JPH06315746A - Method for measuring flow rate of molten metal in double layer casting process - Google Patents

Method for measuring flow rate of molten metal in double layer casting process

Info

Publication number
JPH06315746A
JPH06315746A JP10705693A JP10705693A JPH06315746A JP H06315746 A JPH06315746 A JP H06315746A JP 10705693 A JP10705693 A JP 10705693A JP 10705693 A JP10705693 A JP 10705693A JP H06315746 A JPH06315746 A JP H06315746A
Authority
JP
Japan
Prior art keywords
molten metal
flow rate
time
noise
mold
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP10705693A
Other languages
Japanese (ja)
Inventor
Kunitoshi Watanabe
国俊 渡辺
Takashi Sato
孝 佐藤
Masayuki Araki
誠之 荒木
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nippon Steel Corp
Original Assignee
Nippon Steel Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nippon Steel Corp filed Critical Nippon Steel Corp
Priority to JP10705693A priority Critical patent/JPH06315746A/en
Publication of JPH06315746A publication Critical patent/JPH06315746A/en
Pending legal-status Critical Current

Links

Landscapes

  • Measuring Volume Flow (AREA)
  • Continuous Casting (AREA)

Abstract

PURPOSE:To reduce the responding delay to the measured value while executing the removal of noise, at the time of measuring the flow rate of molten metal poured into a mold in continuous casting of double layer cast slab. CONSTITUTION:At the time of obtaining the flow rate of molten metal by a differential operation of data of the molten metal wt. with time series, the center of gravity in the data with the time series in a fixed time interval from the present to the past and the center of gravity in the data with the time series making the part the starting point by the differential time, are obtd. By executing the differential operation between two centers of gravities, the flow rate of the molten metal is obtd. Further, by executing in advance the removal of the noise in the wt. signal with a noise removal filter, the responding delay to the measured value of the flow rate is restrained to be small.

Description

【発明の詳細な説明】Detailed Description of the Invention

【0001】[0001]

【産業上の利用分野】本発明は複層鋳片の連続鋳造にお
いて、モールド内に注入される溶融金属(溶湯)の流量
を測定する方法に関するものである。
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for measuring the flow rate of molten metal (molten metal) injected into a mold in continuous casting of multi-layer cast pieces.

【0002】[0002]

【従来の技術】表層と内層が異なる金属組成からなる複
層鋳片を連続鋳造により製造するプロセスの構成例を図
8に示す。2組のレードル1a,1b 、レードルノズル2
a,2b、タンディッシュ3a,3b 、注入ノズル4a,4b
がモールド9の上に設けられ、ており、表層溶湯がレー
ドル1a 、レードルノズル2a 、タンディッシュ3a 、
注入ノズル4a を通してモールド9に注入され、内層溶
湯がレードル1b 、レードルノズル2b 、タンディッシ
ュ3b 、注入ノズル4b を通してモールド9に注入され
る。複層鋳造においては、目標の表層厚さを得るため
に、モールド9に注入される表層溶湯およびまたは内層
溶湯の流量を測定し制御する必要がある。そのために、
レードル1a,1b およびタンディッシュ3a,3b の重量
を測定し、従来一般的にはその時間変化から溶湯流量を
求めていた。
2. Description of the Related Art FIG. 8 shows a structural example of a process for producing a multi-layer cast slab having a metal composition having different surface layers and inner layers by continuous casting. 2 sets of ladle 1a, 1b, ladle nozzle 2
a, 2b, tundish 3a, 3b, injection nozzles 4a, 4b
Is provided on the mold 9, and the surface molten metal is ladle 1a, ladle nozzle 2a, tundish 3a,
It is injected into the mold 9 through the injection nozzle 4a, and the inner layer molten metal is injected into the mold 9 through the ladle 1b, the ladle nozzle 2b, the tundish 3b and the injection nozzle 4b. In multi-layer casting, it is necessary to measure and control the flow rates of the surface layer molten metal and / or the inner layer molten metal injected into the mold 9 in order to obtain a target surface layer thickness. for that reason,
The weights of the ladles 1a and 1b and the tundishes 3a and 3b were measured, and conventionally, the flow rate of the molten metal was generally obtained from the change over time.

【0003】溶湯流量演算装置の構成例を図9に示す。
レードル1の重量がレードル重量演算装置11に入力さ
れ、タンディッシュ3の重量がタンディッシュ重量演算
装置12に入力されて、レードル溶湯流量、タンディッ
シュ溶湯流量が出力され、両者の和がモールドへの溶湯
注入流量である。なおプロセスによっては、表層レード
ル1a がない構成もあり、その場合はタンディッシュ3
の重量変化のみで溶湯流量が測定できる。
FIG. 9 shows an example of the structure of a molten metal flow rate calculating device.
The weight of the ladle 1 is input to the ladle weight calculation device 11, the weight of the tundish 3 is input to the tundish weight calculation device 12, the ladle molten metal flow rate and the tundish molten metal flow rate are output, and the sum of the two is sent to the mold. It is the molten metal injection flow rate. Depending on the process, there is a structure without the surface ladle 1a, in which case the tundish 3
The molten metal flow rate can be measured only by the change in weight.

