JP7215603B2 - Induction heating method and induction heating system - Google Patents

Description

The present invention relates to a method and system for induction heating of steel in a hot rolling process.

In general, the mechanical properties of steel manufactured through a hot rolling process are strongly related to the temperature of the steel during the process. Therefore, temperature control of steel materials in the hot rolling process is important from the viewpoint of improving the quality of products.

In recent years, many hot rolling lines have introduced induction heating devices (hereinafter also referred to as “IH devices”). By introducing the IH device, thermal rundown is suppressed. Thermal down is a phenomenon in which the temperature of the steel material decreases toward the tail end on the entry side of the finishing mill, mainly during the production of thin steel material. In addition, the occurrence of temperature unevenness in the longitudinal direction of the steel material due to heat transfer to the struts (skids) that support the slab in the heating furnace is reduced. Furthermore, when producing hard, high-grade steel, raising the temperature of the rolled material before finishing rolling reduces the rolling load, improves the strip threadability, and reduces the load on the finishing rolling mill.

Furthermore, the introduction of an IH device enables local heating of the steel material. The temperature rise by heating the tail end of the steel material has the effect of suppressing the occurrence of tail end constriction. The tail end drawing is a phenomenon in which the tail end meanders when passing through the finishing mill, and buckles when it comes into contact with the device. Since the temperature at the ends of the steel material in the width direction tends to decrease due to heat radiation, the temperature distribution in the width direction becomes uniform by heating. Therefore, an improvement in quality in the width direction is expected. Moreover, when the end portion is heated, it is possible to suppress the occurrence of a structure in which coarse grains and regular grains are mixed (that is, mixed grains) and cracks. In addition, it is also expected to have the effect of suppressing uneven wear of the rolls of the finishing mill.

IH devices are classified into two types, a solenoid type and a transverse type, according to the direction of the main magnetic flux interlinking with the steel material (that is, the alternating magnetic flux generated from the IH device). In the solenoid type IH device, the direction of the main magnetic flux coincides with the conveying direction of the steel material. In the transverse IH device, the direction of the main magnetic flux coincides with the thickness direction of the steel material.

Patent Documents 1 and 2 are exemplified as prior art relating to the transverse IH device. Patent Documents 1 and 2 disclose examples in which three IH devices are arranged along the transport direction. In Patent Literature 1, the overlapping amount of the IH devices in the width direction of the steel material is controlled. In Patent Document 2, the distance between the cores of the IH device in the width direction is controlled. Both the overlapping amount and the distance are controlled by changing the position of the IH device in the width direction.

Japanese Patent Application Laid-Open No. 2004-195497 Japanese Patent No. 4714658

However, the control of the IH device in Patent Document 1 is performed on the premise that the temperature distribution of the steel material in the width direction is bilaterally symmetrical. Therefore, when this precondition is not satisfied, it is difficult to make the temperature distribution uniform in the width direction. Further, in the control of the IH device in Patent Document 2, the heating of the edges in the width direction is ensured by the control of the edge heater which is separately performed. Therefore, in order to make the temperature distribution uniform in the width direction, it is necessary to coordinate the control of the IH device and the edge heater.

The present invention has been made to solve the above-mentioned problems, and provides a technique that can make the temperature distribution uniform only by controlling the IH device even when the temperature distribution in the width direction is not bilaterally symmetrical. intended to provide

The present invention is a transverse induction heating method and has the following features.
In the induction heating method, a combination of electric power supplied to each of a plurality of inductors provided in the conveying direction of the rolling line and heating the steel material and the position of each of the plurality of inductors in the width direction of the steel material. including setting steps.
The step of setting the combination includes:
calculating an entry-side predicted temperature indicating a predicted temperature of each of the plurality of sections of the steel material in the width direction at the position where the steel material starts to be heated by the plurality of inductors;
setting candidates for the combination;
Predicting, using a machine learning model, the amount of temperature rise in each of the plurality of categories when the candidate is adopted;
calculating, based on the predicted entry-side temperature and the amount of temperature increase, a predicted exit-side temperature indicating the predicted temperature of each of the plurality of sections at the end position of heating of the steel material by the plurality of inductors;
The candidate or the evaluation when an evaluation function having variables of the predicted delivery temperature and target temperatures of the plurality of sections at the heating end position as variables is less than a threshold, or the evaluation A step of adopting the candidate when the function is the minimum as the combination;
including.

In the step of calculating the predicted entry temperature, the predicted entry temperature may be calculated based on the actual entry temperature indicating the actual temperature of each section of the steel material upstream from the heating start position.

The steel material includes a current material located upstream of the heating start position and a previous material located downstream of the heating end position and heated by the inductors immediately before the current material is heated by the plurality of inductors. , may include
The step of setting the combination includes calculating a deviation between the predicted delivery-side temperature of the previous material and an actual delivery-side temperature indicating the actual temperature of each section of the previous material downstream of the heating end position. It may further contain:
In the step of calculating the predicted entry-side temperature, the predicted entry-side temperature may be calculated based on the temperature of each of the plurality of sections of the current material upstream from the heating start position and the deviation. good.

