JP2019107660A - Calculation method of rough rolling time of steel plate, calculation device of the same, and manufacturing method of steel plate - Google Patents

Abstract

To provide a calculation method of rough rolling time of a steel plate and a calculation device of the same in which rough rolling time of a steel plate can be accurately calculated regardless of variation of facility state or operation state.SOLUTION: A calculation device 10 of rough rolling time of a steel plate comprises: a database 11 storing data regarding specifications of steel plates rolled in the past and number of rough rolling passes thereof; and a DBM server 12 which constructs a regression equation model representing a relationship between the specifications of the steel plate and the number of rough rolling passes by use of specifications of steel plates rolled in the past and number of rough rolling passes thereof stored in the database 11, calculates number of rough rolling passes of a steel plate of a processing object by entering data of specification of the steel plate of the processing object into the regression equation model, calculates time required for rough rolling of the steel plate of the processing object as a pure rolling time on the basis of the calculated number of rough rolling passes, and calculates rough rolling time of the steel plate of the processing object by adding the pure rolling time, and turn time and width read-in time of the steel plate of the processing object.SELECTED DRAWING: Figure 2

Description

  The present invention calculates a rough rolling time calculation method of thick steel plate for calculating the rough rolling time of thick steel plate in a rolling line of thick steel plate having reverse rough rolling mill and finish rolling machine, rough rolling time calculation device of thick steel plate, The present invention relates to a method of manufacturing a thick steel plate.

  The rolling line of thick steel plate with reverse roughing mill and finishing rolling machine has the feature that production efficiency is high compared with the rolling line of thick steel plate which performs rough rolling to finish rolling with one rolling mill Have. In a rolling line of a thick steel plate having a reverse rough rolling mill and a finishing rolling mill, first, a slab which is a steel material is heated by a heating furnace and then roughly rolled by the reverse rough rolling mill. In rough rolling, width rolling is also carried out, in which the slab is rolled 90 ° in a horizontal plane. Then, the slab after rough rolling is finish-rolled to a product thickness in a finish-rolling mill. In order to carry out finish rolling in a specific temperature range, a cooling process or holding process may be carried out between the reverse rough rolling mill and the finish rolling mill for temperature control of the slab. In addition, accelerated cooling may be performed on a thick steel plate obtained by finish rolling, as necessary.

  In such a thick steel plate rolling line, a continuous heating furnace is widely used as a heating furnace for heating a slab. A continuous heating furnace is a furnace which heats while conveying a plurality of slabs, and the slab at the head of the furnace is extracted in order outside the furnace. In order to heat the slab efficiently in the continuous heating furnace, the in-furnace time (heating time) of the slab may be determined, and the slab may be heated so that the temperature of the slab is just the target temperature in the in-furnace time. If the minimum time required for heating to the target temperature is shorter than the rolling turn waiting time, the slab in-furnace time becomes the rolling turn waiting time. For this reason, the standing time of the slabs charged to the continuous heating furnace is the sum of the rolling times of the slabs already charged. For this reason, in a rolling line of thick steel plate having a reverse rough rolling mill and a finishing rolling mill, the steps from the continuous heating furnace to the reverse rough rolling mill are steps in which all the slabs pass, so Accurately calculating the time required for rough rolling (rough rolling time) to determine the furnace time is extremely important in improving the heating efficiency of the slab (fuel basis unit) and consequently the yield of thick steel plate is there.

  From such a background, Patent Document 1 describes an invention for calculating the total rolling time of a thick steel plate in consideration of the difference in product quality of the thick steel plate.

JP, 2014-28394, A

  However, the invention described in Patent Document 1 does not consider the change in rolling time due to the change in threading speed or rolling load caused by the change in equipment condition or operation condition, so the equipment condition or operation condition changes. If so, the rolling time can not be calculated accurately. In addition, in order to solve such a problem, the method of calculating | requiring the correction coefficient of rolling time beforehand for every change of an installation state or an operation condition, and correcting rolling time using this correction coefficient is also considered. However, with this method, much labor is required, and the rolling time can not be accurately calculated if a change other than the expected change occurs.

