JP3302809B2 - Extraction pitch determination device in hot continuous rolling system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for determining an extraction pitch between two materials, a preceding material and a corresponding material, in a hot continuous rolling system in order to take timing of extracting a material from a heating furnace. More particularly, the present invention relates to an extraction pitch determination device in a hot continuous rolling system that determines an extraction pitch from a heating furnace optimally with high accuracy and enables stable operation.

[0002]

2. Description of the Related Art In general, in a hot continuous rolling system, in order to take the timing of extracting a material from a heating furnace, an extraction pitch between two materials is predicted from a rolling schedule of a preceding material and a rolling schedule of a corresponding material. (Decision).

FIG. 6 is a diagram showing an example of a predicted rolling schedule of a preceding material and a corresponding material obtained from a setting calculation result on which the extraction pitch is determined.

[0004] In Fig. 6, the horizontal axis indicates each facility on the rolling line, and the vertical axis indicates time, and the flow of time when the material is rolled is shown. That is, it shows the flow of time when the material passes through the check points (here, coarse, coil box, finish, coiler) on the rolling line.

In FIG. 6, A indicates a predicted rolling schedule of the preceding material, and B and C indicate predicted rolling schedules of the corresponding material. The corresponding material is rolled after a certain time interval from the preceding material, but at this time, the corresponding material predicted rolling schedule C is moved in parallel to the restricted time interval so that the time interval with the preceding material is minimized, By setting the timing as B, it is possible to set the most efficient time interval in performing rolling.

That is, the time interval when the line C obtained from the predicted rolling schedule is closest to the line A is the line B. At this time, how close it can be is determined by the following method.

In each facility, there is a time interval (gap time: TG _i ) that is restricted for each facility after the preceding material passes and before the succeeding material is rolled.

[0008] Thus, the expected time interval (Δt _i ) when the material is rolled must be greater than the constrained time interval TG _i . Accordingly, the line C in FIG. 1 satisfies Δt _i > TG _i (i = 1, 2, 3, 4), and the line B is translated to a position satisfying the condition of the constraint time interval. Becomes Therefore, for line B, Δt _i ≧ TG _i (i = 1, 2, 3, 4) is satisfied, and the most efficient rolling schedule case can be determined.

At this time, the time interval t ₀ at the extraction end of the line A and the line B is the extraction pitch.

Here, the extraction pitch t ₀ is determined on the basis of the result of the prediction calculation from the rolling schedule of the preceding material and the corresponding material and the constraint time interval. Conventionally, fixed data stored in a table is used. Therefore, if an error occurs in the actual rolling not reflected in the time also predictive calculation of the extracted pitch t _0, always the same error has occurred in the prediction calculation of the extracted pitch t _0.

That is, since the data at the time of actual rolling is not reflected in the prediction calculation, it does not match the time interval at the time of actual rolling, and the extraction pitch is determined with low accuracy.

[0012]

As described above, the conventional method for determining the extraction pitch has a problem that the extraction pitch from the heating furnace cannot be accurately determined.

An object of the present invention is to provide a highly reliable hot continuous rolling mill capable of accurately and optimally determining an extraction pitch from a heating furnace by reflecting actual rolling data in calculation of a predicted extraction pitch. It is an object of the present invention to provide an apparatus for determining an extracted pitch in a system.

[0014]

SUMMARY OF THE INVENTION In order to achieve the above object, in a hot continuous rolling system, in order to take the timing of extracting material from a heating furnace, extraction between a preceding material and a corresponding material is performed. In the extraction pitch determination device for determining a pitch, first, in the invention according to claim 1, a predicted rolling schedule, which is a time required for a preceding material to pass through each facility on a rolling line, is calculated based on a preceding material setting calculation result. Means for calculating a predicted rolling schedule, which is a time required for the material to pass through each facility on the rolling line, based on the result of the relevant material setting calculation; A constraint gap time table that sets the constraint time interval between the preceding material and the corresponding material in each facility on the line, and the pitch between the preceding material and the corresponding material is calculated by learning. Based on the time gap learning term table for setting the time gap learning term to perform and the predicted rolling schedule of the preceding material and the corresponding material calculated by each predicted rolling schedule calculating means, the learning term of the time gap learning term table is used. A predicted extraction pitch calculating means for calculating a predicted extraction pitch by calculating a minimum time interval satisfying the restricted time interval set by the restricted gap time table, and a predicted extraction pitch calculated by the predicted extraction pitch calculating means. The actual rolling data sampling means that extracts the material from the heating furnace based on the actual rolling data and takes in the actual rolling data of the preceding material and the corresponding material after the actual rolling, Each equipment on the rolling line based on the actual rolling data of the material A gap time calculating means for calculating an a definitive preceding material time interval corresponding material,
A gap time error calculating means for calculating an error between the time interval calculated by the gap time calculating means and the time interval calculated from the predicted rolling schedule; and a constraint gap time based on the error calculated by the gap time error calculating means. Correction means for correcting the data in the table.

