the rolls, and the heat balance between stands are recalculated (updated), resulting in highly accurate temperature prediction values.

[0007] Incidentally, a conventional practice is to set an arbitrary point near the tip of the rolled material as a control point, collect FDT measurement values ​​measured when this control point is directly below the delivery thermometer 9, learn a temperature model using the collected FDT measurement values, and reflect the learned values ​​in the setting calculations for the next rolled material (hereinafter also referred to as the "next material"), thereby making the FDT measurement value of the control point at the next material closer to the FDT target value. Furthermore, a conventional practice is to collect rolling load measurement values ​​measured when the control point is directly below each stand, and collect FDH measurement values ​​measured when the control point is directly below the delivery thickness gauge 8, use these measurement values ​​to learn models for deformation resistance, rolling load, and mill elongation at each stand, and reflect the learned values ​​in the setting calculations for the next material, thereby making the FDH measurement value of the control point at the next material closer to the FDH target value.

[0008] Japanese Patent No. 3657750 Japanese Patent Application Laid-Open No. 2000-210708

[0009] However, the temperature of the front end portion of the rolled material extracted from the heating furnace may be locally high (or low). In this case, as shown in Figure 8, the outlet temperature FDT of the front end portion of the rolled material becomes higher (or lower) than the outlet temperature FDT of the tail end portion (steady state portion) of the rolled material. That is, a temperature abrupt change portion, where the temperature transitions from a high temperature portion (or low temperature portion) to a steady state portion, occurs at the front end portion of the rolled material. Such a temperature abrupt change portion occurs because the temperature difference between the surface and the interior of the rolled material increases due to the rapid heating of the rolled material in the heating zone of the heating furnace, or because the time the rolled material spends in the soaking zone of the heating furnace (residence time) is insufficient. In recent years, there has been a trend to shorten the residence time in the furnace in order to shorten the rolling interval (rolling pitch) and increase production, making it virtually impossible to avoid the occurrence of a temperature abrupt change portion.

[0010] When a sudden temperature change occurs at the tip of the rolled material, the learning unit can consider the following first and second learning methods. In the first learning method, the high-temperature portion of the rolled material is designated as a control point (hereinafter referred to as the "first control point") Mp1, and FDT and FDH measurement values ​​measured when the first control point Mp1 is located directly below the exit thermometer 9 and the exit thickness gauge 8 are collected, and learning is performed using the collected measurement values. That is, the setting calculation unit 110 calculates various setting values ​​so that the FDT and FDH measurement values ​​at the first control point Mp1 match the target values. As a result, as shown by the dashed line L1 in FIG. 3 , although the FDT measurement value at the first control point Mp1 matches the target value at time t1, the FDT measurement value drops after time t1. Even if FB-FDTC is started at time t1, its effect does not begin to be seen until time t3, and it takes time for the FDT measurement value, which has once dropped, to recover to within the allowable range at time t4. This is because the FB-FDTC cannot keep up with changes in the FDH measurement value, resulting in a long dead time before the cooling water flow rate or roll speed is changed. Furthermore, as shown by the dashed line L1 in Figure 4, the FDH measurement value at the first control point Mp1 at time t1 coincides with the target value. By storing (locking on) the gauge meter thickness in the area where the deviation between the FDH measurement value and the FDH target value at time t1 is small and executing GM-AGC, this, combined with the use of a highly responsive hydraulic screw-down device as the control terminal, can suppress thickness fluctuations in areas where the temperature suddenly changes. Furthermore, by using the monitor AGC in combination, the FDH measurement value can be maintained close to the FDH target value throughout the entire length of the rolled material.