【0004】流量演算装置11および12における従来
の演算方法を図10に示す。破線を真の重量とすると、
機械振動や電気ノイズのために重量計信号をτ秒毎にサ
ンプリングして得られる測定値は○印のように真値の周
りをばらつく。一例として、タンディッシュ重量は数十
トン程度に対し、溶湯流量は1秒間あたり数Kg程度であ
るために、重量としては小さなノイズであっても、流量
演算過程では非常に大きなノイズとなる。本来、溶湯流
量は重量信号を微分すれば正しく求めることができる。
しかし、上述のごとく重量信号に雑音を含んでいる場合
には演算結果が大きく乱れ、実用に耐えないことが殆ど
である。したがって図10に示すごとく有限な時間幅Δ
t,その間の重量変化ΔW(k) とし、時点kの溶湯流量
Q(k) をΔW(k) /Δtとして求める、いわゆる差分演
算が一般に用いられている。図11は鋼についての溶湯
流量測定の実例であり、差分時間Δtが1秒と10秒で
重量を差分演算したものであるが、1秒差分においても
ノイズの影響を大きく受けており、特に制御信号として
は実用に耐えない。そのため、従来は差分時間Δtを大
きくしてノイズの影響を軽減する方法がとられている。
FIG. 10 shows a conventional calculation method in the flow rate calculation devices 11 and 12. If the broken line is the true weight,
Due to mechanical vibrations and electrical noise, the measured value obtained by sampling the weighing signal every τ seconds varies around the true value as indicated by ○. As an example, the weight of the tundish is about several tens of tons and the flow rate of the molten metal is about several Kg per second. Therefore, even if the noise is small in weight, it becomes very large noise in the flow rate calculation process. Originally, the molten metal flow rate can be accurately obtained by differentiating the weight signal.
However, as described above, when the weight signal contains noise, the calculation result is greatly disturbed, and in most cases it cannot be put to practical use. Therefore, as shown in FIG. 10, a finite time width Δ
A so-called difference calculation is generally used, where t is the weight change ΔW (k) during that time, and the melt flow rate Q (k) at time k is ΔW (k) / Δt. FIG. 11 is an example of the molten metal flow rate measurement for steel, in which the difference time Δt is the difference calculation of the weight at 1 second and 10 seconds. The difference of 1 second is also greatly affected by noise, and the control is particularly effective. Not practical for use as a signal. Therefore, conventionally, a method of increasing the difference time Δt to reduce the influence of noise has been adopted.

【0005】いま、τ秒毎にサンプリングして得られる
時系列重量データを現時点から過去に向かってW(k),W
(k-1),W(k-2),・・・とする。それぞれに対応する真の
重量と、そのときのノイズをそれぞれD(k),D(k-1),D
(k-2),・・・、およびN(k),N(k-1),N(k-2),・・・と
する。このとき、一般に(1)式が成り立つ。 W(k) =D(k) +N(k) (1) 図10より差分時間Δtは、◎ Δt=τd (2) である。時点kの流量Q(k) を求めるための差分演算は
(3)式(4)式で行われる。
Now, the time series weight data obtained by sampling every τ seconds is W (k), W from the present time to the past.
(k-1), W (k-2), ... The true weight corresponding to each and the noise at that time are respectively D (k), D (k-1), D
(k-2), ... And N (k), N (k-1), N (k-2) ,. At this time, the equation (1) is generally established. W (k) = D (k) + N (k) (1) From FIG. 10, the difference time Δt is ⊚Δt = τd (2). The difference calculation for obtaining the flow rate Q (k) at the time point k is performed by the expressions (3) and (4).

【0006】 Q(k) =[W(k-d) −W(k) ]/Δt (3) =[D(k-d) −D(k) ]/Δt+[N(k-d) −N(k) ]/Δt (4) (4)式の右辺第1項はΔt秒間の真の平均溶湯流量Q
r(k) 、第2項はノイズQn(k) である。いま、ノイズ
は白色で分散σ2 のガウス分布に従うとし、ノイズをN
(σ)と表せば、(4)式は次のようになる。 Q(k) =Qr(k) +21/2 N(σ)/Δt (5) すなわち(5)式より、測定溶湯流量Q(k) を真値に近
付けるためにはΔtを大きくすればよいことがわかる。
事実、図11に示すとおり、差分時間を大きくした10
秒差分演算による測定結果Aは、高周波ノイズがよく除
去されている。Bは1秒差分演算による測定結果であ
る。
Q (k) = [W (kd) −W (k)] / Δt (3) = [D (kd) −D (k)] / Δt + [N (kd) −N (k)] / Δt (4) The first term on the right side of the equation (4) is the true average molten metal flow rate Q for Δt seconds.
r (k) and the second term are noise Qn (k). Now, assume that the noise is white and follows a Gaussian distribution with variance σ 2 , and the noise is N
When expressed as (σ), the equation (4) is as follows. Q (k) = Qr (k) +2 1/2 N (σ) / Δt (5) That is, according to the equation (5), Δt may be increased to bring the measured molten metal flow rate Q (k) close to the true value. I understand.
In fact, as shown in FIG. 11, the difference time is increased by 10
In the measurement result A obtained by the second difference calculation, high frequency noise is well removed. B is the measurement result by the 1-second difference calculation.