According to the second aspect of the present invention, a predicted rolling schedule calculation is performed to calculate a predicted rolling schedule, which is a time required for the preceding material to pass through each facility on the rolling line, based on the calculation result of the preceding material setting. Means, a corresponding material predicted rolling schedule calculating means for calculating a predicted rolling schedule, which is a time required for the corresponding material to pass through each facility on the rolling line, based on the corresponding material setting calculation result; A constrained gap time table for setting a constrained time interval between the material and the material, a time gap learning item table for setting a time gap learning item for calculating a pitch between the preceding material and the material, and prediction rolling. Based on the predicted rolling schedule of the preceding material and the corresponding material calculated by the schedule calculating means, the time gap learning term Learning using the learning term of the table, finding the minimum time interval that satisfies the constraint time interval set by the constraint gap time table, and predicting extraction pitch calculating means for calculating the prediction extraction pitch; The material is extracted from the heating furnace based on the calculated predicted extraction pitch, and the actual rolling data sampling means for taking in the actual rolling data of the preceding material and the corresponding material after the actual rolling, and the actual rolling data sampling means for taking in the actual rolling data A gap time calculating means for calculating a time interval between the preceding material and the corresponding material in each equipment on the rolling line based on the actual rolling data of the preceding material and the corresponding material, and a time interval calculated by the gap time calculating means. Gap time error to calculate the error from the time interval calculated from the predicted rolling schedule Means out, based on the error calculated by the gap time error calculating unit, and configured to include an update means for updating the learning term time gap learning term table.

Further, according to the third aspect of the present invention, a predicted rolling schedule calculation is performed to calculate a predicted rolling schedule, which is a time required for the preceding material to pass through each facility on the rolling line, based on the calculation result of the preceding material setting. Means, a corresponding material predicted rolling schedule calculating means for calculating a predicted rolling schedule, which is a time required for the corresponding material to pass through each facility on the rolling line, based on the corresponding material setting calculation result; A constrained gap time table for setting a constrained time interval between the material and the material, a time gap learning item table for setting a time gap learning item for calculating a pitch between the preceding material and the material, and prediction rolling. Based on the predicted rolling schedule of the preceding material and the corresponding material calculated by the schedule calculating means, the time gap learning term The learning using the learning term of the table is performed, the minimum time interval that satisfies the constraint time interval set by the constraint gap time table is determined, and the prediction extraction pitch calculation means for calculating the prediction extraction pitch, and the prediction extraction pitch calculation means The material is extracted from the heating furnace based on the calculated predicted extraction pitch, and the actual rolling data sampling means for taking in the actual rolling data of the preceding material and the corresponding material after the actual rolling, and the actual rolling data sampling means for taking in the actual rolling data A gap time calculating means for calculating a time interval between the preceding material and the corresponding material in each equipment on the rolling line based on the actual rolling data of the preceding material and the corresponding material, and a time interval calculated by the gap time calculating means. Incorrect gap time to calculate the error from the time interval calculated from the predicted rolling schedule Calculating means, correcting means for correcting the data of the constraint gap time table based on the error calculated by the gap time error calculating means, and a time gap learning term table based on the error calculated by the gap time error calculating means. And updating means for updating the learning term.

[0017]

Therefore, in the extraction pitch determining apparatus in the hot continuous rolling system of the present invention, in determining the extraction pitch, the time gap between the preceding material and the relevant material obtained from the actual rolling data and the time gap obtained when calculating the prediction schedule are determined. Reflect the difference from the time gap that was set in the prediction schedule calculation, that is, correct the data in the constraint gap time table based on the difference or update the learning term in the time gap learning term table based on the difference By doing so, it is possible to determine an optimal extraction pitch with high accuracy.