The present invention is a transverse induction heating system and has the following features.
The induction heating system is
a plurality of inductors provided in the conveying direction of the rolling line to heat the steel material;
A control device for controlling the temperatures of the plurality of sections of the steel material in the width direction based on the power supplied to each of the plurality of inductors and the position of each of the plurality of inductors in the width direction of the steel material. and,
Prepare.
The control device performs a setting calculation process for setting the combination of the power and the position.
The control device, in the setting calculation process,
calculating an entry-side predicted temperature indicating the predicted temperature of each of the plurality of sections at the position where heating of the steel material by the plurality of inductors is started;
setting candidates for the combination;
Using a machine learning model, predict the amount of temperature rise in each of the plurality of categories when the candidate is adopted,
calculating a predicted outlet temperature indicating a predicted temperature of each of the plurality of sections at the end position of heating of the steel material by the plurality of inductors, based on the predicted inlet temperature and the amount of temperature increase;
The candidate or the evaluation when an evaluation function having variables of the predicted delivery temperature and target temperatures of the plurality of sections at the heating end position as variables is less than a threshold, or the evaluation The candidate when the function is minimized is adopted as the combination.

The induction heating system may further include an entry-side thermometer that is provided upstream from the heating start position and measures an entry-side actual temperature that indicates the actual temperature of each section of the steel material.
The control device may calculate the predicted entry temperature based on the actual entry temperature in the setting calculation process.

The induction heating system may further include a delivery-side thermometer that is provided downstream of the heating end position and measures an actual delivery-side temperature that indicates the actual temperature of each section of the steel material.
The steel material includes a current material located upstream of the heating start position and a previous material located downstream of the heating end position and heated by the inductors immediately before the current material is heated by the plurality of inductors. , may include
The control device, in the setting calculation process,
calculating a deviation between the predicted output-side temperature of the previous material and the actual output-side temperature of the previous material;
The entry-side predicted temperature may be calculated based on the temperature of each of the plurality of sections of the current material upstream of the heating start position and the deviation.

According to the induction heating method of the present invention, the combination of the power supplied to each of the plurality of inductors and the position of each of the plurality of inductors is set. When setting the combination, the predicted temperature (that is, the predicted entry-side temperature) of each of the plurality of sections of the steel material in the width direction at the heating start position of the steel material is calculated. In addition, combination candidates are set. Furthermore, the amount of temperature increase in each of the plurality of sections when this candidate is adopted is predicted using a machine learning model. Furthermore, based on the calculated entry-side predicted temperature and the predicted amount of temperature rise, the predicted temperature (that is, the predicted delivery-side temperature) of each of the plurality of sections at the heating end position of the steel material is calculated. .

Then, a candidate when the evaluation function having variables of the output-side predicted temperature and the respective target temperatures of the plurality of sections at the heating end position (that is, the output-side target temperature) is less than the threshold, or the evaluation The candidate when the function is minimized is adopted as the combination of power and position. As described above, according to the induction heating method according to the present invention, an evaluation function having variables of the predicted outlet temperature calculated in consideration of the predicted inlet temperature and the amount of temperature increase and the target outlet temperature is used. can be used to determine the appropriate power and position combination. Therefore, even if the temperature distribution in the width direction is not symmetrical, it is possible to make the temperature distribution uniform at the heating end position only by controlling the inductor.

According to the induction heating system of the present invention, setting calculation processing for setting the combination of power and position is performed. According to the setting calculation process, an evaluation function whose variables are the output-side predicted temperature calculated in consideration of the input-side predicted temperature and the amount of temperature increase, and the output-side target temperature, is used to determine the appropriate power and position. A combination can be determined. Therefore, even if the temperature distribution in the width direction is not bilaterally symmetrical, it is possible to make the temperature distribution uniform at the heating end position only by executing the control of the inductor.

BRIEF DESCRIPTION OF THE DRAWINGS It is the schematic which shows an example of a structure of the hot rolling equipment in which the induction heating system which concerns on 1st Embodiment is applied. It is a schematic diagram showing an example of the composition of an induction heating device. It is a schematic diagram showing an example of the configuration of an inductor. FIG. 2 is a schematic diagram showing an example of arrangement of inductors; 1 is a schematic diagram showing an example of a node; FIG. It is a block diagram explaining an example of a structure of the function of a computer. It is the figure which showed an example of distribution of the amount of heat injected|thrown-in to steel material from an induction heating apparatus. 6 is a flowchart for explaining the flow of setting calculation processing according to the first embodiment; It is an example of a data table showing initial values. It is a schematic diagram showing an example of a temperature rise amount model. It is a schematic diagram showing an example of composition of hot-rolling equipment to which an induction heating system concerning a 2nd embodiment is applied. It is a block diagram explaining an example of a structure of the function of a computer. FIG. 4 is a schematic diagram showing an example of the relationship between two types of temperature distributions of previous materials and two types of temperature distributions of current materials. 10 is a flowchart for explaining the flow of setting calculation processing according to the second embodiment;

First Embodiment First, a first embodiment of the present invention will be described in detail with reference to FIGS. 1 to 10. FIG.

1. System configuration 1-1. Configuration of Hot Rolling Equipment FIG. 1 is a schematic diagram showing an example of the configuration of a hot rolling equipment to which an induction heating system (hereinafter also referred to as an "IH system") according to the first embodiment is applied. The induction heating method according to the first embodiment is implemented by the IH system according to the first embodiment.

Hot rolling equipment is equipment for processing steel material SS as a rolled material into a final product having a desired thickness and width. As shown in FIG. 1 , the hot rolling equipment includes an IH device 1 , a rolling line 2 , a calculator 4 and a thermometer 3 . The IH device 1, the computer 4, and the thermometer 3 constitute an IH system according to the first embodiment. For these configurations, see item 1-2. explained in The configuration of the rolling line 2 will be described below.

