JPS63210216A - Method for prediction of furnace heat in blast furnace - Google Patents

Abstract

PURPOSE:To accurately predict furnace heat, especially lowering of furnace hat in a blast furnace without delay by adopting ARMA model introducing a variable related to temp. difference at inner wall in the blast furnace and the max. value of solution loss C quantity, etc., as a variable into MA term. CONSTITUTION:At the time of predicting variation with time of the furnace heat level by ARMA model using the operational result in the blast furnace, the above two kinds of the variables are used into the MA term. That is, the furnace heat is predicted by adopting A: the variable related to the temp. difference at the inner wall measured in every the prescribed time intervals by inner wall thermometers set at the prescribed positions in the blast furnace and B: the max. value of the above C quantity during the prescribed time or the min. value of N content in the top furnace gas component during the prescribed time as the variable into the MA term. Further, it is desirable that the above ARMA model is different to each iron tapping hole.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention relates to a blast furnace furnace heat prediction method for stable operation of a blast furnace.

(Prior art and its problems) It has been known for a long time that in order to maintain stable operation of a blast furnace, it is necessary to keep the temperature of hot metal constant. For this reason, blast furnace operators have always needed to predict changes in blast furnace heat.

Prediction of temperature drop is extremely important in blast furnace heat changes, especially since hot metal may solidify due to temperature drop and may no longer flow out of the blast furnace.

A method for predicting blast furnace heat is to consider the blast furnace as a black box and obtain a model through statistical analysis.
There is one based on the AR (^uto-Regressive) (auto-regressive) model.

Although the AR model described above has excellent characteristics as a model for predicting temporal fluctuations in the actual furnace heat level of a blast furnace, it has the following problems.

FIG. 17 is a diagram showing changes in the actual value and predicted value of hot metal temperature over time, and in the figure, 11 is a curve based on the actual value x 112 is the predicted value. As shown in the figure, the change in the predicted value tends to lag behind the change in the actual value by about τ sampling time (that is, τXΔ is time, where the sampling interval is Δt), and its amplitude tends to be small. For this reason, there was a problem in that the prediction function could not be fully fulfilled, and the accuracy of the predicted value was also not sufficiently accurate.

In order to solve this problem, ARM A (Auto-
Rearessive HOVingAVeraQO)
(autoregressive moving average) model is
0-248804゜"Tetsu to Hagane J1984, No. 1, 5
Although it is disclosed on page 4, etc., the current situation is that no definitive variable to be introduced into the MA term has yet been found.

(Objective of the Invention) The object of the present invention is to provide a blast furnace furnace heat prediction method that eliminates the problems of the above-mentioned prior art and can accurately predict hot metal temperature changes without delay in prediction. .

(Means for Achieving the Object) In order to achieve the above object, the blast furnace furnace heat prediction method according to the present invention uses the blast furnace operation results to predict temporal fluctuations in the blast furnace furnace heat level using the ARMA model. Variables related to the inner wall temperature difference measured at predetermined time intervals by inner wall thermometers installed at predetermined locations in the blast furnace, and the maximum value of the amount of solution loss carbon in a predetermined time width or the amount of nitrogen in the top gas component over a predetermined time. The minimum value in the width is introduced into the MA term.

(Example) 1. First variable introduced into the MA term The following can be considered as the first reason for the decrease in the furnace heat of the blast furnace.

High-temperature air (gas flow) blown up from the blast furnace lining to adjust the temperature and amount of hot metal is usually blown into the center of the furnace. However, due to reasons such as raw material charging conditions and charge distribution, a large amount of gas may suddenly flow to the periphery of the furnace.

As a result, Fe　+C→Fe　+C’0 The endothermic reaction of is promoted and the furnace heat decreases.

By the way, when a large amount of gas flows around the furnace, Na
, when the deposits in the furnace such as Pb and the stagnation layer peel off and fall off the wall, the temperature of the furnace wall in that part increases rapidly. After that, the furnace heat will decrease sooner or later.

Moreover, the following can be considered as the second reason for the decrease in furnace heat of the blast furnace.

When the unloading speed in the blast furnace increases due to raw material charging conditions, charge distribution, etc., the level of the cohesive zone in the blast furnace decreases due to so-called raw ore unloading, resulting in a decrease in furnace heat.

By the way, when the cohesive zone level decreases, the furnace wall temperature at the corresponding portion also decreases rapidly. After that, the furnace heat will decrease sooner or later.