【0007】[0007]

【発明が解決しようとする課題】ところで、(5)式の
Δtを大きくすればノイズを除去できる反面、測定値の
応答性が悪くなる。10秒差分演算による測定値は1秒
差分に比べて応答が遅れていることが図11に明瞭に現
れている。測定値の応答が遅れれば、例えば流量制御を
行う場合に制御性が悪化する。つまり制御の応答性が低
下したり、外乱に対する安定性が低下する。
By the way, if Δt in the equation (5) is increased, noise can be removed, but the responsiveness of the measured value is deteriorated. It is clearly shown in FIG. 11 that the measured value obtained by the 10-second difference calculation has a delayed response compared to the 1-second difference. If the response of the measured value is delayed, the controllability deteriorates, for example, when the flow rate control is performed. That is, the control responsiveness is reduced and the stability against disturbance is reduced.

【0008】本発明は、ノイズ除去を行いながら測定値
の応答遅れを小さくする溶湯流量測定方法を提供するも
のである。
The present invention provides a molten metal flow rate measuring method for reducing noise response and reducing response delay of measured values.

【0009】[0009]

【課題を解決するための手段】本発明は、表層タンディ
ッシュおよび内層タンディッシュから異なる金属組成の
溶湯を、それぞれ注入ノズルを介してモールド内に注入
し、電磁ブレーキにより該モールド内の溶湯に制動力を
与え、該モールド内に形成される境界層によって表層溶
湯および内層溶湯を上下に分離させて複層鋳片を鋳造す
る連続鋳造において、溶湯重量時系列データの差分演算
により溶湯流量を求めるにあたり、第1発明は、現時点
から過去に向かって一定時間幅内の時系列データの重心
と、差分時間だけ過去を起点とする時系列データの重心
とを求め、該2つの重心間で差分演算を行うことによ
り、溶湯流量を求めることを特徴とする複層鋳造プロセ
スにおける溶湯流量測定方法である。また第2発明は、
事前にノイズ除去フィルターで重量信号のノイズ除去を
行うことにより、流量測定値の応答遅れを小さく抑える
ことを特徴とする複層鋳造プロセスにおける溶湯流量測
定方法である。
According to the present invention, molten metal having a different metal composition is poured from a surface tundish and an inner layer tundish into a mold through respective pouring nozzles, and the molten metal in the mold is controlled by an electromagnetic brake. In continuous casting in which power is applied and the surface layer molten metal and the inner layer molten metal are separated vertically by the boundary layer formed in the mold to cast a multi-layer slab, the molten metal flow rate is calculated by the difference calculation of the molten metal weight time series data. According to the first invention, the centroid of time-series data within a certain time width from the present time to the past and the centroid of time-series data starting from the past by a difference time are obtained, and a difference calculation is performed between the two centroids. This is a method for measuring a molten metal flow rate in a multi-layer casting process, which is characterized in that a molten metal flow rate is obtained by carrying out. The second invention is
This is a method for measuring a molten metal flow rate in a multi-layer casting process, which is characterized by suppressing a response delay of a flow rate measurement value by removing noise of a weight signal with a noise removal filter in advance.

【0010】以下、本発明の溶湯流量測定方法について
詳細に説明する。図1に第1発明の方法を示す。現時点
kのデータを含めて過去n個(楕円内のデータ)の重量
データの重心Wm(k)と、同様に差分時間Δtm 秒過去の
データ群n個の重心Wm(k-dm) を計算し、その2つの重
心の差分演算により溶湯流量を求めるものである。2つ
の重心は(6),(7)式で求められる。
The molten metal flow rate measuring method of the present invention will be described in detail below. FIG. 1 shows the method of the first invention. The center of gravity Wm (k) of the past n pieces of data (data within the ellipse) including the data of the present time k and the center of gravity Wm (k-dm) of the past n data groups of the difference time Δtm seconds are calculated in the same manner. The molten metal flow rate is obtained by calculating the difference between the two center of gravity. The two centers of gravity are obtained by the equations (6) and (7).

【0011】[0011]

【数1】 [Equation 1]

【数2】 図1より差分時間Δtm は(8)式となる。 Δtm =τ・dm (8) 時点kの溶湯流量Qm(k)は、重心間の差分演算で得られ
る。
[Equation 2] From FIG. 1, the difference time Δtm is given by equation (8). Δtm = τ · dm (8) The molten metal flow rate Qm (k) at the time point k is obtained by the difference calculation between the centers of gravity.

【0012】[0012]

【数3】 ここで測定値W(k) は(1)式に示す如く真値D(k) と
ノイズN(k) の和で表され、ノイズN(k) は白色で分散
σ2 のガウス分布に従うとし、N(σ)と表せば(5)
式より、(10)式は(11)式のように表される。
[Equation 3] Here, the measured value W (k) is represented by the sum of the true value D (k) and the noise N (k) as shown in the equation (1), and the noise N (k) is white and follows the Gaussian distribution of variance σ 2. , N (σ) (5)
From the equation, the equation (10) is expressed as the equation (11).

【0013】[0013]

【数4】 (11)式の右辺第1項は時点kにおける差分時間Δt
m 秒間の平均流量Qr(k)を含めた、過去n個の平均流量
のある平均である。同第2項はノイズである。前記
(5)式の右辺第2項のノイズと、(11)式右辺第2
項のノイズを等置すれば、(12)式を得る。 Δtm =Δt/n1/2 (12) すなわち、本重心差分法によれば従来の単純差分法に比
べて、同じノイズの大きさで差分時間を小さくできるこ
とが分かる。
[Equation 4] The first term on the right side of the equation (11) is the difference time Δt at the time point k.
It is an average of the past n average flow rates including the average flow rate Qr (k) for m seconds. The second term is noise. The noise of the second term on the right side of the equation (5) and the second noise on the right side of the equation (11)
Equation (12) is obtained by equalizing the noises of the terms. Δtm = Δt / n 1/2 (12) That is, it can be seen that the present centroid difference method can reduce the difference time with the same noise magnitude as compared with the conventional simple difference method.