The rolling line 2 includes a heating furnace 21 , a roughing mill 22 , a finishing mill 23 , a cooling table 24 and a winder 25 . A heating furnace 21 is provided upstream of the rolling line 2 . The heating furnace 21 heats the steel material SS to a predetermined temperature (for example, about 1200° C.). A rectangular parallelepiped steel material SS is extracted from the heating furnace 21 . The steel material SS extracted from the heating furnace 21 has, for example, a thickness of 200-250 mm, a width of 800-2000 mm and a length of 5-12 m.

The rough rolling mill 22 is provided downstream of the heating furnace 21 . The roughing mill 22 has 1 to 3 rolling stands. In the roughing mill 22, forward rolling from the upstream side to the downstream side of the rolling line 2 and reverse rolling from the downstream side to the upstream side are alternately performed multiple times. That is, the steel material SS reciprocates the rough rolling mill 22 multiple times. Thereby, the thickness of the steel material SS is reduced to a predetermined thickness. The rough rolling mill 22 may be attached with a device called an edger for adjusting the width of the steel material SS.

The finishing mill 23 is provided downstream of the IH device 1 . The finishing rolling mill 23 has a plurality of rolling stands (for example, seven). The finishing rolling mill 23 has devices such as rolling rolls and support rolls for each rolling stand. In the finishing mill 23, unidirectional rolling from upstream to downstream of the rolling line 2 is performed. As a result, the thickness of the steel material SS is reduced to the product thickness. During finish rolling, the temperature of the steel material SS decreases due to contact with the rolling rolls and injection of cooling water. The temperature of the steel material SS on the delivery side of the finishing mill 23 is, for example, about 900° C. in the case of ordinary steel.

A cooling table 24 is provided downstream of the finishing mill 23 . The cooling table 24 includes a water cooling device (not shown) called a cooling bank. The cooling table 24 jets cooling water from this cooling bank to lower the temperature of the steel material SS.

A winder 25 is provided downstream of the cooling table 24 . The winding machine 25 winds the steel material SS into a coil. The temperature of the steel material SS before it is wound by the winder 25 is around 600° C. in the case of ordinary steel. In the case of special steel, this temperature may be about 200°C.

1-2. Configuration of IH system 1-2-1. IH Device The IH device 1 is a transverse IH device. FIG. 2 is a schematic diagram showing an example of the configuration of the IH device 1. As shown in FIG. The IH device 1 shown in FIG. 2 includes structural parts 11 t and 11 b, carriages 12 t and 12 b, inductors 13 t and 13 b, a power source 14 and a position controller 15 .

The structural portion 11t accommodates the inductor 13t. The structural portion _11t moves in the width direction WSS within the range of mechanical restrictions. When the structural portion 11t moves, the position of the inductor _13t in the width direction WSS changes. Movement of the structure portion 11t is performed by the position controller 15 changing the position of the carriage _12t in the width direction WSS.

The configuration of the structural portion 11b is the same as that of the structural portion 11t. The configuration of the truck 12b is the same as that of the truck 12t. The configuration of inductor 13b is the same as that of inductor 13t. The configuration of the structural portion 11b is the same as that of the structural portion 11t. The configuration of the truck 12b is the same as that of the truck 12t. Movement of the structural portion 11b is performed by the position controller 15 changing the position of the carriage 12b in the width direction _WSS . The control of the carriages 12t and 12b by the position controller 15 is performed independently.

FIG. 3 is a schematic diagram showing an example of the configuration of inductors 13t and 13b. As shown in FIG. 3, the inductor 13t includes an iron core 16t and a heating coil 17t. The inductor 13b includes an iron core 16b and a heating coil 17b. Alternating current is supplied to the heating coils 17t and 17b from the power supply 14 shown in FIG. Then, a magnetic flux linkage is generated in the thickness direction THSS of the steel material _SS . The interlinking magnetic flux induces an eddy current in the steel material SS. Joule heat is generated by this eddy current, and the steel material SS is heated.

Here, consider the center line L _SS of the steel material SS in the width direction W _SS . In the first embodiment, the position of the center line _LSS is used as a reference position used in various controls. In the following description, the distance xt from the centerline LSS to the axis _L16t of the core _16t and the distance xb from the centerline LSS to the axis _L16b of the core _16b are also referred to as "shift." When the carriages 12t and 12b are controlled by the position controller 15, the distances xt and xb change. Distances xt and xb do not have to match. Further, the distance yt in the thickness direction TH _SS from the steel material SS to the core 16t and the distance yb in the thickness direction TH _SS from the steel material SS to the core 16b may not match.

The IH device 1 is provided with a total of N _IH groups. FIG. 4 is a schematic diagram showing an example of the arrangement of the IH device 1. As shown in FIG. FIG. 4 illustrates an example of arrangement of the inductor 13t. An example of the arrangement of the inductor 13b is basically the same as that of the inductor 13t. Therefore, description of an example of arrangement of the inductor 13b is omitted.

As shown in FIG. 4, the inductors 13t are arranged along the conveying direction DE _SS of the steel material SS. The transport direction DE _SS is a direction orthogonal to the width direction W _SS . The inductor 13t_#1 corresponds to the "head inductor" positioned most upstream in the transport direction DE _SS . An inductor 13t_#2 is positioned downstream of the inductor 13t_#1. The inductor 13t_#N _IH corresponds to the "final inductor" positioned most downstream in the transport direction DE _SS .