[0011] Next, in the second learning method, the tail end of the sudden temperature change portion of the rolled material is designated as a control point (hereinafter referred to as the "second control point") Mp2, and FDT and FDH measurement values ​​measured when the second control point Mp2 is located directly below the outlet thermometer 9 and the outlet thickness gauge 8 are collected, and learning is performed using the collected FDT and FDH measurement values. That is, the setting calculation unit 110 calculates various setting values ​​so that the FDT and FDH measurement values ​​at the second control point Mp2 coincide with their target values. As a result, as shown by the dashed-dotted line L2 in FIG. 3 , the FDT measurement value at the second control point Mp2 coincides with the FDT target value at time t2, and the FDT measurement values ​​at and after the second control point Mp2 are also controlled to be close to the target values. On the other hand, although the deviation between the FDT measurement value and the FDT target value near the first control point Mp1 is relatively large, the length of the temperature defect portion outside the allowable range can be shorter than in the first learning method. Furthermore, as described above, because the FDT deviation at the first control point Mp1 is large, the models for deformation resistance, rolling load, and mill elongation are calculated based on the rolled material temperature at each stand under the assumption that the measured FDT value matches the target FDT value. However, FDT deviations occur during actual rolling (in this example, the measured FDT value is higher than the target FDT value). As a result, the actual rolled material temperature at each stand differs from the prediction (is higher than the prediction), and the deformation resistance, rolling load, and mill elongation also differ from the prediction (are lower than the prediction). As a result, the FDH deviation at the first control point Mp1 becomes large (the measured FDH value is thinner than the target FDH value), as shown by the dashed-dotted line L2 in FIG. 4 . Furthermore, the GM-AGC memorizes (locks on) the gauge meter thickness of the portion with the large FDH deviation near the first control point Mp1 and attempts to maintain that thickness, resulting in a prolonged period of thickness defects outside the allowable range. When the leading edge of the rolled material reaches the delivery thickness gauge 8 and the monitor AGC is started, the thickness deviation gradually decreases from time t3, but the defective thickness portion continues for a relatively long period of time.

[0012] Thus, there was a problem that the first learning method resulted in a longer portion of the outlet temperature defect, and the second learning method resulted in a longer portion of the outlet thickness defect.

[0013] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a rolling control device that can shorten the delivery temperature defective portion as much as possible without lengthening the delivery thickness defective portion, even when a temperature sudden change portion exists at the leading end of the rolled material.

[0014] The first aspect relates to a rolling control device that controls the temperature and thickness of a rolled material at the delivery side of a finishing mill. The finishing mill includes multiple stands having rolls, multiple cooling devices arranged between the stands, and a rolling load measuring instrument that measures the rolling load of each stand. An delivery thickness gauge that measures the thickness of the rolled material and an delivery thermometer that measures the temperature of the rolled material are respectively arranged at the delivery side of the finishing mill. The rolling control device includes a setting calculation unit, a first result collection unit, a second result collection unit, a first learning unit, and a second learning unit. Before rolling the rolled material, the setting calculation unit calculates the roll gap setting value and roll peripheral speed setting value of each stand and the cooling water flow setting value of each cooling device using a target delivery thickness value, a target delivery temperature value, and a prediction model. A portion at the tip of the rolled material where the temperature measured by the delivery thermometer transitions from a high-temperature portion or a low-temperature portion, which is higher or lower than the steady-state portion, to the steady-state portion is defined as a temperature sudden change portion, and the high-temperature portion or the low-temperature portion is defined as a first control point. The first result collection unit collects, during rolling, first exit thickness measurements taken by an exit thickness gauge at a first control point and rolling load measurements of each stand taken by a rolling load measuring instrument. The second result collection unit collects, during rolling of the rolled material, second exit temperature measurements taken by an exit thermometer at a second control point on the tail end side of the first control point. After rolling the rolled material, the first learning unit learns the deformation resistance model, rolling load model, and mill elongation model included in the prediction model based on the first exit thickness measurements and rolling load measurements collected by the first result collection unit. After rolling the rolled material, the second learning unit learns the temperature model included in the prediction model based on the second exit temperature measurements collected by the second result collection unit. The setting calculation unit is configured to calculate the roll gap setting value, roll peripheral speed setting value and cooling water amount setting value before rolling the next material using the deformation resistance model, rolling load model and mill elongation model learned by the first learning unit and the temperature model learned by the second learning unit.

[0015] The second aspect has the following features in addition to the first aspect. The first result collection unit further collects first outlet temperature measurement values ​​measured by an outlet thermometer at the first control point. The rolling control device further includes an outlet temperature prediction unit that predicts the finish mill outlet temperature at the first control point based on an outlet temperature target value, and a third learning unit that, after rolling the rolled material, learns a predicted outlet temperature value predicted by the outlet temperature prediction unit based on a difference between the first outlet temperature measurement value and the second outlet temperature measurement value. The setting calculation unit is configured to predict the rolling load and mill elongation of each stand and calculate a roll gap setting value, before rolling the next material, using the predicted outlet temperature value at the first control point learned by the third learning unit and the deformation resistance model, rolling load model, and mill elongation model learned by the first learning unit.

