JP2000202512A - Automatically extracting method for heating furnace in hot rolling - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make not only the average value of the error of prediction error of rolling time but variance small and to optimize the extracting timing of a material from a heating furnace. SOLUTION: A prediction equation of a passing time through each equipment for executing prediction calculation of the time from the extracting point of time of the material till passing through each equipment is described as a function of the number of passes and rolling length, the prediction equation is corrected using an online identifying method based on the actual value of the passing time through each equipment of a preceding material, the extracting time from the heating furnace of the next material is decided and the material is automatically extracted from the heating furnace.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for automatically extracting a heating furnace in hot rolling.
Predict and calculate the time from material extraction to passing through each facility, determine the material extraction time so that there is no interference between the preceding and succeeding materials, and automatically extract the material from the heating furnace On how to do it.

[0002]

2. Description of the Related Art In hot rolling, it is important to appropriately determine the timing of extracting a material from a heating furnace from the viewpoint of rolling efficiency and energy efficiency. If the extraction timing is too early, interference with the preceding material occurs in the rate-determining process, causing problems such as material collision, damage to the rolling machine, energy loss due to waiting time on the mill line, and conversely, the extraction timing is too late. If this is the case, waste time for transporting the material occurs, and the rolling efficiency decreases.

[0003] Generally, as described in Japanese Patent Application Laid-Open No. 61-259818, a method of predicting and calculating the passage time of each facility based on a rolling schedule is used to determine the timing of extracting a material from a heating furnace. ing.

If the actual value of the rolling time differs from the predicted calculated value and an error occurs, as described in Japanese Patent Application Laid-Open No. Sho 62-30812, the actual value of the preceding material at each facility passing time is used. There is a method of correcting the extraction time of the next extraction material based on the difference from the predicted value. Thereby, it is possible to automatically optimize the extraction timing with respect to a sudden failure of the equipment or a change in the rolling time due to manual intervention of an operator.

[0005]

However, in order to improve the biting property and the quality (shape and size), the rolling speed and
If the acceleration / deceleration rate, the number of passes, and the like are adjusted, the rolling time prediction formulas created so far do not match, and there is a problem that an error occurs between the rolling time prediction value and the actual value.

If the extraction timing of the next material is corrected based on the difference between the actual time and the predicted time as in the method of the conventional Japanese Patent Application Laid-Open No. Sho 62-30812, the average value of the rolling time prediction error generated due to such factors is considered. However, there is a problem that the variance showing the variation cannot be reduced.

[0007] The magnitude of the prediction error is generally evaluated by a root mean square error.

[0008] Assuming that N data can be collected,

[0009]

[Outside 1]

The prediction error ei is defined by the following equation.

[0011]

(Equation 1)

Then, the root mean square error MSE is defined by the following equation.

[0013]

(Equation 2)

The average value of the prediction error is

[0015]

[Outside 2] Then, the root mean square error MSE can be transformed as the following equation.

[0016]

(Equation 3)

The first term on the right side of equation (4) represents the variance (variation from the average value), and the second term represents the square of the average value. That is, the root mean square error representing the magnitude of the prediction error is
It is represented by the sum of the square of the variance and the average value. Therefore, in order to reduce the prediction error, it is necessary to make the average value of the error close to 0 and reduce the variance.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned conventional problems, and it is an object of the present invention to reduce not only the average value of the rolling time prediction error but also the variance, and to optimize the extraction timing from the heating furnace. And

[0019]

According to the present invention, in order to solve the above-mentioned problems, each equipment passage time prediction formula for predicting and calculating the time from the time of material extraction to each equipment passage is calculated by using the number of passes and pressure. Describe the extension function, modify the prediction formula using the online identification method based on the actual passing time of each equipment of the preceding material, determine the heating furnace extraction time of the next material, and automatically output the material from the heating furnace. Is extracted.

Each equipment passing time represented by the rolling time is as follows:
There is a correlation between the number of passes and the elongation of pressure, and this characteristic does not change even if the rolling speed or the acceleration / deceleration rate fluctuates. Therefore, by performing multiple regression on the actual value of each equipment transit time with the number of passes and pressure extension, and sequentially correcting each equipment transit time prediction formula, the average of each equipment transit time prediction error caused by fluctuations in rolling speed and acceleration / deceleration rate is obtained. By reducing both the value and the variance, the heating furnace extraction timing can be further optimized.

[0021]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, with reference to the drawings, an embodiment in which the present invention is applied to section steel rolling will be described in detail.

As shown in FIG. 1, the section steel rolling line comprises a heating furnace 10, a breakdown mill 12, a rough mill 14 including a universal mill U1 and an edger E1, and a finishing mill (UF) 16, Of these, the roughing mill 14 is often the rate-determining step.

In the present embodiment, the roughing mill 14 is a rate-determining step, and the preceding and succeeding materials do not interfere with each other in the roughing mill 14,
In addition, a method for determining an optimum heating furnace extraction timing so as to minimize the time from when the preceding material passes through the coarse mill 14 to when the following material enters the coarse mill 14 will be described. here,
For the sake of simplicity, it is assumed that the rolling is always performed at the same time in the breakdown mill 12. That is, it is assumed that the time from when the material is extracted from the heating furnace 10 to when it reaches the roughing mill 14 is always constant.