A private room (that is, a pulpit) side for operators provided in the rolling line 2 is called an operator side OS. The opposite side of this private room is called the Drive Side DS. In the example shown in FIG. 4, the inductor 13t_#1 is located on the drive side DS. The inductor 13t_#N _IH is located on the operator side OS.

1-2-2. Thermometer A thermometer 3 is provided upstream of the leading inductor. The thermometer 3 measures the actual temperature of the surface of the steel material SS at the node ND_#i as the "entrance-side actual temperature ^Teimeas " (where i satisfies _1≤i≤NND ). FIG. 5 is a schematic diagram showing an example of the node ND. In the example shown in FIG. 5, _NND nodes _{ND_} #i are set along the width direction WSS. Therefore, the data of the entry-side actual temperature Tei ^meas constitutes the information of the temperature distribution ΔTei ^meas of the steel material _SS in the width direction WSS. The thermometer 3 sends information on the temperature distribution ΔTei ^meas to the computer 4 .

1-2-3. Computer Computer 4 is typically a computer comprising a processor, memory and input/output interfaces. The computer 4 performs calculation processing regarding a series of rolling processes until the steel material SS extracted from the heating furnace 21 is wound by the winder 25 . FIG. 6 is a block diagram illustrating an example of the functional configuration of the computer 4. As shown in FIG. As shown in FIG. 6 , the calculator 4 includes a temperature calculator 41 , an induction heating controller (hereinafter also referred to as “IH controller”) 42 , and a database 43 . Some or all of these functions are implemented by executing various programs stored in the memory by the processor of the computer 4 .

The temperature calculator 41 calculates the temperature distribution ΔT of the steel material SS in each step of the rolling process. Rolling command information MIL and actual temperature information TEM are used to calculate the temperature distribution ΔT. The rolling order information MIL includes data on the steel grade Gp, width Wp and thickness THp of the final product. The rolling command information MIL also includes data on the target width W ^tgt and target thickness TH ^tgt of the steel material SS in each process. The rolling command information MIL further includes data of the target temperature FDT of the steel material SS on the delivery side of the finishing mill 23 and data of the target temperature CT of the steel material SS before winding by the winder 25 . The actual temperature information TEM includes information on the temperature distribution ΔTei ^meas .

For example, based on the data of the target temperatures FDT and CT, the temperature calculator 41 sets the target temperature of the node ND_#i (1≦i≦N _ND ) downstream of the final inductor as the “delivery target temperature Tdi ^tgt ”. calculate. The data of the delivery-side target temperature ^Tdi_tgt constitutes information of the temperature distribution ^ΔTdi_tgt of the steel material _SS in the width direction WSS. The temperature calculator 41 may calculate the output side target temperature ^Tdi_tgt based on the data input by the operator. The temperature calculator 41 may calculate the target output temperature Tdi- ^tgt based on the data of the target temperatures FDT and CT and the data input by the operator. The temperature calculator 41 sends information on the temperature distribution ^ΔTdi_tgt to the IH controller 42 .

The temperature calculation unit 41 also calculates the predicted temperature of the steel material SS at the heating start position of the steel material SS by the IH device 1 as the "entrance side predicted temperature Tei ^calf " based on the information of the temperature distribution ΔTei ^meas . The position of the leading inductor is exemplified as the heating start position. The data of the predicted entrance temperature Tei ^calf constitutes information of the temperature distribution ΔTei ^calf of the steel material _SS in the width direction WSS. The temperature calculator 41 sends information on the temperature distribution ΔTei ^calf to the IH controller 42 .

The IH control unit 42 controls the IH device 1 by setting each power supplied to the inductors 13t and 13b and each shift of the inductors 13t and 13b. As a configuration for controlling the IH device 1 , the IH control section 42 includes a data acquisition section 44 and a setting calculation section 45 .

The data acquisition unit 44 acquires information from the temperature calculation unit 41 . The information from the temperature calculator 41 includes information on the temperature distribution ΔTdi ^tgt and information on the temperature distribution ΔTei ^calf . The data acquisition unit 44 also acquires information from the database 43 . Information from database 43 includes model parameters and data tables for various controls in the rolling process. The data acquired by the data acquisition unit 44 constitutes setting calculation information SET. The data acquisition unit 44 sends the setting calculation information SET to the setting calculation unit 45 .

The setting calculation unit 45 performs a “setting calculation process” for calculating the optimum power command value CMp and the optimum shift command value CMx based on the setting calculation information SET. For details of the setting calculation process, see item 2-2. explained in The setting calculator 45 sends the power command value CMp to the power supply 14 . The setting calculator 45 sends the shift command value CMx to the position controller 15 .

2. Features of the first embodiment 2-1. Problem FIG. 7 is a diagram showing an example of the distribution in the width direction WSS of the amount of heat applied from the IH device 1 to the steel material _SS (that is, the amount of temperature increase of the steel material SS). Distribution (I) is an example of the distribution when the distances xt and xb are 0 mm (that is, when the shift is 0 mm). When the positions of the structures 11t and 11b are moved to the drive side DS, the heat quantity distribution changes from distribution (I) to distribution (II). When these positions are moved to the operator side OS, the heat quantity distribution changes from distribution (I) to distribution (III). Thus, changing the shift changes the tendency of the heat quantity distribution. It should be noted that the heat quantity distribution itself changes depending on the design of the inductor and the like.