For the above two reasons, it can be said that the difference in furnace wall temperature has a predictive value on the temperature change in the blast furnace. Therefore, an ARMA model can be considered in which a variable related to the furnace wall temperature difference is introduced into the MA term.

FIGS. 1(a) and 1(b) are a side sectional view and a plan sectional view showing the arrangement of the inner wall temperature 1 used in the present invention and the embodiment, respectively. As shown in Figure (a), there are 7 inner wall thermometers in the height direction of the blast furnace 1 (3 on the back).

(2 abdominal parts, 2 morning glory parts), and 4 pieces are installed in the circumferential direction of the blast furnace 1 as shown in FIG. In other words, a total of 28 inner wall thermometers 3 are installed in four directions and seven levels.

The inner wall thermometer is, for example, disclosed in Japanese Utility Model Application No. 57-81 by the present applicant.
531. The one disclosed in Japanese Utility Model Publication No. 59-16816 may be used, and FIG. 2 is a conceptual diagram showing an inner wall thermometer (hereinafter referred to as "rFM sensor") disclosed in the latter.

In the figure, reference numeral 4 denotes a sheath type thermometer having two conductive wires 5 buried in parallel insulatively and having a temperature sensing part 6 on the front end side. Place the book on each thermosensor 6.
are arranged in parallel so that they are placed at different locations in the length direction, and a sheath type dummy rod 7 is connected to the tip of the temperature sensing section 6 to align the leading ends. The sheath type dummy rod 7 has two conductive wires 5 buried in parallel insulating manner, and has substantially the same thermal conductivity as the sheath type temperature measuring element 4. The FM sensor 3 is re-formed by keeping the sheath type temperature measuring body 4 in a non-contact manner with the insulation i8 and housing it in the sheath tube 9.

FIG. 3 is an explanatory diagram of the installation of the FM sensor 3. In the figure, 10 to 13 are the walls of the blast furnace, 10 is a brick, and 1
1 is a stave, 12 is a stamp, and 13 is an iron skin. F
As shown in the figure, the M sensor 3 is installed inside the furnace wall by welding to a packing 14 and a welded portion 15. Note that 16 is a filler, and 17 is a milk injection 1'' where the filler 16 is poured.

Note that due to the installation and structure of the FM sensor 3 described here, the FM sensor 3 itself is eroded along with the erosion of the furnace wall, and the sheath type temperature sensing element 4 may be exposed inside the furnace near the furnace wall. In addition to "furnace wall temperature", it also includes "furnace temperature near the furnace wall".
is being measured. Hereinafter, a concept including both will be described as "furnace wall temperature." As mentioned above, compared to conventional thermometers such as sheathed thermocouples, the FM sensor 3 has a large number of measurement points, satisfies rapid temperature measurement response, is capable of long-term continuous temperature measurement, and is highly reliable. The aim is to improve performance, durability, and workability.

Each FM sensor 3 measures the inner wall temperature of the blast furnace 1 at every predetermined sampling time Δt, as shown in FIG. Here, the inner wall temperature of the first FM sensor 3 at time j is Tj
, , and if one sampling time Δ at time j is the previous inner wall temperature, then J-1, I T... and T, ・Which inner wall temperature difference (difference value) ΔTJ,
I J-1,1... is J,1 8T...=T,...-' j-1,i...(
1) J, I J, 1. This state is shown in FIG.

This difference value ΔTj, . Furthermore, the difference value 6
Ding 1. For those where is negative, V.

J, 1 10, and other values are also multiplied by a positive/negative parameter V· indicating v,=1 to obtain a corrected difference ■ value (positive difference value) CTj, at time j.

CT,,=W,-V,・ΔT,,・(2)J,
I I I J, primary, corrected difference value CTj, , to all FM sensors 3, and next (
According to formula 4), if the value of the total difference value 5T1j becomes larger than the predetermined threshold value ε1, it can be assumed that there has been a rapid temperature rise of the furnace wall due to wall fall of deposits in the furnace, and this event can be expressed numerically. By doing so, it can be introduced as a variable in the MA term.

ST1・≧81 (4) Also, in equation (2), the positive and negative parameters V are the difference values Δ
For those where T... is positive, set v, = O1J, 1 For other cases, set v, = 1, and then calculate the absolute value of all FM sensors 3 of the corrected difference value CTj, . Take the total sum for 5T2j.

ST2. =ΣlcT,,l,...
(3).