[0016] Incidentally, there are cases where a plurality of heating furnaces for heating the rolled material are arranged in parallel upstream of the finishing mill. The occurrence of the sudden temperature change portion near the tip of the rolled material depends on the operation of the heating furnaces, as described above. Therefore, the third aspect further has the following features in addition to the second aspect: at least one of the first learning unit, the second learning unit, and the third learning unit has a stratified learning value table divided for each heating furnace, and during learning, updates the learning value for the division corresponding to the heating furnace from which the rolled material to be learned is extracted, and the setting calculation unit calculates each setting value using the learning value for the division of the heating furnace from which the rolled material to be learned is extracted.

[0017] According to the first aspect, by learning a temperature model used to calculate the cooling water flow rate setting value in the setting calculation unit based on the second outlet temperature measurement value measured at the second control point, it is possible to minimize the deviation between the second outlet temperature measurement value at the second control point of the next strip and the target outlet temperature. As a result, although an outlet temperature deviation occurs at the first control point of the next strip, the length of the temperature-defective portion at the tip of the rolled strip can be significantly shortened compared to when learning is performed using the first outlet temperature measurement value measured at the first control point. Furthermore, by learning a deformation resistance model, a rolling load model, and a mill elongation model used to calculate the roll gap setting value in the setting calculation unit based on the first outlet thickness measurement value measured at the first control point, it is possible to minimize the deviation between the first outlet thickness measurement value at the first control point of the next strip and the target outlet thickness. Therefore, even if a temperature-change portion exists at the tip of the rolled strip, it is possible to minimize the outlet temperature-defective portion without lengthening the outlet thickness-defective portion. Furthermore, by further using the existing gauge meter AGC and monitor AGC, it is possible to make the outlet thickness measurement value coincide with the outlet thickness target value over the entire longitudinal direction of the next material.

[0018] According to the second aspect, by calculating the delivery temperature prediction value based on the difference between the first delivery temperature measurement value and the second delivery temperature measurement value, in combination with learning the deformation resistance model, the rolling load model, and the mill elongation model, the roll gap setting value can be calculated with higher accuracy. As a result, the delivery thickness measurement value can be made to more closely match the delivery thickness target value.

[0019] According to the third aspect, when a plurality of heating furnaces are arranged in parallel, each learning unit can learn using an optimum learning value according to the operation of each heating furnace.

[0020] FIG. 1 is a schematic diagram showing the configuration of a hot rolling line to which a rolling control device according to an embodiment is applied. FIG. 2 is a schematic diagram showing the configuration of a process control computer which is a rolling control device according to an embodiment. FIG. 3 is a diagram for explaining changes in finish mill delivery temperature. FIG. 4 is a diagram for explaining changes in finish mill delivery plate thickness. FIG. 5 is a diagram showing an example of the hardware configuration of a process control computer which implements the rolling control device. FIG. 6 is a diagram showing a learning value table. FIG. 7 is a schematic diagram showing the configuration of a conventional rolling control device. FIG. 8 is a diagram for explaining a sudden temperature change portion which occurs at the tip of a rolled material.

[0021] Hereinafter, an embodiment of the present disclosure will be described with reference to the drawings, taking as an example a case where the present disclosure is applied to a hot rolling line RL. Note that elements common to the various drawings are assigned the same reference numerals, and duplicated explanations will be omitted.

[0022] 1 is a schematic diagram showing the configuration of a hot rolling line RL to which a rolling control device according to an embodiment of the present invention is applied. The hot rolling line RL includes, as main equipment, at least one heating furnace 1, a roughing mill 2, a descaling device 3, a finishing mill 4, a water-cooled table 5, and a coiler 6.

[0023] The heating furnace 1 heats a rolled material (slab) M. The rolled material M is a metal material such as steel having a thickness of approximately 170 mm to 280 mm. The heating furnace 1 heats the interior of the furnace to a high temperature by burning heavy oil or gas, and the rolled material M is heated to approximately 1200°C by passing through the furnace for, for example, approximately 3 to 4 hours. The first half of the furnace is at a high temperature and is called the heating zone, and is responsible for raising the temperature of the rolled material M. The second half of the furnace is at a lower temperature than the heating zone and is called the soaking zone, and is responsible for uniforming the temperature of the rolled material M through heat conduction within the rolled material M. As mentioned above, if the rolled material M is rapidly heated in the heating zone and its residence time in the soaking zone is short, a sudden temperature change will occur at the tip of the rolled material M removed from the heating furnace 1.