As a method for uniforming the temperature distribution in the width direction _WSS after heating by the IH device 1, preparation of a data table or the like based on prior examination can be mentioned. In this approach, the respective shifts of the NIH-based _IH device 1 are preset in consideration of the width Wp and thickness THp data of the final product. However, this method cannot take into consideration the change in the temperature distribution in the width direction _WSS at the heating start position. Therefore, if this temperature distribution differs from that in the preliminary examination, there is a problem that it is difficult to make the temperature distribution in the width direction _WSS after heating by the IH device 1 uniform.

Therefore, in the first embodiment, the setting calculation unit 45 performs setting calculation processing. Details of the setting calculation process will be described below.

2-2. Details of setting calculation processing 2-2-1. Flow of Processing FIG. 8 is a flowchart for explaining the flow of setting calculation processing according to the first embodiment. The routine shown in FIG. 8 is repeatedly executed at predetermined intervals by the processor of computer 4 .

In the routine shown in FIG. 8, first, setting calculation information SET is obtained (step S11). The setting calculation information SET is information from the temperature calculation unit 41 and the database 43 described above.

Following step S11, the predicted entry temperature Tei ^calf is calculated (step S12). The entry-side predicted temperature Tei ^calf is calculated, for example, based on the data of the entry-side actual temperature Tei ^meas and the heat release amount from the installation position of the thermometer 3 to the heating start position. This heat release amount is calculated based on model parameters such as the arrival time from the installation position to the heating start position, the thickness and width of the steel material, and the steel grade Gp.

Following step S12, initial values INxj and INpj are set (step S13). The initial value INxj is the initial value of the shift of the inductors 13t and 13b positioned j-th in the transport direction DE _SS (where j satisfies _1≤j≤NIH ). The initial value INpj is the initial value of the power supplied to the j-th inductors 13t and 13b (where j satisfies _1≤j≤NIH ).

FIG. 9 is an example of a data table showing combinations of initial values INxj and INpj. In the example shown in FIG. 9, the combination (x _{mn_k} , p _{mn_k} ) of the initial values INxj and INpj is based on the N _TH divided thickness sections and the N _W divided width sections. (However, m satisfies _1≤m≤NW , n satisfies _1≤n≤NTH , and k satisfies _1≤k≤NIH ). The thickness section and width section are set based on the data of width Wp and thickness THp.

The combination (x _{mn — k} , p _{mn — k} ) referenced by the table is only used as an initial value in the setting calculation process. Therefore, the numbers _NTH and _NW may be small. Moreover, a division based on the data of the steel type Gp, the target temperature FDT, and the target temperature CT may be added to the divisions shown in FIG.

After step S13, a temperature increase amount QTij of the node ND_#i (1≤i≤N _ND ) by the j-th (1≤j≤N _IH ) IH device 1 is calculated (step S14). The temperature increase amount QTij is calculated using a temperature increase amount model. For an example of the temperature rise amount model and an example of calculation of the temperature rise amount QTij, see item 2-2-2. explained in The data on the amount of temperature increase QTij constitutes information on the distribution ΔTij indicating the amount of temperature change on the surface of the steel material SS when the steel material SS is heated by the IH device 1 .

式（１）に示されるｘ_ｊ ^ｍｉｎは、ｊ番目に位置するインダクタのシフトの下限制約であり、ｘ_ｊ ^ｍａｘは当該シフトの上限制約である。ｐ_ｊ ^ｍｉｎはｊ番目に位置するインダクタに供給される電力の下限制約であり、ｐ_ｊ ^ｍａｘは当該電力の上限制約である。Following step S14, the predicted temperature of the steel material SS at the end position of heating of the steel material SS by the IH device 1 is calculated as the "delivery side predicted temperature Tdi ^cal " (step S15). The position of the final inductor is exemplified as the heating end position. Specifically, the output-side predicted temperature ^{Tdi cal} is expressed by the following formula ( 1).

x _j ^min shown in equation (1) is the lower bound constraint for the shift of the j-th inductor, and x _j ^max is the upper bound constraint for that shift. p _j ^min is the lower bound constraint on the power supplied to the j-th inductor, and p _j ^max is the upper bound constraint on that power.

式（２）に示されるｗ_ｊは、ノードＮＤ＿＃ｉ（１≦ｉ≦Ｎ_ＮＤ）に対する重み付け係数である。収束条件は、例えば、評価関数ｆ^ｏｂｊが閾値未満であることを含んでいる。After step S15, it is determined whether or not the convergence condition is satisfied for the output-side predicted temperature Tdi ^cal calculated in step S15 (step S16). Determination of whether or not the convergence condition is satisfied is performed using an evaluation function f ^obj defined by the following equation (2).

w _j shown in equation (2) is a weighting factor for node ND_#i (1≦i≦N _ND ). Convergence conditions include, for example, that the evaluation function f ^obj is less than a threshold.

If the evaluation function f ^obj is equal to or greater than the threshold, a combination (x _{mn_k} , p _{mn_k} ) of the candidate values CAxj and CApj is set (step S17), and steps S14 to S16 are performed. The candidate value CAxj is a candidate value for shifting the j-th inductor. The candidate value CApj is a candidate for power supplied to the j-th inductor. The processes of steps S14 to S17 are repeatedly executed when it is determined that the evaluation function f ^obj is equal to or greater than the threshold. When the processing of steps S14 to S17 is repeated, the combination (x _{mn_k} , p _{mn_k} ) is set so as not to overlap with that set in this routine. A more appropriate combination (x _{mn — k} , p _{mn — k} ) is searched for by repeating the processing of steps S14 to S17.