【0014】図2は本発明法の効果を示す鋼についての
実例である。Aは差分時間Δtが10秒の単純差分演算
による従来法の流量測定結果であり、Cは差分時間Δt
m が5秒で重心演算データ個数nが3の重心差分による
本発明法の流量測定結果を示している。同図から明らか
なように、両者ともノイズ除去の点では差はないが、本
発明の重心差分法は単純差分法に比べて応答遅れが小さ
いことが分かる。また、図3に両者の周波数応答を示
す。図3から明らかなように、従来の単純差分法Aに比
べて、本発明の重心差分法Cの位相遅れが小さく応答性
が改善されていることが分かる。
FIG. 2 is an example of steel showing the effect of the method of the present invention. A is the flow rate measurement result of the conventional method by the simple difference calculation with the difference time Δt of 10 seconds, and C is the difference time Δt.
The flow rate measurement result of the method of the present invention is shown by the difference of the center of gravity when m is 5 seconds and the number of gravity center calculation data n is 3. As is clear from the figure, although there is no difference in terms of noise removal, the centroid difference method of the present invention has a smaller response delay than the simple difference method. Further, FIG. 3 shows the frequency responses of both. As is clear from FIG. 3, the phase difference of the center-of-gravity difference method C of the present invention is smaller than that of the conventional simple difference method A, and the response is improved.

【0015】つぎに第2発明について説明する。図4は
ノイズ除去フィルターの一つであるカルマンフィルター
の構成例である。本フィルターは重量測定値W(k) が得
られる度に、決められた手順で計算を行うことにより真
の重量の最適推定値Y(k) を得ることができる。まず、
カルマンフィルターによるノイズ除去を行うためのシス
テムモデルをつぎのように定める。
Next, the second invention will be described. FIG. 4 is a configuration example of a Kalman filter which is one of noise removal filters. This filter can obtain the optimum estimated value Y (k) of the true weight by performing the calculation in the determined procedure each time the weight measurement value W (k) is obtained. First,
The system model for noise removal by Kalman filter is defined as follows.

【0016】 X(k) =X(k-1) +B(k-1) U(k-1) +e(k-1) (13) W(k) =X(k) +f(k) (14) ただし、X:真の重量 B:流量係数 U:ストッパー開度 e:システムノイズ W:重量測定値 f:観測ノイズ (13)式はシステム方程式であり、ストッパーまたは
SN(スライディングノズル)操作による溶湯流量が積
分されて重量変化が起こることを示している。システム
ノイズeはストッパーまたはSN流量係数Bの変動をノ
イズとして表現している。一般に、連続鋳造設備のスト
ッパーまたはSN流量係数Bは変動が大きい割りに重量
Xの変動は緩やかである。このような場合、(13)式
の右辺第2項全体をノイズとみなし、それを第3項に含
めても真の重量Xを推定することは可能である。したが
って、(13)式をつぎのように表す。 X(k) =X(k-1) +e(k-1) (15) 一方(14)式は観測方程式で、真の重量Xに測定ノイ
ズfが重畳したものが重量測定値Wとして観測されるこ
とを示す。
X (k) = X (k-1) + B (k-1) U (k-1) + e (k-1) (13) W (k) = X (k) + f (k) (14) ) However, X: True weight B: Flow coefficient U: Stopper opening e: System noise W: Weight measurement value f: Observation noise Equation (13) is a system equation, and is a molten metal due to stopper or SN (sliding nozzle) operation. It shows that the flow rate is integrated and a weight change occurs. The system noise e expresses the fluctuation of the stopper or the SN flow coefficient B as noise. Generally, the stopper or the SN flow coefficient B of the continuous casting equipment has a large fluctuation, but the fluctuation of the weight X is gentle. In such a case, it is possible to estimate the true weight X even if the entire second term on the right side of the equation (13) is regarded as noise and included in the third term. Therefore, the equation (13) is expressed as follows. X (k) = X (k-1) + e (k-1) (15) On the other hand, the equation (14) is an observation equation, in which the measurement noise f superimposed on the true weight X is observed as the weight measurement value W. Indicates that