  This invention is made in view of the said subject, Comprising: The objective is rough-rolling of the thick steel plate which can calculate the rough-rolling time of a thick steel plate accurately, even when an installation state or an operating condition changes. It is providing a time calculation method and a rough rolling time calculation device for thick steel plates. Another object of the present invention is to provide a method of manufacturing a thick steel plate capable of manufacturing a thick steel plate with high yield.

  The method of calculating the rough rolling time of thick steel plate according to the present invention is a method of calculating the rough rolling time of thick steel plate which calculates the rough rolling time of thick steel plate in the rolling line of thick steel plate having reverse roughing mill and finish rolling machine. Build a regression model that shows the relationship between the specifications of thick steel plate and the number of rough rolling passes, using data of the specifications of the thick steel plate rolled in the past and the number of rough rolling passes stored in the database Calculating the number of rough rolling passes of the thick steel plate to be processed by inputting the data of the specifications of the thick steel plate to be processed into the regression model, and calculating the number of rough rolling passes based on the calculated number of rough rolling passes Calculating the time required for the rough rolling of the thick steel plate as a pure rolling time, and the sum of the pure rolling time and the turn time and width reading time of the thick steel plate to be treated as the rough rolling time of the thick steel plate to be treated Calculation Characterized in that it comprises the steps of, a.

  In the rough rolling time calculation method of a thick steel plate according to the present invention, in the above-described invention, the specifications of the thick steel plate to be treated and the data of the number of rough rolling passes are Storing the data as data in the database.

  The rough rolling time calculation device for thick steel plate according to the present invention is a rough rolling time calculation device for thick steel plate which calculates rough rolling time of thick steel plate in a rolling line of thick steel plate having reverse rough rolling mill and finish rolling mill. There is a database storing data of the specifications of the thick steel plate rolled in the past and the number of rough rolling passes, and data of the specifications of the thick steel plate rolled in the past stored in the database and the number of rough rolling passes Is used to construct a regression model that indicates the relationship between the specifications of the thick steel plate and the number of rough rolling passes, and the data of the specifications of the thick steel plate to be processed is input to the regression model. The number of rough rolling passes of the steel plate is calculated, and the time required for the rough rolling of the thick steel plate to be treated is calculated as a pure rolling time based on the calculated number of rough rolling passes, and the pure rolling time and the turn of the thick steel plate to be treated Time and width And database modeling server that calculates a rough rolling time thick steel sum processed with time narrowing seen, characterized in that it comprises a.

  A method of manufacturing a thick steel plate according to the present invention controls a heating furnace for heating a slab to be rolled in the rolling line based on a rough rolling time calculated by the method of calculating a rough rolling time of a thick steel plate according to the present invention. It is characterized by including the step of manufacturing a steel plate.

  According to the rough rolling time calculation method of a thick steel plate and the rough rolling time calculation device of a thick steel plate according to the present invention, the rough rolling time can be accurately calculated even when the equipment state or the operation condition changes. Moreover, according to the method of manufacturing a thick steel plate according to the present invention, a thick steel plate can be manufactured with a high yield.

FIG. 1 is a schematic view showing a configuration of a rolling line to which a rough rolling time calculation device for thick steel plate which is an embodiment of the present invention is applied. FIG. 2: is a schematic diagram which shows the structure of the rough rolling time calculation apparatus of the thick steel plate which is one Embodiment of this invention. FIG. 3 is a flowchart showing the flow of rough rolling time calculation processing according to an embodiment of the present invention. FIG. 4 is a scatter diagram showing the relationship between the rough rolling time calculated by the prior art and the present invention and the actual value of the rough rolling time.

  Hereinafter, with reference to the drawings, the configuration and operation of a rough rolling time calculation device for thick steel plate which is an embodiment of the present invention will be described.

〔line formation〕
First, with reference to FIG. 1, the structure of the rolling line to which the rough rolling time calculation apparatus of the thick steel plate which is one Embodiment of this invention is applied is demonstrated.