The convergence condition includes that the number of repetitions of steps S14 to S17 has reached an upper limit. If the evaluation function f ^obj is less than the threshold value, or if the number of repetitions of steps S14 to S17 reaches the upper limit, it is determined that the convergence condition is satisfied.

If it is determined that the convergence condition is satisfied, optimum values OPx and OPp are determined (step S18). The optimum value OPx is the shift command value CMx input to the position controller 15 . The optimum value OPp is the power command value CMp input to the power supply 14 . The optimum values OPx and OPp are determined in consideration of the content of the determination made in step S16. Specifically, when it is determined that the evaluation function f ^obj is less than the threshold, the combination (x _{mn_k} , p _{mn_k} ) used to calculate the evaluation function f ^obj is adopted as the optimum values OPx and OPp. . When it is determined that the number of iterations has reached the upper limit, the combination (x _{mn — k} , p _{mn — k} ) used in the calculation when the evaluation function f ^obj is minimized is adopted as the optimum values OPx and OPp.

2-2-2. Calculation of Temperature Rise QTij Using Temperature Rise Model FIG. 10 is a schematic diagram showing an example of a temperature rise model. In the first embodiment, a machine learning model is used as the temperature increase model. In the example shown in FIG. 10, a machine learning model is constructed by a neural network composed of an input layer IPT, an intermediate layer MID and an output layer OPT. A heating factor for the steel material SS is input to the input layer IPT. Examples of heating factors include steel type Gp, width Wp, thickness THp, delivery side target temperature ^Tdi_tgt , and entry side actual temperature Tei ^meas . The output layer OPT outputs a temperature rise amount QTij at the node ND_#i (1≤i≤N _ND ) by the j-th (1≤j≤N _IH ) IH device 1 .

When constructing a machine learning model, for example, a case study for various heating factors is carried out in advance by numerical analysis of three-dimensional magnetic fields and heat quantities using the finite element method or the like. According to the case studies, the temperature of the surface at the node ND_#i at the heating end position is obtained in each case. A machine learning model is constructed by using this temperature as teacher data.

3. Effect of First Embodiment According to the IH system according to the first embodiment, setting calculation processing is performed to determine the optimum values OPx and OPp. In the setting calculation process, repetitive calculation using the evaluation function f ^obj is performed. The output-side predicted temperature Td ^cal is used as a variable of the evaluation function f ^obj (see formula (2)). A temperature increase amount QTij calculated using a temperature increase amount model is used as a variable for the delivery side predicted temperature Tdi ^cal (see formula (1)). A combination of the candidate values CAxj and CApj is taken into account in the calculation of the temperature increase amount QTij. Therefore, by controlling the IH device 1 based on the optimum values OPx and OPp, it is possible to homogenize the temperature distribution ΔT of the steel material _SS in the width direction WSS at the heating end position.

Also, the predicted entry temperature Tei ^calf is used as the variable of the predicted delivery temperature Tdi ^cal (see equation (1)). The data of the actual entry temperature Tei ^meas is used for the calculation of the predicted entry temperature Tei ^calf . That is, the entry-side actual temperature Tei ^meas is taken into account in the calculation of the entry-side predicted temperature Tei ^calf . Therefore, it is possible to improve the accuracy of prediction of the output-side predicted temperature ^Tdical and improve the reliability of the optimum values OPx and OPp. Therefore, it is possible to improve the reliability of the control of the IH device 1 performed based on the optimum values OPx and OPp.

Second Embodiment Next, a second embodiment of the present invention will be described in detail with reference to FIGS. 11-14. It should be noted that explanations overlapping with the explanation of the first embodiment will be omitted as appropriate.

1. System configuration 1-1. Configuration of IH System FIG. 11 is a schematic diagram showing the configuration of a hot rolling facility to which the IH system according to the second embodiment is applied. In addition, the induction heating method according to the second embodiment is realized by an IH system described below.

As shown in FIG. 11, the hot rolling equipment includes an IH device 1, a rolling line 2, a calculator 4, and a thermometer 5. The IH device 1, the computer 4, and the thermometer 5 constitute an IH system according to the second embodiment.

A thermometer 5 is provided downstream of the final inductor. The thermometer 5 measures the actual temperature of the surface of the steel material SS at the node ND_#i as the "delivery side actual temperature Tdi ^meas " (1≦i≦N _ND ). The concept of node ND_#i is as explained in FIG. The data of the delivery side actual temperature Tdi ^meas constitutes the information of the temperature distribution ΔTdi ^meas of the steel material _SS in the width direction WSS. The thermometer 5 sends information on the temperature distribution ΔT di ^meas to the computer 4 .

1-1-1. Computer FIG. 12 is a block diagram illustrating an example of the functional configuration of the computer 4 . As shown in FIG. 12, the calculator 4 includes a temperature calculator 41, an IH controller 42, a database 43, and a temperature distribution predictor . Some or all of these functions are implemented by executing various programs stored in the memory by the processor of the computer 4 .

The function of the temperature calculator 41 is basically the same as that in the first embodiment. However, in the second embodiment, the temperature calculator 41 calculates the predicted temperature of the steel material SS at the heating start position as the "entrance-side predicted temperature Tei ^cals ". Information other than the information of the temperature distribution ΔTei ^meas is used to calculate the predicted entry temperature Tei ^cals . The information of the temperature distribution ΔTei ^meas is the information used to calculate the predicted entry temperature Tei ^calf in the first embodiment.