【0017】このとき、システムノイズeの分散はS、
および観測ノイズfの分散はRであるとし、そのスペク
トルは白色とする。(14),(15)式をもとにカル
マンフィルターを適用すると、真の重量Xの最適推定値
Yは(16)〜(18)式で求められる。これらの式
は、初期値が与えられれば重量測定値W(k) が得られる
度に、Y,G,Pがサイクリックに計算できる。 Y(k) =Y(k-1) +G(k)[W(k) −Y(k-1)] (16) G(k) =[P(k-1) +S] [ P(k-1) +S+R]-1 (17) P(k) =[1−G(k)] [P(k-1) +S] (18) ただし、G(k) はフィルターゲインである。また、P
(k) は最適推定値Y(k) の推定誤差の分散である。ま
た、初期値はつぎのように設定する。 P(0)=W(0)2 (19) Y(0)=0 (20) また、ノイズの分散はつぎのように表される。 S=e2 (21) R=f2 (22) これらのSとRはパラメータとして予め設定しておく。
もし、ノイズeとfが明確でない場合には、試行錯誤的
に値を定める。以上の処理により、重量推定値Wに比べ
て、より真値に近い重量の最適推定値が得られる。
At this time, the variance of the system noise e is S,
And the variance of the observation noise f is R, and its spectrum is white. When the Kalman filter is applied based on the equations (14) and (15), the optimum estimated value Y of the true weight X is obtained by the equations (16) to (18). In these equations, Y, G, and P can be cyclically calculated every time the weight measurement value W (k) is obtained if an initial value is given. Y (k) = Y (k-1) + G (k) [W (k) -Y (k-1)] (16) G (k) = [P (k-1) + S] [P (k- 1) + S + R] -1 (17) P (k) = [1-G (k)] [P (k-1) + S] (18) where G (k) is a filter gain. Also, P
(k) is the variance of the estimation error of the optimum estimated value Y (k). The initial value is set as follows. P (0) = W (0) 2 (19) Y (0) = 0 (20) Further, the variance of noise is expressed as follows. S = e 2 (21) R = f 2 (22) These S and R are preset as parameters.
If the noises e and f are not clear, the values are determined by trial and error. By the above processing, the optimum estimated value of weight closer to the true value than the estimated weight value W is obtained.

【0018】つぎに、重量の最適推定値Yを用いて次式
で溶湯流量Qm を計算する。 Qm =[Y(k-dm)−Y(k) ]/Δtm (23) ただし、tm は1以上の整数。 Δtm =τdm (24) このようにYを用いて流量演算を行えば、図10に示す
差分時間Δtを小さくできる。つまり、ノイズを効果的
に除去しつつ、応答遅れの小さい流量測定が可能にな
る。
Next, the molten metal flow rate Qm is calculated by the following equation using the optimum estimated value Y of the weight. Qm = [Y (k-dm) -Y (k)] / Δtm (23) where tm is an integer of 1 or more. Δtm = τdm (24) By performing the flow rate calculation using Y in this way, the difference time Δt shown in FIG. 10 can be reduced. That is, it is possible to measure the flow rate with a small response delay while effectively removing the noise.

【0019】図5は本発明法の効果を示す鋼についての
実例である。Aは差分時間Δtが10秒の単純差分演算
による従来法の流量測定結果であり、Dはカルマンフィ
ルターによる重量信号のノイズ除去後、差分時間Δtm
が1秒で求めた本発明法の流量測定結果を示している。
図5から明らかなように、両者ともノイズ除去の点では
差はないが、本発明法は従来の単純差分法に比べて応答
遅れが小さいことが分かる。また、カルマンフィルター
のノイズの分散SとRは、それぞれ20と300の例で
ある。一般にS,Rを大きくすればノイズは小さくなる
反面、応答遅れも増加するので、それらにも最適値が存
在する。
FIG. 5 is an example of steel showing the effect of the method of the present invention. A is the flow rate measurement result of the conventional method by the simple difference calculation with the difference time Δt of 10 seconds, and D is the difference time Δtm after noise removal of the weight signal by the Kalman filter.
Shows the flow rate measurement result of the method of the present invention obtained in 1 second.
As is clear from FIG. 5, although there is no difference in terms of noise removal, the method of the present invention has a smaller response delay than the conventional simple difference method. The Kalman filter noise variances S and R are 20 and 300, respectively. Generally, when S and R are increased, noise is reduced, but response delay is also increased, and therefore, optimum values also exist for them.

【0020】また、図3に両者の周波数応答を示す。図
から明らかなように、従来の単純差分法Aに比べて、本
発明のカルマンフィルターのよる方法Dは位相遅れが小
さく応答性が改善されていることが分かる。
FIG. 3 shows the frequency response of both. As is clear from the figure, compared with the conventional simple difference method A, the method D according to the Kalman filter of the present invention has a smaller phase delay and improved response.

【0021】[0021]

【実施例】まず、第1発明の実施例を示す。図6は、図
9に示す流量演算装置11および12の詳細構成例であ
る。この流量演算装置13は、図8に示す各重量計に対
してそれぞれ設けられている。まず、(8)式の差分時
間Δtm またはそれに対応するデータ個数dm 、および
重心計算用データ個数nを、パラメータ入力部14より
あらかじめ入力しておく。データ入力部15でサンプリ
ング周期τ秒毎にレードルまたはタンディッシュの重量
計信号を読み込む。記憶演算部16では読み込まれた時
系列データを少なくとも(dm +n)個以上記憶する。
また記憶演算部16において、新たな重量データが読み
込まれる度に、(6),(7)および(9)式の演算を
行い、現時点kにおける溶湯流量Qm(k)を得る。さらに
重量時系列信号W(k) をW(k-1) へ、またW(k-1) をW
(k-2) へと順に過去側にシフトする。データ出力部17
では、次回の流量演算値が得られるまでQm(k)を保持す
る。このようにして、図8に示す各重量計毎に溶湯流量
が得られるので、図9に示す如くレードル溶湯流量とタ
ンディッシュ溶湯流量を加算すれば、モールドに注入さ
れる内層(または表層)流量を求めることができる。
First, an embodiment of the first invention will be described. FIG. 6 is a detailed configuration example of the flow rate calculation devices 11 and 12 shown in FIG. The flow rate calculation device 13 is provided for each weighing scale shown in FIG. First, the difference time Δtm in the equation (8) or the number of data dm corresponding thereto and the number of gravity center calculation data n are input in advance from the parameter input unit 14. The data input unit 15 reads a ladle or tundish weighing scale signal every sampling period τ seconds. The storage operation unit 16 stores at least (dm + n) or more pieces of read time series data.
In addition, the storage calculation unit 16 calculates the equations (6), (7) and (9) each time new weight data is read, and obtains the molten metal flow rate Qm (k) at the present time k. Furthermore, the weight time series signal W (k) is changed to W (k-1) and W (k-1) is changed to W
Shift to the past side in order to (k-2). Data output unit 17
Then, Qm (k) is held until the next flow rate calculation value is obtained. In this way, since the molten metal flow rate can be obtained for each weighing scale shown in FIG. 8, if the ladle molten metal flow rate and the tundish molten metal flow rate are added as shown in FIG. 9, the inner layer (or surface layer) flow rate injected into the mold is obtained. Can be asked.