  FIG. 1 is a schematic view showing a configuration of a rolling line to which a rough rolling time calculation device for thick steel plate which is an embodiment of the present invention is applied. As shown in FIG. 1, a rolling line 1 to which a rough rolling time calculation device for thick steel plate according to an embodiment of the present invention is applied includes a heating furnace 2, a reverse rough rolling mill 3, a finishing rolling mill 4, and acceleration. A cooling device 5 is provided as a main component. In the rolling line 1, first, the slab S which is a steel material is heated by the heating furnace 2 which is a continuous heating furnace and then roughly rolled by the reverse rough rolling mill 3. Then, the slab S after rough rolling is finish-rolled to a product thickness in a finish-rolling mill 4 and then accelerated and cooled by an accelerated cooling device 4 as necessary. Thereby, the thick steel plate P is manufactured.

〔Device configuration〕
Next, with reference to FIG. 2, the structure of the rough rolling time calculation apparatus of the thick steel plate which is one Embodiment of this invention is demonstrated.

  FIG. 2: is a schematic diagram which shows the structure of the rough rolling time calculation apparatus of the thick steel plate which is one Embodiment of this invention. As shown in FIG. 2, the rough rolling time calculation device 10 for a thick steel plate according to an embodiment of the present invention includes a database 11 and a DBM (DataBase Modeling) server 12 as main components.

  The data base 11 includes the specifications (weight, heating furnace extraction temperature, thickness, intermediate thickness (thickness after rough rolling) of slabs S (hereinafter referred to as past materials) rolled in the rolling line 1 shown in FIG. 1 ), Target finish temperature, finish width, finish length, finish thickness etc., actual values of rough rolling pass number (hereinafter referred to as actual rough rolling pass number), and turn time (slab S in width direction and length direction) The data of the actual value (hereinafter referred to as actual turn time) of the time required to turn for rolling is stored.

  The DBM server 12 is configured by an information processing apparatus such as a workstation and is electrically connected to the database 11 so as to be able to communicate information with the database 11. The DBM server 12 executes various processes such as a rough rolling time calculation process described later when an arithmetic processing unit such as a central processing unit (CPU) in the information processing apparatus executes a computer program.

  In the rough rolling time calculation device 10 for a thick steel plate having such a configuration, the DBM server 12 executes the rough rolling time calculation processing described below, and the slab is obtained even if the equipment state or the operating condition changes. The rough rolling time of S is accurately calculated. The operation of the DBM server 12 when executing the rough rolling time calculation process will be described below with reference to the flowchart shown in FIG.

[Rough rolling time calculation process]
FIG. 3 is a flow chart showing a flow of rough rolling time calculation processing according to an embodiment of the present invention. The flowchart shown in FIG. 3 starts at the timing when information on the specifications of the slab S (hereinafter referred to as the material) to be processed is input to the DBM server 12, and the rough rolling time calculation process is the process of step S1. Go to

  In the process of step S1, the DBM server 12 reads data of the specifications of past material stored in the database 11, the number of actual rough rolling passes, and the actual turn time. Thus, the process of step S1 is completed, and the rough rolling time calculation process proceeds to the process of step S2.

  In the process of step S2, the DBM server 12 constructs a regression model (prediction model) indicating the relationship between the specifications of the past material and the number of rough rolling passes using the data read in the process of step S1. Specifically, a regression model of the number of rough rolling passes constructed based on the specifications of the slab S is generally non-linear, and a large amount of calculation is required to construct a non-linear regression model. So, in this embodiment, reduction and speeding up of calculation amount are achieved focusing on the fact that a continuous non-linear function can be linearly approximated within a minute section. That is, a regression model is constructed using only data in the specification space in the vicinity of the specification of the material.

Specifically, the number of past materials whose data is stored in the database 11 is n, the type of specifications is m types, and the specifications of the i (= 1 to n) -th past materials are m × 1 matrix X Let _i , X = (X ₁ ,..., X _n ), and let the number of actual rough rolling passes of past material be 1 × n matrix Y. At this time, the linear regression equation shown in the following equation (1) can be derived by solving the minimum problem shown in the following equation (2) and finding the values of the parameters α and β. Where α is a 1 × m coefficient matrix and β is a 1 × n constant matrix.