This additional information consists of measured or predicted surface temperature data at node ND_#i upstream of the lead inductor, such as data for the actual surface temperature at node ND_#i on the delivery side of the roughing mill 22. The information to be provided is exemplified. The data of the predicted entrance temperature Tei cals constitutes information of the temperature distribution ^{ΔTei cals} of the steel material _SS in the width direction WSS. The temperature calculation unit 41 sends information on the temperature distribution ^{ΔTei cals} to the temperature distribution prediction unit 46 .

The function of the IH control section 42 is also basically the same as that in the first embodiment. That is, the IH control unit 42 (setting calculation unit 45) performs setting calculation processing based on the setting calculation information SET. However, in the second embodiment, the IH control unit 42 (setting calculation unit 45) uses the information of the temperature distribution ΔTdi ^cal used as the variable of the evaluation function ^obj when the optimum values OPx and OPp are adopted as the temperature distribution prediction. 46.

The temperature distribution prediction unit 46 calculates the temperature distribution ΔT of the steel material SS in the width direction W _SS at the heating start position as "temperature distribution ΔTei ^recal ". The temperature distribution ΔTei ^recal is composed of data of the re-predicted temperature of the steel material SS in the width direction W _SS at the heating start position (hereinafter also referred to as "entrance-side re-predicted temperature Tei ^recal "). The entry-side re-predicted temperature Tei ^recal is specifically calculated based on a calculation formula (formula (3) below) representing the sum of the data of the entry-side predicted temperature Tei ^cals and the deviation T ^error .

式（３）に示されるαは、前回材ＳＳＡが今回材ＳＳＢに与える影響を表す係数である。式（４）に示される出側実温度Ｔｄｉ^ｍｅａｓのデータは、温度計算部４１から入力された前回材ＳＳＡの温度分布ΔＴｄｉ^ｍｅａｓの情報に含まれている。出側予測温度Ｔｄｉ^ｃａｌのデータは、設定計算部４５から入力された前回材ＳＳＡの温度分布ΔＴｄｉ^ｃａｌの情報に含まれている。The deviation T ^error is the data of the predicted delivery temperature Tdi ^cal of another steel material SS heated by the IH apparatus 1 immediately before the steel material SS is heated by the IH apparatus 1 and the actual delivery side temperature of the other steel material SS. ^Tdi_meas data. For convenience of explanation, the steel material SS located upstream of the heating start position and to be heated by the IH device 1 will be referred to as "current material SSB". The steel material SS located downstream of the heating end position and heated by the IH device 1 immediately before the current material SSB is referred to as "previous material SSA".

α shown in Equation (3) is a coefficient representing the influence of the previous material SSA on the current material SSB. The data of the delivery side actual temperature Tdi ^meas shown in Equation (4) is included in the information of the temperature distribution ΔTdi ^meas of the previous material SSA input from the temperature calculation unit 41 . The data of the delivery-side predicted temperature Tdi ^cal is included in the information of the temperature distribution ΔTdi ^cal of the previous material SSA input from the setting calculator 45 .

2. Features of the second embodiment 2-1. Problem FIG. 13 is a schematic diagram showing an example of the relationship between the temperature distributions ΔTdi ^cal and ΔTdi ^meas of the previous material SSA and the temperature distributions Tei ^cals and ΔTei ^recal of the current material SSB. In the IH system according to the first embodiment, information on the temperature distribution ΔTei ^meas is obtained from the thermometer 3 . In contrast, the IH system according to the second embodiment is not provided with the thermometer 3 . Therefore, when these systems are compared, it is assumed that the accuracy of prediction of the temperature distribution ^{ΔTei cals} is low in the second embodiment.

Therefore, in the second embodiment, prediction information of the temperature distribution of the previous material SSA downstream of the final inductor (that is, information of the temperature distribution ΔTdi ^cal of the previous material SSA) and result information (that is, temperature distribution ΔTdi of the previous material SSA The temperature distribution ^{ΔTei cals} is predicted “again” using the error from the information of ^meas ). This compensates for the decrease in prediction accuracy of the temperature distribution ^{ΔTei cals} .

2-2. Details of Setting Calculation Processing FIG. 14 is a flowchart for explaining the flow of setting calculation processing according to the second embodiment. The routine shown in FIG. 14 is executed after the heating of the previous material SSA by the IH device 1 is finished and before the heating of the current material SSB is started. The processing of steps S11 and S13 to S18 shown in FIG. 14 is as explained in FIG. Therefore, the details of the processing of steps other than these steps will be described below.

Following step S11, the predicted entry temperature Tei ^cals is calculated (step S21). The predicted entry temperature Tei ^cals is calculated, for example, based on the data constituting the "other information" described above. Consider the case where the data constituting the “other information” is the data of the actual temperature of the surface at the node ND_#i on the delivery side of the roughing mill 22 . In this case, the entry-side predicted temperature Tei ^cals is calculated.

Following step S21, the deviation T ^error is calculated (step S22). The deviation T ^error is calculated by inputting the data of the predicted delivery temperature Tdi ^cal of the previous material SSA and the data of the actual delivery temperature Tdi ^meas of the previous material SSA into the equation (4).