[0018]

DESCRIPTION OF THE PREFERRED EMBODIMENTS First, the concept of the present invention will be described.

FIG. 3 is a diagram showing an example of the actual rolling schedule of the preceding material and the corresponding material in the hot continuous rolling system.

In FIG. 3, line D shows the actual rolling schedule of the preceding material, and line E shows the actual rolling schedule of the corresponding material. A broken line F indicates a predicted rolling schedule of the preceding material, and a broken line G indicates a predicted rolling schedule of the corresponding material. At this time, the actual gap time between the preceding material and the corresponding material is represented by Δt _iACT , and the predicted gap time is represented by Δt _i . As a result, the gap time in the finishing equipment was predicted at the same time interval as the constraint gap time at the time of prediction, but it can be understood that the gap time was actually the result of Δt _iACT .

Here, the rolling in the equipment set based on the restricted time interval is hereinafter referred to as a neck step.

First, the learning in the neck step will be considered.

Assuming that the line of the predicted rolling schedule of the preceding material is regulated by the finishing equipment based on the constraint gap time of the neck process, the predicted gap time (Δt _k ) at the finishing equipment (i = _k ) is limited by the constraint gap ( TG _k )
Has a value equal to.

The error (zt _k ) between the time gap (Δt _k ) at the time of the rolling schedule prediction in the neck process and the time gap (Δt _kACT ) of the actual rolling data is as follows: zt _k = Δt _{k −Δt kACT} (k = neck Process). Thus, optimizing constraints gap time more efficient _{value, TG KNW = TG k + α} * (zt k -TG k) α: a smoothing gain.

Next, consider learning at another facility other than the neck step.

_Assuming that the expected rolling time in these facilities obtained from the predicted rolling schedule calculation is calculated by T _i = f (P1, P2,...), The actual measured rolling time is T _iACT , when the rolling time is assumed to be T _i, the error of both at this time, is determined by _{zt i} = T _i -T iACT. As a result, in the calculation of the predicted rolling schedule, a deviation corresponding to this error occurs, and therefore, it is considered that this error is reflected as a learning term.

That is, the new learning term is calculated by zt _iNEW = zt _iOLD + β * (zt _i -zt _iOLD ), and the scheduled rolling time obtained from the predicted rolling schedule calculation is: T _iNEW = f (P1, P2,...) ) + _{Zt iNEW} .

As described above, the actual rolling data is reflected in the calculation of the predicted rolling schedule, and the accuracy of the predicted rolling schedule calculation can be improved.

Therefore, in the present invention, when determining the extraction pitch, an error between the time gap between the preceding material and the corresponding material obtained from the actual rolling data and the time gap obtained when calculating the predicted rolling schedule is calculated by the predicted rolling. By reflecting the result in the schedule calculation, the extraction pitch is accurately predicted.

Hereinafter, an embodiment of the present invention based on the above concept will be described in detail with reference to the drawings.

FIG. 1 is a functional block diagram showing a configuration example of an extraction pitch determination device in a hot continuous rolling system according to the present invention.

That is, as shown in FIG. 1, the apparatus for determining an extracted pitch according to the present embodiment comprises:
2, the constraint gap time table 13, the time gap learning term table 14, and the predicted extraction pitch calculation means 1
5, actual rolling data sampling means 16, gap time calculating means 17, gap time error calculating means 18, correcting means 19, and updating means 20.

Here, the preceding material predicted rolling schedule calculating means 11 calculates a predicted rolling schedule which is the time when the preceding material passes through each facility on the rolling line, based on the result of the preceding material setting calculation.

The corresponding material predicted rolling schedule calculating means 12 calculates a predicted rolling schedule, which is a time required for the corresponding material to pass through each facility on the rolling line, based on the corresponding material setting calculation result.

Further, the constraint gap time table 13
Is to set a restricted time interval between the preceding material and the corresponding material in each facility on the rolling line.

Further, the time gap learning term table 14 is for setting a time gap learning term for calculating a pitch between the preceding material and the corresponding material by learning.