Specifically, the number of past materials whose data is stored in the database 11 is n, the type of specifications is m types, and the specifications of the i (= 1 to n) -th past materials are m × 1 matrix X Let _i , X = (X ₁ ,..., X _n ), and let the number of actual rough rolling passes of past material be 1 × n matrix Y. Based on the past material data, it is considered to optimize the number of rough rolling passes of a material having certain specifications. In order to obtain the optimum value of the rough rolling pass number y (y is a scalar value) for a material having certain parameters x (x is an m × 1 matrix) by the linear regression equation shown in the following equation (1), It is sufficient to solve the minimum problem shown by the equation (2) of and calculate the values of the parameters α and β. Is a 1 × m coefficient matrix, and β is a 1 × n constant matrix in which the values of all the elements thereof are scalar values β ^* .

Next, assuming that the specification matrix of the material is A (= (a ₁ ,..., A _m )) and the distance from the past material to the material in the specification space is L, the distance L is It becomes n × 1 matrix shown in).

Next, a weighting matrix W is defined by the following equation (4) for the distance L shown by the equation (3). However, shown in Equation (4), _{L i} is the matrix L i (= 1~n) th row, sigma (L) is the standard deviation of the matrix L, p and constant.

  The weighting matrix W is a Gaussian function centered on the specifications of the material. Therefore, by weighting with the weighting matrix W, linear approximation in a minute space around the specifications of the material can be performed. That is, (Y-αX-β) in Formula (2) is a 1 × n matrix, and if (Y-αX-β) is K, the minimum problem shown in Formula (2) is It is expressed as 5).

  As described above, by assigning weights to the minimization objective function of linear regression, it is possible to construct a regression model in which the number of rough rolling passes which is a non-linear function is approximated around the required point. This regression equation model can all be constructed by automatic calculation by an arithmetic processing unit. Thus, the process of step S2 is completed, and the rough rolling time calculation process proceeds to the process of step S3.

  In the process of step S3, the DBM server 12 predicts the rough rolling time of the material by inputting data of specifications of the material into the regression model constructed in the process of step S2. Thus, the process of step S3 is completed, and the rough rolling time calculation process proceeds to the process of step S4.

  In the process of step S4, a rolling schedule for the DBM server 12 to roughly roll the slab S to a predetermined thickness with the number of rough rolling passes calculated in the process of step S3 using data of specifications of the material (Conveying speed and rolling load of slab S in each rough rolling pass) are calculated. Then, the DBM server 12 calculates the rough rolling time of the material as a pure rolling time based on the calculated rolling schedule, and the calculated pure rolling time and the turn time and width reading time of the material (slab S is predetermined The time with which it is measured whether or not it can be rolled to the dimension width of 2) is calculated as the rough rolling time of this material. In the present embodiment, the average value of the actual turn times stored in the database 11 is used as the turn time of the material. Further, since the width reading is performed fully automatically, the width reading time is considered to be constant regardless of the slab S, and a predetermined time is set as the width reading time. Thus, the process of step S4 is completed, and the rough rolling time calculation process proceeds to the process of step S5.

  In the process of step S5, the DBM server 12 stores data of the material of the material, the number of rough rolling passes, and the actual value of the turn time in the database 11. When the number of data points in the database 11 is equal to or more than a predetermined number, the DBM server 12 desirably deletes the oldest data. Thereby, the process of step S5 is completed and a series of rough rolling time calculation processes are complete | finished.

  As apparent from the above description, the DBM server 12 uses the data of the past material specifications and the number of rough rolling passes stored in the database 11 to determine the relationship between the specifications of the past material and the number of rough rolling passes. The regression model is constructed and the data of the specification of the material is input to the regression model to calculate the number of rough rolling passes of the material, and the rough of the material is calculated based on the calculated number of rough rolling passes. The time required for rolling is calculated as the pure rolling time, and the sum of the pure rolling time and the turn time and width reading time of the material is calculated as the rough rolling time of the material. Even if there is, the rough rolling time of the material can be calculated with high accuracy.