[0024] The roughing mill 2 has one or two stands R1 and R2. The roughing mill 2 repeatedly rolls the rolled material M while reversing the direction of travel, processing it into a rolled material (intermediate bar) M with a thickness of approximately 20 mm to 45 mm. The descaling device 3 removes oxide scale from the surface of the rolled material M by spraying high-pressure water (descaling water). The descaling device 3 has multiple headers that spray high-pressure water, and each header can be turned on and off. The outlet temperature FDT can be changed by controlling the number of headers that are turned on and changing the flow rate of high-pressure water.

[0025] The finishing mill 4 includes a plurality of stands Fi (i = 1 to 7) (seven in the example shown in FIG. 1 ). Each stand Fi has a pair of upper and lower work rolls (rolling rolls) 41, a pair of upper and lower backup rolls 42, and a motor 43 for driving the work rolls 41 to rotate. The backup rolls 42 are provided with hydraulically controlled screw-down devices (hydraulic cylinders) 44 so that the roll gap can be changed. Each stand Fi also has a rolling load measuring device 45 that measures the rolling load of each stand Fi. The rolling load measuring device 45 can be configured as a load cell, but can also be configured to calculate the rolling load from the hydraulic pressure of the hydraulic screw-down device 44.

[0026] A plurality of cooling devices ISCi (i = 1 to 6) are arranged between the stands Fi of the finishing mill 4. Each cooling device ISCi is also called an inter-stand spray, and the cooling water flow rate can be changed using a proportional valve or the like. Changing the cooling water flow rate of any cooling device ISCi changes the outlet temperature FDT. Although not shown, induction heating devices (IH) can be installed on the inlet side of the finishing mill 4 or between the stands Fi. In this case, the outlet temperature FDT can be changed by controlling the supplied power to change the heating capacity.

[0027] The water-cooled table 5 uses cooling water to cool the rolled material M that has passed through the finishing mill 4. The coiler 6 winds the rolled material M into a coil product.

[0028] An entry thermometer 7 is arranged at the entry side of the finishing mill 4, and is capable of measuring the temperature FET of the rolled material M at the entry side of the finishing mill 4 (hereinafter also referred to as the "entry temperature") . An exit thickness gauge 8 and an exit thermometer 9 are arranged at the exit side of the finishing mill 4, and are capable of measuring the exit thickness FDH and the exit temperature FDT.

[0029] The hot rolling line RL is operated by a control system using a computer. The computer includes a host computer 10 and a process control computer 11, which are connected to each other via a network. An interface screen 12, which is an operation screen for an operator, is connected to the process control computer 11 via the network. The operator can input control conditions and the like on the interface screen 12.

[0030] The host computer 10 determines the steel type of the rolled material M, the target finish mill entry thickness value, the target finish mill exit thickness value (product thickness), and the target finish mill exit temperature value in accordance with the operation plan, and transmits the determined items as initial information to the process control computer 11. The initial information also includes the heating conditions of the heating furnace 1.

[0031] In the hot rolling line RL, the outlet thickness FDH and outlet temperature FDT of the rolled material M are important product indices, and it is desirable to shorten as much as possible the defective portion where the deviation from the target values ​​is outside the allowable range.

[0032] The process control computer 11 calculates setting values ​​for each piece of equipment that will achieve the FDH target value and the FDT target value, based on the initial information from the computer 10 and the control conditions and the like provided from the interface screen 12, and sets the calculated setting values ​​for each piece of equipment. While each piece of equipment is operating, the process control computer 11 corrects (changes) each setting value in accordance with values ​​obtained from various measuring instruments. The process control computer 11, which is a rolling control device, will be described in detail below.

[0033] Figure 2 is a schematic diagram showing the configuration of a process control computer 11, which is a rolling control device according to an embodiment. Figure 3 is a diagram for explaining changes in the finish mill delivery temperature FDT. Figure 4 is a diagram for explaining changes in the finish mill delivery thickness FDH. As shown in Figures 3 and 4, the portion at the front end of the rolled material M where the FDT measurement value transitions from a high-temperature portion higher than the steady-state portion to the steady-state portion is defined as the temperature abrupt change portion, the high-temperature portion of the rolled material M is defined as the first control point, and the second control point is defined as the tail end side of the temperature abrupt change portion.

[0034] The process control computer 11 includes a setting calculation unit 110, a first result collection unit 111, a second result collection unit 112, a first learning unit 113, a second learning unit 114, a third learning unit 115, an outlet temperature prediction unit 116, an AGC 117, and an FDTC 118. The feedback thickness control unit (AGC) and the feedback temperature control unit (FDTC) have already been described, and therefore detailed description thereof will be omitted here.