On the other hand, the predicted extraction pitch calculating means 15 uses the learning term of the time gap learning term table 14 based on the predicted rolling schedule of the preceding material and the corresponding material calculated by the predicted rolling schedule calculating means 11 and 12. Learning, and constraint gap time table 1
The minimum time interval that satisfies the constraint time interval set in Step 3 is obtained, and the predicted extraction pitch is calculated.

The actual rolling data sampling means 1
Reference numeral 6 is for extracting a material from the heating furnace based on the predicted extraction pitch calculated by the predicted extraction pitch calculation means 15 and taking in the actual rolling data of the preceding material after the actual rolling and the corresponding material.

Further, the gap time calculating means 17
The time interval between the preceding material and the corresponding material in each facility on the rolling line is calculated based on the actual rolling data of the preceding material and the corresponding material taken in by the actual rolling data sampling means 16.

On the other hand, the gap time error calculating means 18
Calculates the error between the time interval calculated by the gap time calculation means 17 and the time interval calculated from the predicted rolling schedule.

The correcting means 19 corrects the data of the constraint gap time table 13 based on the error calculated by the gap time error calculating means 18.

Further, the updating means 20 updates the learning term of the time gap learning term table 14 based on the error calculated by the gap time error calculating means 18.

Next, the operation of the extraction pitch determining device in the hot continuous rolling system of the present embodiment configured as described above will be described with reference to the flowchart shown in FIG.

First, the preceding material predictive rolling schedule calculating means 11 is a time for the material of the preceding material to pass through each facility based on the result of the preceding material setting calculation which is data for setting in the rolling system. A rolling schedule is required.

Similarly, the corresponding material prediction rolling schedule calculation means 12 calculates the rolling schedule, which is the time for the material of the corresponding material to pass through each facility, on the basis of the corresponding material setting calculation result which is data for setting in the rolling system. Is required.

Next, the predicted extraction pitch calculating means 15
Based on the predicted rolling schedule of the preceding material and the corresponding material calculated by the predicted rolling schedule calculating means 11 and 12, learning is performed using the learning term of the time gap learning term table 14, and the constraint gap time table 1
The minimum time interval that satisfies the constraint time interval set in Step 3 is obtained, and the predicted extraction pitch is determined from this.

Next, the material is extracted from the heating furnace based on the predicted extraction pitch, and the rolling is actually performed based on the set calculation result calculated in advance.

Then, the actual rolling data sampling means 16 takes in the actual rolling data of the preceding material and the corresponding material after the completion of the rolling.

Next, the gap time calculating means 17
On the basis of the actual rolling data of the preceding material and the corresponding material taken in by the actual rolling data sampling means 16, the time interval between the preceding material and the corresponding material in each facility on the rolling line is calculated.

Next, the gap time error calculating means 18 calculates an error between the time interval calculated by the gap time calculating means 17 and the time interval calculated from the predicted rolling schedule.

Next, the correcting means 19 corrects the data of the constraint gap time table 13 based on the error calculated by the gap time error calculating means 18.

Further, the updating means 20 updates the learning term of the time gap learning term table 14 based on the error calculated by the gap time error calculating means 18.

Hereinafter, a description will be given of the learning in the neck step and the learning other than the neck step.

(A) Learning in Neck Process FIG. 4 is a diagram showing an example of actual values and predicted values of a preceding material and a corresponding material in a certain facility.

In FIG. 4, line H shows the actual rolling schedule of the preceding material, and line I shows the actual rolling schedule of the corresponding material. A broken line J indicates a predicted rolling schedule of the preceding material, and a broken line K indicates a predicted rolling schedule of the corresponding material.

The error z between the time interval Δt in the predicted rolling schedule calculation and the time interval Δt _ACT in the actual rolling is calculated.
t is determined.

Zt = Δt−Δt _{ACT Based} on this, the constraint gap time value is corrected. TG _NEW = TG + α * (zt−TG) TG: Constrained gap time Δt: Time interval in predicted rolling schedule calculation Δt _ACT : Time interval based on actual time α: Smooth gain TG _NEW : Constrained gap time after learning (b) Learning other than the neck process FIG. 5 is a diagram illustrating an example of the actual value of the corresponding material in a certain facility.