J Shrine J,1 Then, according to the following equation (4), if the value of the sum of the difference values 5T2j based on the equation (3) is larger than the predetermined threshold ε2, it is determined that there has been a rapid temperature drop due to the descent of raw ore. By quantifying this phenomenon, it can be introduced as a variable into the MA term.

ST2・≧ε2　　　　　　　　　　...(4).

If an event that satisfies formula (4) (formula (4)' is also acceptable) exists in the past by a time width NΔ from the current time t, then F
It is conceivable to introduce a variable F1 (t), which sets (t)=1 and, if it does not exist, Fl (t)=-1, as an MA term.

In this case, predicted hot metal temperature value y(t+Δt2) at one point in time
can be estimated using the following equation.

'Sl' (t+Δt2) =Ia1. y (t −iΔt2)+b1F1 (t)
A=ρ (5) The first term on the right side is the AR term, and the second term is the MA term.

Note that all and bl are coefficients.

Figures 6 and 7 show examples of prediction using equation (5), where (a) is the hot metal temperature and (b) is the total positive difference value 5T1j
, (c) is a graph showing changes in Fl (t) over time.

Also, from the current time t, the time width NΔ is moved backwards into the past, (
4) Variable F2(t
) may be introduced as an MA term as shown in the following equation (6).

V(t+Δt2) = Σa, y(t-iΔt2)+b2F2 < t
),,, 2+...(6) In addition, a21. b2 is a coefficient.

Figures 8 and 9 show examples of prediction using equation (6), where (a) is the hot metal temperature and (b) is the positive difference value sum ST1.
, (C) is a graph showing changes in F2(t) over time.

It has been known that the increase or decrease in the amount of solution loss carbon (hereinafter referred to as "Tru loss CIJ"), which indicates the quality of the reduction state of the blast furnace, has a strong correlation with the temperature of the blast furnace heat. ing. In other words, an increase in the amount of tsuru loss C indicates that the so-called tsuru loss reaction shown below is promoted.C1002→2GO This reaction is an endothermic reaction. The fever will drop.

Therefore, it can be said that an increase or decrease in the amount of truss loss C can be predicted with respect to changes in blast furnace heat.

Therefore, AR that introduces the maximum value of the amount of truss loss C into the MA term
An MA model can be considered.

FIGS. 10(a) and 10(b) show changes over time in the instantaneous value of the hot metal temperature and the average value of Sol loss C1, respectively. In the same figure (a), Ll is an actual value, and F2 is a predicted value. The instantaneous value of the loss (0il (K9/l-1)) is calculated at each sampling time of gas chromatography based on the proportions of Co, CO2, N2, etc. in the top gas components, ventilation conditions, raw material charging conditions, etc. is calculated, and with reference to FIG. 10(b),
The maximum value x(t) of the amount of vine loss C is determined by the average value Δ1 of the amount of vine loss C for each section Δt1 within the predetermined time width NΔt1.
1) This is calculated as the maximum value A of (k-1 to N) every sampling time Δt2. On the other hand, the instantaneous value y(t) of the hot metal temperature is determined at every sampling time Δt2. Note that t is the current time.

Using such variables, the predicted hot metal temperature value V(t+Δ12) at one point in time can be estimated using the following equation using the ARMA model.

'Si' (ten Δj2) -Σ a-y (t-iΔt2)+cx (t)il, 1 ...(7) The first term on the right side is the AR term, and the second term is the MA term.

Note that a. and C are coefficients.

1 [1, Two variables to be introduced into the MA term The variables described in ``e, n'' can be expected to have a considerable effect even if each of them is introduced alone into the MA term, but in order to further improve prediction accuracy, It is conceivable to introduce two variables simultaneously as MA terms.

In this case, the predicted value V(t+Δ12) of the hot metal temperature one point in time can be estimated by the following equation.

V (t+Δt2) 10b F (t) + C1X (j)...
(8) V(t+Δj2) = Σ ay(t−iΔt2) ,,, 2+ 10b F (t)+c2x (t)...(9) In addition, a, l), CI, a2H, b2
, C2 is the coefficient 1i 1 number.

FIG. 11 is a flowchart showing a prediction processing procedure based on equation (8). In step S1, the positive difference sum 5TIj is calculated using equations (1) to (3), and then in step S2
Step back up the time width NΔ from the current time t to the past, (4)
Determine whether the formula is satisfied. If equation (4) is satisfied even once, set Fl (t)-1 in step S3; if equation (4) is not satisfied even once, step S
4, let Fl (t)=-1.