[0035] Based on initial information (e.g., the steel type of the rolled material M, the FET target value, the FDH target value, and the FDT target value) input from the host computer 10, the setting calculation unit (FSU) 110 uses a prediction model to calculate at least one set value of the roll peripheral speed of each stand Fi, the cooling water flow rate of the interstand cooling device ISCi, the descaling water flow rate of the descaling device 3, and the heating power of the induction heating device, as well as the roll gap set value of each stand Fi, before rolling the rolled material M (before the rolled material M enters the finishing mill 4), and sets these values ​​for each device. These set values ​​are calculated taking into account water-cooling heat transfer by the interstand cooling water in the finishing mill 4, heat radiation, cooling by air convection, processing heat generated due to deformation of the rolled material M in each stand Fi, frictional heat generation and contact heat removal generated between the work roll 41 and the rolled material M, and a temperature model, which is a set of mathematical formulas representing the influence of these factors. Temperature models are publicly known, as disclosed in, for example, Patent Document 2. The setting calculation unit 110 first sets the predetermined set values ​​of the roll peripheral speed and cooling water flow rate of the final stand F7 as initial values, and if there is sufficient margin in the cooling water flow rate range, adjusts the set value of the cooling water flow rate so that the outlet temperature FDT at the tip of the rolled material M becomes the FDT target value. If the set value of the cooling water flow rate reaches its limit and simply changing the cooling water flow rate is not sufficient, the setting calculation unit 110 changes the roll peripheral speed of the final stand F7 so that the outlet thickness FDT becomes the FDT target value. At this time, the set value of the roll peripheral speed of each stand Fi is calculated so that the volume velocity of the rolled material M between each stand Fi is equal. In this way, the set value of the roll peripheral speed of the final stand F7 and the set value of the cooling water flow rate between each stand Fi are determined, and then each device of the finishing mill 4 is set based on these set values ​​until the tip of the rolled material M reaches the finishing mill 4.

[0036] During rolling, the first performance data collection unit 111 collects (samples) the first FDT measurement value measured by the delivery thermometer 9 and the rolling load measurement value measured by the rolling load measuring device 45 of each stand Fi at the first control point Mp1. The position of the first control point Mp1 is determined in advance so that it is located before the sudden temperature change portion (e.g., the high-temperature portion) of the rolled material M, which has a sudden temperature change portion near its leading edge. In each stand Fi, the position of the sudden temperature change portion is tracked taking into account elongation due to changes in thickness caused by rolling. When determining the first control point Mp1, it should be noted that if the first control point Mp1 is too close to the leading edge, the measurement of the first FDT measurement value by the delivery thermometer 9 may become unstable. Sampling is performed based on the position of the rolled material M detected by sensors (not shown) on the hot rolling line RL and transport time information. Since the method for detecting the position of the rolled material M is well known, a detailed description thereof will be omitted here.

[0037] The second result collection unit 112 collects a first FDT measurement value measured by the delivery thermometer 9 at the first control point Mp1 during rolling, and also collects a second FDT measurement value measured by the delivery thermometer 9 at the second control point Mp2. The position of the second control point Mp2 is determined to be closer to the tail end than the range of variation, taking into account the variation in the position of the sudden temperature change part.

[0038] The first learning unit 113 learns the rolling force model based on the rolling force measurement values ​​of each stand Fi at the first control point Mp1 collected by the first result collection unit 111. There are various methods for this learning, but for example, as shown in the following formula (1), a ratio is calculated by dividing the rolling force measurement value by the rolling force prediction value (predicted value before correction by learning), and this is multiplied by a first learned value Z1b before updating and a predetermined proportional division rate (β １ ) to obtain a new (updated) first learned value Z1a. The first learned value Z1 is, for example, a parameter of a group of equations describing the rolling force model. Z1a=Z1b×(1−β １ ) + (measured rolling load value / predicted rolling load value) × β １ ...(1)

[0039] The second learning unit 114 learns the temperature model based on the second FDT measurement value of the second control point Mp2 collected by the second result collection unit 112. There are various methods for this learning, but for example, as shown in the following formula (2), a ratio is calculated by dividing the difference between the second FDT measurement value and the actual inlet temperature value at the second control point Mp2 by the difference between the FDT predicted value (the predicted value before correction by learning) and the actual inlet temperature value, and the calculated ratio is multiplied by a predetermined proportional ratio (β ２ ) to obtain a new (updated) second learned value Z2a. Z2a=Z2b×(1−β ２ ) + {(FET actual value - second FDT measured value) / (FET actual value - FDT predicted value)} × β ２ ... (2)

[0040] Here, the FET actual value that serves as the starting point for the calculation of the temperature model can be, but is not limited to, the FET measurement value measured by the inlet thermometer 7. The FET actual value can be a proportional division of the FET measurement value and the FET estimated value. The FET estimated value can be estimated based on the state of the rolled material M on the outlet side of the roughing mill 2 and the transport state from the roughing mill 2 to the finishing mill 4 (transport time, state of the heat-retaining cover, etc.).