In FIG. 5, line L indicates the actual rolling schedule, and broken line M indicates the predicted rolling schedule.

The scheduled rolling time obtained by the predicted rolling schedule calculation in each facility is as follows: T = f (P1, P2,...) T: Scheduled rolling time f: Rolling schedule calculation formula P1, P2,. If the actual rolling time at this time is T _ACT , the error between the predicted value and the actual value at this time is zt = T−T _ACT , and the learning term is zt = zt _OLD + β * ( zt−zt _OLD ) Therefore, in the next predicted rolling schedule calculation, by performing the calculation as T _NEW = f (P1, P2,...) + zt _NEW , the actual rolling data is reflected and accurate prediction can be performed. .

As described above, the extraction pitch determining device in the hot continuous rolling system according to the present embodiment uses the predicted rolling time, which is the time required for the preceding material to pass through each facility on the rolling line, based on the calculation result of the preceding material setting. Predicted material rolling schedule calculating means 11 for calculating a schedule, and a predicted material rolling schedule calculation for calculating a predicted rolling schedule, which is a time required for the relevant material to pass through each facility on the rolling line, based on the relevant material setting calculation result. Means 12, a constraint gap time table 13 for setting a constraint time interval between the preceding material and the corresponding material in each facility on the rolling line, and a time gap learning term for calculating the pitch between the preceding material and the corresponding material by learning. Time gap learning term table 1 for setting
4 and each predicted rolling schedule calculating means 11 and 12
Time gap learning term table 14 based on the predicted rolling schedule of the preceding material and the corresponding material calculated by
A predicted extraction pitch calculating means 15 for obtaining a minimum time interval satisfying the constraint time interval set by the constraint gap time table 13 and calculating a predicted extraction pitch; 15. A rolling material sampling means 16 for extracting a material from the heating furnace based on the predicted extraction pitch calculated by 15 and taking in the actual rolling data of the preceding material and the corresponding material after the actual rolling, and the actual rolling data sampling Means 1
6, the gap time calculating means 17 for calculating the time interval between the preceding material and the corresponding material in each facility on the rolling line based on the actual rolling data of the preceding material and the corresponding material taken in by the step 6 and the gap time calculating means 17 Time error calculating means 18 for calculating an error between the set time interval and the time interval calculated from the predicted rolling schedule.
Correction means 19 for correcting the data in the constraint gap time table 13 based on the error calculated by the gap time error calculation means 18; and time gap learning based on the error calculated by the gap time error calculation means 18. An updating means 20 for updating a learning item in the item table 14 is provided.

Therefore, the purpose of reflecting the actual rolling data in the calculation of the predicted extraction pitch is to reflect the error between the time actually measured by rolling and the predicted schedule time in the constraint gap time and the time gap learning term to optimize the calculation. Therefore, the extraction pitch can be determined according to the actual rolling, and it is possible to determine the extraction pitch with extremely high accuracy.

Thus, a stable operation can be performed in the hot continuous rolling system.

The present invention is not limited to the above embodiment, but can be implemented in the following manner.

(A) In the above-described embodiment, a learning method generally used has been described as a learning method. However, the present invention is not limited to this, and learning is performed using, for example, a neurosystem as another learning method. It is also possible.

(B) In the above embodiment, the types of equipment on the rolling line are classified into four types: coarse, coil box, finish, and coiler.
Although the case of equipment has been described, the present invention is not limited to this, and equipment can be classified according to the system.

(C) In the above embodiment, in order to improve the accuracy of the extraction pitch, a method of correcting the constraint gap time by an error between the time actually measured by rolling and the predicted schedule time, and a time gap learning term by the error. The case where both of the updating methods are applied has been described. However, the present invention is not limited to this, and the accuracy of the extraction pitch is similarly improved by applying only the former method or the latter method. Is possible.