  Also, by manufacturing the thick steel plate P while controlling the heating furnace 2 that heats the slab S rolled in the rolling line 1 based on the rough rolling time calculated by the rough rolling time calculation process, the thick steel plate P is obtained with high yield. It can be manufactured.

  In order to confirm the effect of this invention, rough rolling time was estimated using the prior art and this invention about the rolling data of 10978 thick steel plates. 4 (a) and 4 (b) show the actual values (rough results rough rolling time) of rough rolling time predicted by the prior art and the present invention (prediction rough rolling time prediction and rough rolling time prediction by DBM) and rough rolling time, respectively. When there is a point on the straight line in the figure, it means that the predicted value and the actual value match. As shown in FIGS. 4A and 4B, the determination coefficients of the prior art and the present invention are 0.03 and 0.22, respectively, and the standard deviations of the errors of the prior art and the present invention are 38.8 seconds and It became 30.8 seconds (average value of actual rough rolling time is 117.9 seconds). From this, it was confirmed that according to the present invention, the rough rolling time can be calculated with high accuracy.

  The embodiment to which the invention made by the present inventors has been applied has been described above, but the present invention is not limited by the description and the drawings that form a part of the disclosure of the present invention according to this embodiment. That is, other embodiments, examples, operation techniques and the like made by those skilled in the art based on the present embodiment are all included in the scope of the present invention.

DESCRIPTION OF SYMBOLS 1 rolling line 2 heating furnace 3 reverse rough rolling mill 4 finish rolling mill 5 accelerated cooling device 10 coarse steel rolling time calculating device 11 database 12 DBM (DataBase Modeling) server S slab P thick steel plate

Claims (4)

A rough rolling time calculation method of thick steel plate for calculating rough rolling time of thick steel plate in a rolling line of thick steel plate having reverse rough rolling mill and finish rolling mill,
Constructing a regression model indicating the relationship between the specifications of the thick steel plate and the number of rough rolling passes using the specifications of the thick steel plate rolled in the past and the data of the number of rough rolling passes stored in the database; ,
The number of rough rolling passes of the thick steel plate to be treated is calculated by inputting data of specifications of the thick steel plate to be treated into the regression model, and the thick steel plate to be treated is calculated based on the calculated number of rough rolling passes. Calculating a time required for rough rolling as a pure rolling time;
Calculating the sum of the pure rolling time and the turn time and width reading time of the thick steel plate to be treated as the rough rolling time of the thick steel plate to be treated;
The rough rolling time calculation method of the thick steel plate characterized by including.   Storing the specification of the thick steel plate to be processed and the data of the number of rough rolling passes in the database as the data of the thick steel plate rolled in the past and the number of rough rolling passes; The rough rolling time calculation method of the thick steel plate as described in. A rough rolling time calculation device for thick steel plate which calculates rough rolling time of thick steel plate in a rolling line of thick steel plate having reverse rough rolling mill and finishing rolling machine,
A database for storing data of the specifications of the steel plate rolled in the past and the number of rough rolling passes;
By using the data of the specification of the thick steel plate and the number of rough rolling passes that were rolled in the past and stored in the database, a regression model indicating the relationship between the specification of the thick steel and the number of rough rolling passes is constructed. The number of rough rolling passes of the thick steel plate to be treated is calculated by inputting data of specifications of the thick steel plate to be treated into the regression model, and the thick steel plate to be treated is calculated based on the calculated number of rough rolling passes. A database modeling server which calculates the time required for rough rolling as a pure rolling time and calculates the sum of the pure rolling time and the turn time and width reading time of the thick steel plate to be treated as the rough rolling time of the thick steel plate to be treated;
An apparatus for calculating a rough rolling time of a thick steel plate, comprising:   The step which manufactures a thick steel plate, controlling the heating furnace which heats the slab rolled by the said rolling line based on the rough rolling time calculated by the rough rolling time calculation method of the thick steel plate of Claim 1 or 2 A method of manufacturing a thick steel plate characterized by

Images