[0041] The third learning unit 115 calculates the difference between the first FDT measurement value (FDT1) of the first control point Mp1 and the second FDT measurement value (FDT2) of the second control point Mp2, for example, as in the following equation (3), and divides the calculated difference by a predetermined proportional ratio (β ３ ) and set as a new (updated) third learned value Z3a. Z3b=Z3b×(1−β ３ )+(FDT1-FDT2)×β ３ ...(3)

[0042] Before rolling the next rolled material (next material) M, the third learning unit 115 corrects the FDT predicted value predicted by the outlet temperature prediction unit 116 (described later) using the temperature model learned by the second learning unit 114, for example, as in the following equation (4). Note that if the FET actual value has already been obtained at the time of the setting calculation, the FET actual value can be used instead of the FET predicted value. FDT predicted value (after learning is reflected) = FET predicted value - (FET predicted value - FDT predicted value (before learning is reflected) × second learning value (4)

[0043] The setting calculation unit 110 calculates at least one setting value of the roll peripheral speed, the ISC cooling water flow rate, the descaling water flow rate of the descaling device, and the heating power of the induction heating device so that the FDT predicted value (after learning is reflected) coincides with the FDT target value. The roll peripheral speed of each stand Fi (i = 1 to 6) other than the final stand F7 can be calculated from the roll peripheral speed of the final stand F7 using the forward slip ratio and the law of constant volume velocity, as shown in the following equation (5), for example: Fi roll peripheral speed = {(1 + F7 forward slip ratio) × F7 roll peripheral speed × F7 delivery thickness} / {(1 + Fi forward slip ratio) × Fi delivery thickness} ... (5)

[0044] The forward slip is the deviation rate between the roll peripheral speed of the stand Fi and the rolling material speed at the exit of the stand Fi. The forward slip can be calculated based on the entry and exit thicknesses of the stand Fi, etc., using a theoretically or experimentally derived forward slip model.

[0045] The outlet temperature prediction unit 116 predicts the outlet temperature of the finishing mill 4 at the first control point Mp1 based on the target outlet temperature value. The outlet temperature prediction unit 116 calculates an FDT predicted value using the learning result of the third learning unit 115, as shown in the following equation (6): FDT predicted value = FDT predicted value (after learning is reflected) + third learned value (6)

[0046] The setting calculation unit 110 also calculates the temperature of each stand Fi at the first control point Mp1 using the FDT predicted value, and calculates the rolling load predicted value (after learning is reflected) of each stand Fi using the learning result of the first learning unit 113, as shown in the following formula (7): Rolling load predicted value (after learning is reflected) = Rolling load predicted value (before learning is reflected) × First learned value (7)

[0047] The setting calculation unit 110 uses the rolling load prediction value (after learning is reflected) of each stand Fi to predict the mill elongation of each stand Fi and calculates the roll gap setting value of each stand Fi.

[0048] By performing the setting calculation for the next material in this manner, it is possible to reduce the FDT deviation over the entire length of the rolled material M, including near the tip. Although FDT deviation may occur in the portion before the temperature abrupt change portion, the length of the temperature-defective portion where the FDT measurement value falls outside the allowable range is significantly shorter than with the conventional first learning method. Furthermore, the FDH measurement value at the first control point Mp1 can be accurately matched to the FDH target value. Here, by reflecting the learning results in the temperature calculation of the first control point Mp1 of each stand (expressed as a position tracked taking into account elongation due to rolling), it is possible to predict and calculate with high accuracy not only the temperature of the first control point Mp1 of each stand, but also the deformation resistance, rolling load, and roll gap calculated using the temperature. This allows the FDH measurement value at the first control point Mp1 to be accurately matched to the FDH target value. Furthermore, during rolling, a gauge meter AGC is used to estimate changes in the elastic deformation of the rolling rolls based on the rolling load measurement value and to control the thickness by manipulating the roll gap. This suppresses thickness fluctuations in areas where the temperature suddenly changes, thereby maintaining a constant outlet thickness FDH. This gauge meter AGC operates the roll gap using a hydraulic screw-down device 44 with a fast response speed, so it can quickly respond to sudden changes in the rolling load measurement value and suppress fluctuations in the outlet thickness FDH. A monitor AGC is also used in conjunction with this, which operates the roll gap of each stand Fi based on the FDH measurement value to reduce FDH deviation. However, there is a distance of several meters between the final stand F7 of the finishing mill 4, which is the target of the finishing operation, and the outlet thickness gauge 8, and the monitor AGC cannot be applied to the area corresponding to this distance. Despite these limitations, the FDH deviation can be further reduced by also using the monitor AGC.