[0067]

As described above, according to the present invention, based on the preceding material setting calculation result, the preceding material predictive rolling for calculating the predicted rolling schedule which is the time when the preceding material passes through each facility on the rolling line. A schedule calculating means, a corresponding material predicted rolling schedule calculating means for calculating a predicted rolling schedule which is a time required for the corresponding material to pass through each facility on the rolling line, based on the corresponding material setting calculation result, and each facility on the rolling line A constraint gap time table that sets a constraint time interval between the preceding material and the corresponding material, a time gap learning item table that sets a time gap learning term for calculating the pitch between the preceding material and the relevant material by learning,
Based on the predicted rolling schedule of the preceding material and the corresponding material calculated by each predicted rolling schedule calculating means, learning using the learning term of the time gap learning term table is performed, and the constraint time interval set by the constraint gap time table is used. The minimum time interval to be satisfied is determined, the predicted extraction pitch calculation means for calculating the predicted extraction pitch, and the material is extracted from the heating furnace based on the predicted extraction pitch calculated by the predicted extraction pitch calculation means, and the rolling is actually performed. Based on the actual rolling data sampling means that captures the actual rolling data of the preceding and subsequent materials and the actual rolling data of the preceding material and the applicable material that are captured by the actual rolling data sampling means, the preceding rolling at each facility on the rolling line is performed. A gap time calculating means for calculating a time interval between the material and the corresponding material; A gap time error calculating means for calculating an error between the time interval calculated by the time calculating means and the time interval calculated from the predicted rolling schedule; and a constraint gap time table based on the error calculated by the gap time error calculating means. And the updating means for updating the learning term of the time gap learning term table based on the error calculated by the gap time error calculating means. An extremely reliable extraction pitch determination device in a hot continuous rolling system that can accurately and optimally determine an extraction pitch from a heating furnace by reflecting actual rolling data can be provided.

[Brief description of the drawings]

FIG. 1 is a functional block diagram showing an embodiment of an extraction pitch determination device in a hot continuous rolling system according to the present invention.

FIG. 2 is a flowchart for explaining the operation in the embodiment.

FIG. 3 is a diagram showing an example of an actual rolling schedule of a preceding material and a corresponding material in a hot continuous rolling system for explaining the concept of the present invention.

FIG. 4 is a view showing an example of a prediction / actual schedule of a preceding material / corresponding material for explaining an operation in the embodiment.

FIG. 5 is a view showing an example of an actual value and a predicted value for explaining the operation in the embodiment.

FIG. 6 is a diagram showing an example of a predicted rolling schedule of a preceding material and a corresponding material obtained from a setting calculation result serving as a basis for determining an extraction pitch.

[Explanation of symbols]

11: Predicted material predicted rolling schedule calculating means, 12: Applicable material predicted rolling schedule calculating means, 13: Constrained gap time table, 14: Time gap learning term table, 15: Predicted extraction pitch calculating means, 16: Actual rolling data sampling means , 17: gap time calculating means, 18: gap time error calculating means, 19: correcting means, 20: updating means.

Continuation of the front page (72) Inventor Toshie Yamamura 1 Toshiba-cho, Fuchu-shi, Tokyo Inside the Toshiba Fuchu factory (58) Fields investigated (Int. Cl. ⁷ , DB name) B21B 37/00 C21D 1/00 C21D 9/00

Claims (3)