After step S22, the entry-side re-predicted temperature Tei ^recal is calculated (step S23). The entry-side re-predicted temperature Tei ^recal is calculated by inputting the deviation T ^error calculated in step S22 and the entry-side predicted temperature Tei ^cals of the current material SSB into equation (3).

3. Effects of the Second Embodiment According to the IH system according to the second embodiment, even if a thermometer (that is, the thermometer 3 shown in FIG. 1) is not provided upstream of the leading inductor, Calculation of the entry-side re-predicted temperature Tei ^recal performed using data different from the data of the actual temperature Tei ^meas compensates for the decrease in prediction accuracy of the entry-side predicted temperature Tei ^cals . Therefore, according to the IH system according to the second embodiment, it is possible to obtain the same effect as that according to the first embodiment.

1 induction heating device 13t, 13b inductor 14 power supply 15 position controller 2 rolling line 3, 5 thermometer 4 calculator 41 temperature calculation section 42 induction heating control section 43 database 44 data acquisition section 45 setting calculation section 46 temperature distribution prediction section DE _SS transfer Direction SS Steel material SSA Previous material SSB Current material Tdi ^meas output side actual temperature Tdi ^cal output side predicted temperature Tdi ^tgt output side target temperature Tei ^meas entry side actual temperature Tei ^calf , ^{Te cals entry side} predicted temperature , ΔTdi ^meas , ΔTei ^meas , ΔTdi ^cal , ΔTdi ^tgt , ΔTei ^calf , ΔTei ^cals , ΔTei ^recal Temperature distribution TH _SS thickness direction W _SS width direction

Claims (6)

A transverse induction heating method,
setting a combination of electric power supplied to each of a plurality of inductors provided in the conveying direction of the rolling line and heating the steel material and positions of each of the plurality of inductors in the width direction of the steel material;
The step of setting the combination comprises:
calculating an entry-side predicted temperature indicating a predicted temperature of each of the plurality of sections of the steel material in the width direction at the position where the steel material starts to be heated by the plurality of inductors;
setting candidates for the combination;
Predicting, using a machine learning model, the amount of temperature rise in each of the plurality of categories when the candidate is adopted;
calculating, based on the predicted entry-side temperature and the amount of temperature increase, a predicted exit-side temperature indicating the predicted temperature of each of the plurality of sections at the end position of heating of the steel material by the plurality of inductors;
The candidate or the evaluation when an evaluation function having variables of the predicted delivery temperature and target temperatures of the plurality of sections at the heating end position as variables is less than a threshold, or the evaluation A step of adopting the candidate when the function is the minimum as the combination;
An induction heating method comprising: The induction heating method according to claim 1,
In the step of calculating the predicted entry temperature, the predicted entry temperature is calculated based on the actual entry temperature indicating the actual temperature of each section of the steel material upstream from the heating start position. induction heating method. The induction heating method according to claim 1,
The steel material includes a current material located upstream of the heating start position, and a previous material located downstream of the heating end position and heated by the inductors immediately before the current material is heated by the plurality of inductors. , including
The step of setting the combination includes calculating a deviation between the predicted delivery-side temperature of the previous material and the actual delivery-side temperature indicating the actual temperature of each section of the previous material downstream of the heating end position. further includes
In the step of calculating the predicted entry-side temperature, the predicted entry-side temperature is calculated based on the temperature of each of the plurality of sections of the current material upstream from the heating start position and the deviation. An induction heating method characterized by: A transverse induction heating system,
a plurality of inductors provided in the conveying direction of the rolling line to heat the steel material;
A control device for controlling the temperatures of the plurality of sections of the steel material in the width direction based on the power supplied to each of the plurality of inductors and the position of each of the plurality of inductors in the width direction of the steel material. and,
with
The control device performs setting calculation processing for setting a combination of the power and the position,
The control device, in the setting calculation process,
calculating an entry-side predicted temperature indicating the predicted temperature of each of the plurality of sections at the position where heating of the steel material by the plurality of inductors is started;
setting candidates for the combination;
Using a machine learning model, predict the amount of temperature rise in each of the plurality of categories when the candidate is adopted,
calculating a predicted outlet temperature indicating a predicted temperature of each of the plurality of sections at the end position of heating of the steel material by the plurality of inductors, based on the predicted inlet temperature and the amount of temperature increase;
The candidate or the evaluation when an evaluation function having variables of the predicted delivery temperature and target temperatures of the plurality of sections at the heating end position as variables is less than a threshold, or the evaluation An induction heating system, wherein the candidates for which the function is minimized are adopted as the combination. An induction heating system according to claim 4, wherein
An entry-side thermometer is provided upstream from the heating start position and measures an entry-side actual temperature indicating the actual temperature of each section of the steel material,
The induction heating system, wherein the control device calculates the predicted entry-side temperature based on the actual entry-side temperature in the setting calculation process. An induction heating system according to claim 4, wherein
further comprising a delivery-side thermometer that is provided downstream from the heating end position and measures an actual delivery-side temperature that indicates the actual temperature of each section of the steel material,
The steel material includes a current material located upstream of the heating start position, and a previous material located downstream of the heating end position and heated by the inductors immediately before the current material is heated by the plurality of inductors. , including
The control device, in the setting calculation process,
calculating a deviation between the predicted output-side temperature of the previous material and the actual output-side temperature of the previous material;
The induction heating system, wherein the predicted entry temperature is calculated based on the temperature of each of the plurality of sections of the current material upstream from the heating start position and the deviation.

Images