In the present embodiment, the following equation is used as a model for predicting the rolling time in the rough mill 14.

Rolling time = a × number of passes + b × pressure extension + c (5)

Here, the number of passes means the number of times of rolling in the coarse mill 14, and the elongation means the material length after completion of the rough mill rolling, a, b, c.
Is a para router of the rolling time prediction model. These parameters a, b, and c are classified according to the web height and the flange width of the section steel as the rolled product, and are determined off-line from the actual rolling time using an identification method such as a least square method.

The time required for extracting the preceding material from the heating furnace 10 and extracting the next material from the heating furnace 10 (hereinafter referred to as the extraction pitch) is determined as follows.

Extraction pitch = Rough mill rolling time of preceding material + Bar to Bar time (6)

Here, the rough mill rolling time of the preceding material is the time from when the preceding material enters the rough mill 14 to when it leaves the rough mill 14. This is calculated using equation (5) based on the planned path and the planned pressure extension of the preceding material. or,
The Bar to Bar time is the time from when the preceding material exits the coarse mill 14 to when the next material can enter the coarse mill 14 (for example, when the preceding material exits the coarse mill and Time to set the equipment to the next material rolling position)
Give as a constant.

Based on the extracted pitch calculated by the equation (6),
If the material is extracted from the heating furnace 10, the optimum extraction timing can be realized if the rolling time prediction model of the rough mill 14 is suitable.

FIG. 2 shows the result of identifying a rough mill rolling time prediction model at a certain size by an off-line least squares method. The extraction pitch was determined using this model.

FIG. 3 shows a comparison between the predicted rolling time after a lapse of six months and the actual value. It can be seen that the rolling time prediction error has increased. This is because the rolling time and the acceleration / deceleration rate were adjusted, so that the rolling time prediction model did not fit. If the extraction pitch is determined using such a model, the rolling efficiency and the energy efficiency decrease.

Next, a sequential least squares method will be described as an example of the online identification method. The rough mill rolling time of the Nth material is expressed as y (N), the number of passes is expressed as p (N), and the elongation is expressed as q (N). Also, a parameter vector of the rolling time prediction model calculated based on the results up to the N-th material is defined by the following equation.

[0034]

(Equation 4)

The calculation algorithm for obtaining the parameters of the rolling time prediction model based on the rolling performance up to the N-th material is given as follows (sequential least squares method).

[0036]

(Equation 5)

Using the results identified online, the parameters of the rolling time prediction model are corrected each time the material is rolled according to the above equation.

FIG. 4 shows the result of modifying the rolling time prediction model by this method. It can be seen that the average value and the variance of the rolling time prediction error are smaller than those in FIG.

As a result, the extraction pitch per material could be shortened by an average of 4 seconds, and the rolling efficiency could be improved. Further, since the dispersion is reduced, the waiting time of the material before the rough mill can be reduced, and the energy loss can be reduced.

Note that ９9 and □ 11 in FIGS.
The former means a material having a total number of passes of 9 in a coarse mill, and the latter means a material having an eleven passes.

In the above embodiment, the present invention is applied to the prediction of the rolling time of a section steel. However, the application of the present invention is not limited to this. Can be used to predict

[0042]

According to the present invention, it is possible to reduce the average value and the variance of the prediction error of each equipment passing time represented by the rolling time. Therefore, the extraction timing of the material from the heating furnace can be further optimized, the extraction pitch can be shortened, and the rolling efficiency can be improved. In addition, it is less necessary for the material to wait before a rate-determining step such as a rough mill, so that energy loss can be reduced.

[Brief description of the drawings]

FIG. 1 is a process diagram showing a section steel mill rolling line, which is an example of an object to which the present invention is applied.

FIG. 2 is a diagram showing an example of a rolling time prediction error when a rough mill rolling time prediction model is identified offline by a conventional method.

FIG. 3 is a diagram showing an example of a rolling time prediction error when the rolling time prediction model used in FIG. 2 does not fit due to aging.

FIG. 4 is a diagram showing an example of a rolling time prediction error when the present invention is applied.

[Explanation of symbols]

 10: Heating furnace 12: Breakdown mill 14: Rough mill U1: Universal mill E1: Edger 16: Finishing mill (UF)

Claims (1)

[Claims] 1. A method for predicting and calculating the time from the extraction of a material to the passage of each equipment in a heating furnace, and determining the extraction time of the material so that interference between a preceding material and a succeeding material does not occur. In the method of automatically extracting materials from heating furnaces, each equipment transit time prediction formula is described as a function of the number of passes and pressure extension, and online identification is used based on the actual equipment transit time values of the preceding material. A method for automatically extracting a heating furnace in hot rolling, wherein the prediction formula is corrected to determine a heating furnace extraction time of the next material.