【0022】つぎに、第2発明の実施例を示す。図6に
おいて、(21),(22)式のパラメータS,R、お
よび(24)式の差分時間Δtm に対応するデータ個数
dmをパラメータ入力部14よりあらかじめ入力してお
く。データ入力部15でサンプリング周期τ秒毎にレー
ドルまたはタンディッシュの重量計信号を読み込む。記
憶演算部16では重量の最適推定値Yと、それに基づく
溶湯流量Qm を計算する。データ出力部17では次回の
流量演算値が得られるまでQm(k)を保持する。
Next, an embodiment of the second invention will be shown. In FIG. 6, the parameters S and R in the equations (21) and (22) and the data number dm corresponding to the difference time Δtm in the equation (24) are input in advance from the parameter input unit 14. The data input unit 15 reads a ladle or tundish weighing scale signal every sampling period τ seconds. The storage calculation unit 16 calculates an optimum weight estimation value Y and a molten metal flow rate Qm based on the optimum weight estimation value Y. The data output unit 17 holds Qm (k) until the next calculated flow rate value is obtained.

【0023】ここで、重量の最適推定値Yを求める方法
を示す。図7はその手順を示している。重量測定値W
(k) が読み込まれたら、まず(17)式によりフィルタ
ーゲインG(k) を計算する。このとき、(21),(2
2)式のパラメータS,RとYの誤差分散の前回値P(k
-1) を記憶部から取り出して計算する。つぎに、前回の
重量の最適推定値Y(k-1) を記憶部から取り出し、(1
6)式で今回の重量の最適推定値Y(k) を計算する。今
回のY(k) は出力部に出力されるとともに、次回のY計
算用に記憶される。最後に、記憶部からP(k-1),Sを取
り出し、今回のP(k) を計算し記憶する。Pは常に最新
のデータを1つだけ記憶しておけばよい。Yについては
差分による流量演算のために少なくとも(dm +1)個
記憶する。時点kのY(k) が得られたら、記憶部のY
(k) をY(k-1) へ、Y(k-1) をY(k-2) へと順に過去側
へシフトする。
Here, a method for obtaining the optimum estimated value Y of weight will be described. FIG. 7 shows the procedure. Weight measurement value W
When (k) is read, first the filter gain G (k) is calculated by the equation (17). At this time, (21), (2
2) The previous value P (k of the error variance of the parameters S, R and Y of the equation)
-1) is taken out from the storage unit and calculated. Next, the optimum estimated value Y (k-1) of the previous weight is retrieved from the storage unit, and (1
The optimum estimated value Y (k) of the weight this time is calculated by the equation 6). The current Y (k) is output to the output unit and stored for the next Y calculation. Finally, P (k-1), S is taken out from the storage unit, and the current P (k) is calculated and stored. It is sufficient for P to always store only the latest data. For Y, at least (dm +1) are stored for flow rate calculation by the difference. When Y (k) at time k is obtained, Y in the storage unit
Shift (k) to Y (k-1) and Y (k-1) to Y (k-2) in order to the past.

【0024】上記手順で時点kの重量最適推定値Y(k)
が得られたら、(23)式により溶湯流量Qm を求め
る。このようにして、図8に示す各重量計毎に溶湯流量
が得られるので、図9に示すごとくレードル溶湯流量と
タンディッシュ溶湯流量を加算すれば、モールドに注入
される内層(または表層)流量を求めることができる。
The weight optimum estimated value Y (k) at the time point k is calculated by the above procedure.
When is obtained, the melt flow rate Qm is calculated by the equation (23). In this way, since the molten metal flow rate can be obtained for each weighing scale shown in FIG. 8, if the ladle molten metal flow rate and the tundish molten metal flow rate are added as shown in FIG. 9, the inner layer (or surface layer) flow rate injected into the mold is obtained. Can be asked.

【0025】[0025]

【発明の効果】本発明法によれば、複層鋳造プロセスに
おける溶湯の流量測定が、ノイズを十分に除去し、かつ
応答性よく行うことができる。その結果、複層鋳造プロ
セスにおいて必要となる表層(または内層)の溶湯流量
制御の安定度向上および流量制御偏差の減少等の効果が
ある。特に表層(または内層)注入ノズルの介在物剥離
時の大きな外乱が制御系に加わった場合などに、制御の
乱れを小さく抑えることができ、表層厚みや表内層の溶
湯成分などの鋳片品質を良好に維持することができる。
According to the method of the present invention, the flow rate of the molten metal in the multi-layer casting process can be sufficiently removed of noise and responsively. As a result, there are effects such as improvement in stability of melt flow rate control of the surface layer (or inner layer) and reduction in flow rate control deviation, which are required in the multi-layer casting process. Especially when a large disturbance is added to the control system when inclusions in the surface layer (or inner layer) injection nozzle are peeled off, the control disturbance can be suppressed to a small level, and the slab quality such as the surface layer thickness and the melt component of the surface inner layer can be reduced. Can be maintained well.