[0049] The specific structure of the process control computer 11 is not limited, but may be as follows, for example. FIG. 5 is a diagram showing an example of the hardware configuration of the process control computer 11. The functions of the process control computer 11 can be realized by the processing circuit shown in FIG. 5. This processing circuit may be dedicated hardware 11a. This processing circuit may include a processor 11b and a memory 11c. This processing circuit may be partially formed as dedicated hardware 11a and further include a processor 11b and a memory 11c. In the example of FIG. 5, part of the processing circuit is formed as dedicated hardware 11a, and the processing circuit also includes a processor 11b and a memory 11c.

[0050] At least part of the processing circuitry may be at least one piece of dedicated hardware 11a, which may be, for example, a single circuit, a multiple circuit, a programmed processor, a parallel programmed processor, an ASIC, an FPGA, or a combination thereof.

[0051] The processing circuit may include at least one processor 11b and at least one memory 11c. In this case, each function of the process control computer 11 is realized by software, firmware, or a combination of software and firmware. The software and firmware are written as programs and stored in the memory 11c. The processor 11b realizes the functions of each part of the rolling control device 11 by reading and executing the programs stored in the memory 11c.

[0052] The processor 11b is also called a CPU (Central Processing Unit), central processing unit, processing unit, arithmetic unit, microprocessor, microcomputer, or DSP. The memory 11c is, for example, a non-volatile or volatile semiconductor memory such as a RAM, a ROM, a flash memory, an EPROM, or an EEPROM.

[0053] In this way, the processing circuit can realize each function of the rolling control device 11 by hardware, software, firmware, or a combination of these.

[0054] As described above, according to the present disclosure, by learning the temperature model used to calculate the cooling water flow rate set value in the setting calculation unit 110 based on the second FDT measurement value measured at the second control point Mp2, it is possible to minimize the deviation between the second FDT measurement value at the second control point Mp2 of the subsequent material and the FDT target value. As a result, although an FDT deviation occurs in the portion of the subsequent material at the first control point Mp1, the length of the temperature-defective portion at the tip of the rolled material can be significantly shortened compared to when learning is performed using the first FDT measurement value measured at the first control point Mp1. Furthermore, by learning the deformation resistance model, rolling load model, and mill elongation model used to calculate the roll gap set value in the setting calculation unit 110 based on the first FDH measurement value measured at the first control point Mp1, it is possible to minimize the deviation between the first FDH measurement value at the first control point Mp1 of the subsequent material and the FDH target value. In this way, by changing the first control point Mp1 where FDH measurement values ​​are collected and the second control point Mp2 where FDT measurement values ​​are collected, even if a temperature sudden change portion exists at the tip of the rolled material, it is possible to shorten the temperature defective portion as much as possible without lengthening the thickness defective portion (i.e., without reducing FDH controllability). Furthermore, by using FB-FDTC in combination, it is possible to match the FDT measurement value with the FDT target value over the entire longitudinal direction of the next material. Furthermore, by further using existing gauge meter AGC and monitor AGC in combination, it is possible to match the FDH measurement value with the FDH target value over the entire longitudinal direction of the next material.

[0055] Although the embodiments of the present disclosure have been described above, the present disclosure is not limited to the above embodiments and can be implemented in various modifications without departing from the spirit of the present disclosure. When the numbers, quantities, amounts, ranges, etc. of each element are mentioned in the above embodiments, the present invention is not limited to the mentioned numbers unless otherwise specified or clearly specified in principle. Furthermore, the structures, etc. described in the above embodiments are not necessarily essential to the present invention unless otherwise specified or clearly specified in principle.