(57) [Claims] 1. An extraction pitch determining apparatus for determining an extraction pitch between two materials, a preceding material and a corresponding material, in order to take timing of extracting a material from a heating furnace in a continuous hot rolling system. Based on the calculation result, a preceding material predicted rolling schedule calculating means for calculating a predicted rolling schedule that is a time when the preceding material passes through each facility on the rolling line, and based on the corresponding material setting calculation result, the relevant material is subjected to the rolling. Applicable material predicted rolling schedule calculation means for calculating a predicted rolling schedule that is a time to pass through each facility on the line, and a constraint gap time table for setting a constraint time interval between the preceding material and the corresponding material in each facility on the rolling line. And time gap learning for setting a time gap learning term for calculating a pitch between the preceding material and the corresponding material by learning. Learning using the learning term of the time gap learning term table based on the table and the predicted rolling schedule of the preceding material and the corresponding material calculated by each of the predicted rolling schedule calculating means, set by the constraint gap time table A minimum time interval that satisfies the set constraint time interval, a predicted extraction pitch calculation unit that calculates a predicted extraction pitch, and extracts a material from the heating furnace based on the predicted extraction pitch calculated by the predicted extraction pitch calculation unit. The actual rolling data sampling means to capture the actual rolling data of the preceding material and the corresponding material after actually performing the rolling, based on the actual rolling data of the preceding material and the applicable material taken in by the actual rolling data sampling means, Calculate the time interval between the preceding material and the corresponding material in each equipment on the rolling line A gap time calculation means for outputting, a gap time error calculation means for calculating an error between the time interval calculated by the gap time calculation means and a time interval calculated from the predicted rolling schedule, and the gap time error calculation means Correction means for correcting data of the constraint gap time table based on the calculated error. An extraction pitch determination apparatus for a continuous hot rolling system, comprising: 2. An extraction pitch determining device for determining an extraction pitch between two materials, a preceding material and a corresponding material, in order to take timing of extracting a material from a heating furnace in a continuous hot rolling system. Based on the calculation result, a preceding material predicted rolling schedule calculating means for calculating a predicted rolling schedule that is a time when the preceding material passes through each facility on the rolling line, and based on the corresponding material setting calculation result, the relevant material is subjected to the rolling. Applicable material predicted rolling schedule calculation means for calculating a predicted rolling schedule that is a time to pass through each facility on the line, and a constraint gap time table for setting a constraint time interval between the preceding material and the corresponding material in each facility on the rolling line. And time gap learning for setting a time gap learning term for calculating a pitch between the preceding material and the corresponding material by learning. Learning using the learning term of the time gap learning term table based on the table and the predicted rolling schedule of the preceding material and the corresponding material calculated by each of the predicted rolling schedule calculating means, set by the constraint gap time table A minimum time interval that satisfies the set constraint time interval, a predicted extraction pitch calculation unit that calculates a predicted extraction pitch, and extracts a material from the heating furnace based on the predicted extraction pitch calculated by the predicted extraction pitch calculation unit. The actual rolling data sampling means to capture the actual rolling data of the preceding material and the corresponding material after actually performing the rolling, based on the actual rolling data of the preceding material and the applicable material taken in by the actual rolling data sampling means, Calculate the time interval between the preceding material and the corresponding material in each equipment on the rolling line A gap time calculation means for outputting, a gap time error calculation means for calculating an error between the time interval calculated by the gap time calculation means and a time interval calculated from the predicted rolling schedule, and the gap time error calculation means Updating means for updating a learning term of the time gap learning term table based on the calculated error. An extraction pitch determining apparatus in a hot continuous rolling system, comprising: 3. An extraction pitch determining apparatus for determining an extraction pitch between two materials, a preceding material and a corresponding material, in order to take timing of extracting a material from a heating furnace in a continuous hot rolling system. Based on the calculation result, a preceding material predicted rolling schedule calculating means for calculating a predicted rolling schedule that is a time when the preceding material passes through each facility on the rolling line, and based on the corresponding material setting calculation result, the relevant material is subjected to the rolling. Applicable material predicted rolling schedule calculation means for calculating a predicted rolling schedule that is a time to pass through each facility on the line, and a constraint gap time table for setting a constraint time interval between the preceding material and the corresponding material in each facility on the rolling line. And time gap learning for setting a time gap learning term for calculating a pitch between the preceding material and the corresponding material by learning. Learning using the learning term of the time gap learning term table based on the table and the predicted rolling schedule of the preceding material and the corresponding material calculated by each of the predicted rolling schedule calculating means, set by the constraint gap time table A minimum time interval that satisfies the set constraint time interval, a predicted extraction pitch calculation unit that calculates a predicted extraction pitch, and extracts a material from the heating furnace based on the predicted extraction pitch calculated by the predicted extraction pitch calculation unit. The actual rolling data sampling means to capture the actual rolling data of the preceding material and the corresponding material after actually performing the rolling, based on the actual rolling data of the preceding material and the applicable material taken in by the actual rolling data sampling means, Calculate the time interval between the preceding material and the corresponding material in each equipment on the rolling line A gap time calculation means for outputting, a gap time error calculation means for calculating an error between the time interval calculated by the gap time calculation means and a time interval calculated from the predicted rolling schedule, and the gap time error calculation means Correcting means for correcting the data of the constraint gap time table based on the calculated error; and updating the learning term of the time gap learning term table based on the error calculated by the gap time error calculating means. Means for determining an extraction pitch in a continuous hot rolling system.