Then, in step S5, a time period NΔt from the current time t
1. Going backwards in the past, calculate N average values u (Δt1 at t-) (k=1 to N) for each section Δt1 of the true loss CI. Next, in step S6, the maximum value of the average value u (Δt1 for t-) obtained in step S5 is obtained from the equation (10) below, and this is set as xB).

x (t)=max (u (Δt1 to t-))k-1
,...,N...(10) Note that steps 81 to 84 described above. The processes of steps 85 to S6 are performed every sampling time Δt2.

Then, in step S7, the coefficient a11°blCl of equation (8) is determined from the actual hot metal temperature value y(t) at the current time t by comparing it with the predicted result V(1). Next, in step S8, the predicted value V(t+Δt2) of the hot metal temperature one point ahead is calculated using the coefficient a11. calculated in step S7 described above. The blast furnace furnace heat is predicted by calculating from equation (8) using b1C1 and outputting this predicted value 'Sl' (t+Δt2) in the next step S9. From then on, step S
Returning to step 1, the furnace heat is predicted while repeating steps 81 to S9.

FIG. 12 is a flowchart showing a prediction processing procedure based on equation (9). In the same figure, in step 811, the sum of positive difference values ST1 is calculated using equations (1) to (3), and in step 812, it is determined whether equation (4) is satisfied during the period NΔt1 from the current time t. judge. (
4) If the formula is satisfied, in step S13, ST
1, if the number of times ε1 occurs is p, then F2 (t
)=p. On the other hand, if equation (4) is not satisfied even once, in step 814, F2(t)=O is set,
Thereafter, steps 815 to S19 are step 85 in FIG.
~ The same processing as S9 is performed.

In this way, by introducing variables related to the furnace wall temperature mark difference (difference value) and the average value of the truss loss C amount into the MA term, which are highly predictive of changes in furnace heat, more accurate furnace heat can be predicted.

■Consideration of temperature deviation due to tapholes. Normally, blast furnaces alternately tap two tapholes, so for different tapholes A and B, the hot metal temperature for each tap as shown in Deviations often occur in the hot metal temperature as shown in changes over time, and in such cases, prediction accuracy is affected.

Therefore, AR using formulas (8) and (9) for each taphole A and B is
One possible method is to set up an MA model and predict the hot metal temperature. If the taping at one point in time is through taphole A, then using the temperature yA(1) of the hot metal tapped from taphole A, yA (t+Δt3) - Σa-y(t-iΔt3) −＾ +A +bA Fl (t)+cA x (t) −
The hot metal temperature is estimated using (11). On the other hand, in the case of tapping through taphole B, the temperature of hot metal tapped from taphole B is y,
Using (t), estimate the hot metal temperature using V, (t+Δt3) = Σa=y(tiΔt3), B+ 8 +bBF1 (t)+cBx (t) (12). In addition, aAi, bA.

CA, aBi, b8. CB is a coefficient, and Δt3 is the sampling time for each tapping date.

FIG. 14 is a flowchart showing a prediction processing procedure using an ARMA model by tapping date in which variables F, (t) are introduced into the MA term. In the figure, in steps 321 to S24,
The variable F1 (t) is found in the same way as above. Next, in steps S25 and S26, the maximum value x (t) of the amount of trail loss C is determined.

Then, in step 827, the taphole that is one point ahead is distinguished, and if the taphole that is one point ahead is taphole A, step 828. In S29, the predicted hot metal temperature value 'Sl'A (t+Δt3) of the taphole A is determined based on equation (11), and after outputting VA (t+Δt3) in step S30, the process returns to step 821.

On the other hand, in step 827, if the next tap is taphole B, in steps 831 and 832, the predicted hot metal temperature 'Sl'B (t+Δt3) is calculated based on equation (12), and in step S33, the prediction value is calculated. The value VB (+Δt3) is output, and the process returns to step S21.

It goes without saying that the ARMA model by tapping date in which the variable F2(t) is introduced into the MA term can also be realized by setting up the ARMA model based on equation (6) for each tapping date, and the processing means are shown in Fig. 15. This is shown in the flowchart below. Step 8
Steps 41 to 852 correspond to steps 321 to 832 described above.

By providing an ARMA model for each tap date in this manner, the influence of hot metal temperature deviation due to the difference in the tap hole can be removed, and the prediction accuracy is considerably improved.