【図面の簡単な説明】[Brief description of drawings]

【図1】本発明(第1発明)の説明図である。FIG. 1 is an explanatory diagram of the present invention (first invention).

【図2】本発明(第1発明)の効果を示す図である。FIG. 2 is a diagram showing an effect of the present invention (first invention).

【図3】本発明(第1、第2発明)の効果を示す図であ
る。
FIG. 3 is a diagram showing effects of the present invention (first and second inventions).

【図4】本発明(第2発明)の説明図である。FIG. 4 is an explanatory diagram of the present invention (second invention).

【図5】本発明(第2発明)の効果を示す図である。FIG. 5 is a diagram showing an effect of the present invention (second invention).

【図6】本発明(第1、第2発明)の詳細構成図であ
る。
FIG. 6 is a detailed configuration diagram of the present invention (first and second inventions).

【図7】本発明(第2発明)の説明図である。FIG. 7 is an explanatory diagram of the present invention (second invention).

【図8】本発明の対象プロセスの説明図である。FIG. 8 is an explanatory diagram of a target process of the present invention.

【図9】本発明の対象プロセスの説明図である。FIG. 9 is an explanatory diagram of a target process of the present invention.

【図10】従来法の説明図である。FIG. 10 is an explanatory diagram of a conventional method.

【図11】従来法の説明図である。FIG. 11 is an explanatory diagram of a conventional method.

【符号の説明】[Explanation of symbols]

1:レードル 2:レードルノズル 3:タンディッシュ 4:注入ノズル 5:レードル重量計 6:レードルスライディングノズル 7:タンディッシュ重量計 8:タンディッシュストッパー 9:モールド 10:電磁ブレーキ 11:レードル流量演算装置 12:タンディッシュ流量演算装置 13:流量演算装置 14:パラメータ入力部 15:データ入力部 16:記憶演算部 17:データ出力部 A:従来法の差分時間Δtが10秒の単純差分演算によ
る測定結果 B:従来法の差分時間Δtが1秒の単純差分演算による
測定結果 C:本発明法(第1発明)の差分時間Δtm が5秒の重
心差分演算による測定結果 D:本発明法(第2発明)のカルマンフィルターによる
重量信号のノイズ除去後、差分時間Δtm が1秒の測定
結果
1: Ladle 2: Ladle Nozzle 3: Tundish 4: Injection Nozzle 5: Ladle Weighing Scale 6: Ladle Sliding Nozzle 7: Tundish Weighing Scale 8: Tundish Stopper 9: Mold 10: Electromagnetic Brake 11: Ladle Flow Calculator 12 : Tundish flow rate calculation device 13: Flow rate calculation device 14: Parameter input part 15: Data input part 16: Memory calculation part 17: Data output part A: Measurement result by simple difference calculation with difference time Δt of 10 seconds in conventional method B : Measurement result by simple difference calculation with difference time Δt of 1 second in conventional method C: Measurement result by center of gravity difference calculation with difference time Δtm of 5 seconds in the method of the present invention (first invention) D: Method of the present invention (second invention) After the noise removal of the weight signal by the Kalman filter in), the measurement result with a difference time Δtm of 1 second

Claims (2)

【特許請求の範囲】[Claims] 【請求項1】 表層タンディッシュおよび内層タンディ
ッシュから異なる金属組成の溶湯を、それぞれ注入ノズ
ルを介してモールド内に注入し、電磁ブレーキにより該
モールド内の溶湯に制動力を与え、該モールド内に形成
される境界層によって表層溶湯および内層溶湯を上下に
分離させて複層鋳片を鋳造する連続鋳造において、溶湯
重量時系列データの差分演算により溶湯流量を求めるに
あたり、現時点から過去に向かって一定時間幅内の時系
列データの重心と差分時間だけ過去を起点とする時系列
データの重心とを求め、該2つの重心間で差分演算を行
うことにより、溶湯流量を求めることを特徴とする複層
鋳造プロセスにおける溶湯流量測定方法。
1. A molten metal having a different metal composition from the surface tundish and the inner layer tundish is injected into a mold through an injection nozzle, and a braking force is applied to the molten metal in the mold by an electromagnetic brake, and the molten metal is injected into the mold. In continuous casting in which the surface molten metal and the inner molten metal are separated into upper and lower layers by the formed boundary layer, and the molten metal flow rate is calculated by the difference calculation of the molten metal time series data, it is constant from the present time to the past. It is characterized in that the center of gravity of the time series data within the time width and the center of gravity of the time series data starting from the past by the difference time are obtained, and the difference between the two centers of gravity is calculated to obtain the molten metal flow rate. Method for measuring melt flow rate in layer casting process.
【請求項2】 表層タンディッシュおよび内層タンディ
ッシュから異なる金属組成の溶湯を、それぞれ注入ノズ
ルを介してモールド内に注入し、電磁ブレーキにより該
モールド内の溶湯に制動力を与え、該モールド内に形成
される境界層によって表層溶湯および内層溶湯を上下に
分離させて複層鋳片を鋳造する連続鋳造において、溶湯
重量時系列データの差分演算により溶湯流量を求めるに
あたり、事前にノイズ除去フィルターで重量信号のノイ
ズ除去を行うことにより、流量測定値の応答遅れを小さ
く抑えることを特徴とする複層鋳造プロセスにおける溶
湯流量測定方法。
2. A molten metal having a different metal composition from the surface tundish and the inner layer tundish is injected into a mold through an injection nozzle, and a braking force is applied to the molten metal in the mold by an electromagnetic brake so that the molten metal is injected into the mold. In continuous casting, in which the surface molten metal and the inner molten metal are separated into upper and lower layers by the formed boundary layer to cast a multi-layer slab, the weight of the noise removal filter is used in advance to determine the molten metal flow rate by the difference calculation of the molten metal time series data. A method for measuring a molten metal flow rate in a multi-layer casting process, characterized in that a response delay of a flow rate measurement value is suppressed to a small level by removing signal noise.
JP10705693A 1993-05-07 1993-05-07 Method for measuring flow rate of molten metal in double layer casting process Pending JPH06315746A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP10705693A JPH06315746A (en) 1993-05-07 1993-05-07 Method for measuring flow rate of molten metal in double layer casting process