[0056] In the above embodiment, the hot rolling line RL has been described as having one heating furnace 1, but there may be cases where multiple heating furnaces (not shown) are installed side by side. The occurrence of the sudden temperature change portion at the tip of the rolled material M depends on the operation of the heating furnace 1, as described above. In this case, at least one of the first learning unit 113, the second learning unit 114, and the third learning unit 115 can be configured to have a stratified learning value table divided for each heating furnace (divided into three sections 1A, 1B, and 1C in the figure), as shown in FIG. 6 . FIG. 6 is a diagram showing the learning value table. During learning, the learning values ​​Z1 to Z3 of the section corresponding to the heating furnace from which the rolled material M to be learned is extracted are updated, and the setting calculation unit 110 can be configured to calculate each setting value using the learning values ​​Z1_1 to Z3_3 of the sections 1A, 1B, and 1C of the heating furnace 1 from which the rolled material M to be learned is extracted. This allows each of the learning units 113 to 115 to learn using the optimum learning value according to the operation of each heating furnace 1.

[0057] 1...heating furnace, 2...roughing mill, 3...descaling device, 4...finishing mill, Fi...stand, 41...work roll, 42...backup roll, 43...motor, 44...screw down device, 45...rolling load measuring instrument, Fi...stand, ISCi...cooling device, 5...water-cooled table, 6...coiler, 7...entry thermometer, 8...exit thickness gauge, 9...exit thermometer, 11...rolling control device, process control computer, 110...setting calculation unit, 111...first result collection unit, 112...second result collection unit, 113...first learning unit, 114...second learning unit, 115...third learning unit, 116...exit temperature prediction unit, 117...feedback thickness control unit (AGC), 118...feedback temperature control unit (FDTC), M...rolled material

Claims

1. A rolling control device for controlling the temperature and thickness of a rolled material on the output side of a finishing mill, wherein the finishing mill includes a plurality of stands having rolls, a plurality of cooling devices arranged between the stands, and a rolling load measuring device for measuring the rolling load of each stand, and an output side thickness gauge for measuring the thickness of the rolled material and an output side thermometer for measuring the temperature of the rolled material are respectively arranged on the output side of the finishing mill. Before rolling the rolled material, a setting calculation unit calculates the roll gap setting value and the roll peripheral speed setting value of each stand, and the cooling water amount setting value of each cooling device by using the output side thickness target value, the output side temperature target value, and a prediction model. A portion where the temperature measured by the output side thermometer at the tip of the rolled material transitions from a high temperature part or a low temperature part that is higher or lower than the steady part to the steady part is defined as a temperature rapid change part, the high temperature part or the low temperature part is defined as a first management point, and during rolling, a first performance collection unit collects the first output side thickness measurement value measured by the output side thickness gauge and the rolling load measurement values of each stand measured by the rolling load measuring device at the first management point. During rolling of the rolled material, a second performance collection unit collects the second output side temperature measurement value measured by the output side thermometer at a second management point on the trailing end side of the first management point. After rolling the rolled material, a first learning unit performs learning of the deformation resistance model, the rolling load model, and the mill elongation model included in the prediction model based on the first output side thickness measurement value and the rolling load measurement values collected by the first performance collection unit. After rolling the rolled material, a second learning unit performs learning of the temperature model included in the prediction model based on the second output side temperature measurement value collected by the second performance collection unit. The setting calculation unit is configured to calculate the roll gap setting value, the roll peripheral speed setting value, and the cooling water amount setting value by using the deformation resistance model, the rolling load model, and the mill elongation model learned by the first learning unit and the temperature model learned by the second learning unit before rolling the next material.

2. The rolling control device according to claim 1, wherein the first performance collection unit further collects, at the first management point, a first outlet temperature measurement value measured by the outlet thermometer. A first outlet temperature prediction unit that predicts the outlet temperature of the finishing mill at the first management point based on the outlet temperature target value; A third learning unit that learns the outlet temperature prediction value predicted by the outlet temperature prediction unit based on the difference between the first outlet temperature measurement value and the second outlet temperature measurement value after rolling the rolled material. The setting calculation unit predicts the rolling load and mill elongation of each stand and calculates the roll gap setting value using the outlet temperature prediction value of the first management point learned by the third learning unit, the deformation resistance model, the rolling load model, and the mill elongation model learned by the first learning unit before rolling the next material. A rolling control device configured as such.

3. The rolling control device according to claim 2, wherein a plurality of heating furnaces for heating the rolled material are arranged in parallel upstream of the finishing mill. At least one of the first learning unit, the second learning unit, and the third learning unit has a hierarchical learning value table classified for each heating furnace, and during learning, the learning value of the section corresponding to the heating furnace from which the rolled material to be learned is extracted is updated. The setting calculation unit is configured to calculate each setting value using the learning value of the section of the heating furnace from which the rolled material to be the setting calculation target is extracted.

Images