■, Supplementary note: In this example, we have described an example in which the event that satisfies equation (4) is digitized and introduced into the MA term, but the same effect can be obtained by digitizing the event that satisfies equation (4) and introducing it into the MA term. It has the effect of

In addition, in this example, an FM sensor was used as the inner wall thermometer, but a normal temperature sensor (for example, a sheathed thermocouple) can be used instead, although there is a problem with the service life. Although the prediction accuracy will be slightly lower due to its reliability and low temperature response, it can be used as a substitute.

Further, in this embodiment, 28 FM sensors 3 are installed in 7 levels and in 4 directions, but it goes without saying that they may be installed appropriately depending on the characteristics of the blast furnace.

Furthermore, when calculating the average of the amount of trailing loss C for each section Δt1, the instantaneous value of the amount of trailing loss C may generate an abnormal value E1°E2 due to noise or the like, as shown in FIG. 16(a). Here, if the amount of vine loss C at time j is X・, and the amount of vine loss C before one sampling time △ is xj-1, then the absolute value of the difference value of the amount of vine loss C △XJ is Δx, = lx = −x ・l
=・(13) J J J−1. By comparing this ΔX with the threshold value ε7 as shown in FIG. 3(b), the abnormal value E1. Find E2 and draw the same figure (C
), a method of smoothing by replacing it with the previous value can be considered. By applying this method, a more accurate maximum value of the 0s C amount can be determined, and as a result, even more accurate prediction becomes possible.

In this example, the maximum value of the average values for each section Δt1 was determined as the maximum value of the amount of trail loss C within the predetermined period NΔt1, but the maximum value of the instantaneous value of the amount of trail loss C within the predetermined period NΔt1 was simply calculated. May be used. However, since it cannot be denied that it is susceptible to the influence of noise components, it is preferable to perform the above-mentioned noise removal process at the same time.

In addition, the amount of nitrogen (%) in the top gas detected by gas chromatography (hereinafter referred to as "gas chromatography N2") has a strong negative correlation with the amount of tsuru loss C, and the minimum amount of gas chromatography A similar effect can be expected by using the MA term instead of the maximum value of the truss loss C amount.

(Effects of the Invention) As explained above, according to the present invention, ARMA is performed using a variable related to the temperature difference on the inner wall of the blast furnace in the MA term, and the maximum value of the amount of turquoise C or the minimum value of the amount of gas chromatin N2 as a variable.
The model makes it possible to predict blast furnace furnace heat, especially furnace heat decline, without delay and accurately.

[Brief explanation of the drawing]

FIGS. 1(a) and (b) are a side sectional view and a plan sectional view showing the arrangement of an FM sensor used in an embodiment of the present invention, FIGS. 2 and 3 are conceptual diagrams of the FM sensor, respectively, 4 is a graph showing changes in inner wall temperature over time, FIG. 5 is a graph showing changes in inner wall temperature difference over time, and FIGS. 6 to 9 each show a prediction example in an embodiment of the present invention. show,
(a) is the hot metal temperature, (b) is the sum of positive difference values, and (C) is a graph showing the change over time of variables related to the inner wall temperature difference.
Figures 0 (a) and (b) are graphs showing changes over time in the instantaneous value of the hot metal temperature and the average value of the amount of trunnion C, respectively, and Figures 11 and 12 are the processing steps of the prediction method in the embodiment of the present invention, respectively. A flowchart showing the process, Figure 13 shows the
- A graph showing the change in hot metal temperature over time when there is a difference in hot metal temperature due to a difference in the mouth, and Figures 14 and 15 respectively show the processing procedure of the ARMA method by tapping day, which are other embodiments of the present invention. Flowchart showing FIG. 16 (a), (b)
, (c) is a graph showing the instantaneous value of the amount of vine loss C including abnormal values, the absolute value of the difference value of the amount of vine loss C, and the instantaneous value of the amount of vine loss C after removing abnormal values, and Fig. 17 is a graph showing the instantaneous value of the amount of vine loss C including abnormal values. 2 is a graph showing prediction results based on actual hot metal temperature values based on the method.

Claims (2)

[Claims] (1) When using the blast furnace operation results to predict temporal fluctuations in the blast furnace heat level using the ARMA model, the inner wall temperature difference measured at predetermined time intervals by inner wall thermometers installed at predetermined locations in the blast furnace. A method for predicting heat in a blast furnace, characterized in that a variable relating to the amount of solution loss carbon and a maximum value of the amount of solution loss carbon in a predetermined time width or a minimum value of the amount of nitrogen in the furnace top gas component in a predetermined time width are introduced into the MA term. (2) The blast furnace furnace heat prediction method according to claim 1, wherein the ARMA model is different depending on the tap hole.