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP10705693A JPH06315746A (en) 1993-05-07 1993-05-07 Method for measuring flow rate of molten metal in double layer casting process

Publications (1)

Publication Number Publication Date
JPH06315746A true JPH06315746A (en) 1994-11-15

Family

ID=14449388

Family Applications (1)

Application Number Title Priority Date Filing Date
JP10705693A Pending JPH06315746A (en) 1993-05-07 1993-05-07 Method for measuring flow rate of molten metal in double layer casting process

Country Status (1)

Country Link
JP (1) JPH06315746A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104596622A (en) * 2014-12-15 2015-05-06 中冶长天国际工程有限责任公司 Method and device for calibrating dust removal quantity of electric precipitator

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS61162754A (en) * 1985-01-14 1986-07-23 Oki Electric Ind Co Ltd Speed detection system for moving object
JPH01317668A (en) * 1988-03-15 1989-12-22 Sumitomo Metal Ind Ltd Method for controlling molten metal surface in continuous casting mold
JPH04105759A (en) * 1990-08-27 1992-04-07 Nippon Steel Corp Method for controlling molten steel surface level in continuous casting for complex cast billet
JPH05104223A (en) * 1991-09-18 1993-04-27 Nippon Steel Corp Method for controlling pouring rate in continuous casting for plural layer cast slab

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS61162754A (en) * 1985-01-14 1986-07-23 Oki Electric Ind Co Ltd Speed detection system for moving object
JPH01317668A (en) * 1988-03-15 1989-12-22 Sumitomo Metal Ind Ltd Method for controlling molten metal surface in continuous casting mold
JPH04105759A (en) * 1990-08-27 1992-04-07 Nippon Steel Corp Method for controlling molten steel surface level in continuous casting for complex cast billet
JPH05104223A (en) * 1991-09-18 1993-04-27 Nippon Steel Corp Method for controlling pouring rate in continuous casting for plural layer cast slab

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104596622A (en) * 2014-12-15 2015-05-06 中冶长天国际工程有限责任公司 Method and device for calibrating dust removal quantity of electric precipitator

Similar Documents

Publication Publication Date Title
CA2191242C (en) Quality prediction and quality control of continuously-cast steel
JPH06315746A (en) Method for measuring flow rate of molten metal in double layer casting process
JPH06264B2 (en) Level control method in continuous casting
JP5958311B2 (en) Hot water level control device, method and program for continuous casting machine
JP5831145B2 (en) Continuous casting machine mold level control device and control method
JP2634106B2 (en) Metal surface level control method in continuous casting
JP2664572B2 (en) Temperature prediction method for unsolidified part of slab in continuous casting
JPH0722812B2 (en) Method and apparatus for controlling molten metal level in continuous casting
JP2001259814A (en) Method for controlling molten metal surface level in mold for continuous caster
JPH0531560A (en) Method for controlling molten metal surface level in continuous casting
JPS6045026B2 (en) Mold content steel level control method
JPH0587351B2 (en)
JPH04105759A (en) Method for controlling molten steel surface level in continuous casting for complex cast billet
JPH05228580A (en) Continuous casting method
JP2668872B2 (en) Breakout prediction method in continuous casting.
JPH05177321A (en) Device for controlling mold level
CN112139468B (en) Method and device for detecting liquid level of strip continuous casting distribution bag
Chen et al. Investigating centerline bridging in continuous casting during speed drops with conoffline
JPH05189009A (en) Controller
JP7073932B2 (en) Solidification completion position detection method for slabs and solidification completion position detection device for slabs
JP2984171B2 (en) Mold level control device
JPH07178525A (en) Method for controlling pouring quantity of molten steel for surface layer in continuous casting of double layered steel plate
JP2019209366A (en) Method and apparatus for controlling bilayer-billet continuous-casting process, and program
JP2023032229A (en) Cast iron solidification latent heat calculation method and solidification latent heat calculation device
Lagerstedt On the shrinkage of metals and its effect in solidification processing

Legal Events

Date Code Title Description
A02 Decision of refusal

Free format text: JAPANESE INTERMEDIATE CODE: A02

Effective date: 19990330