JPS63210223A - Predicting on of lowering of furnace heat of blast furnace - Google Patents

Abstract

PURPOSE:To quickly and accurately execute the prediction, by prediction while timely changing threshold value for five predicting means giving evaluated point for the prescribed time, etc., when summary value of positive part in temp. difference at an inner wall in every the prescribed time intervals exceeds the threshold value. CONSTITUTION:The temp. difference at the inner wall in a blast furnace is measured in every the prescribed time intervals. One means or more among five means, such as the first means giving the evaluated point for the prescribed time when the summary value of the positive parts in the temp. difference exceeds the threshold value, the second means based on the negative parts as the same manner as the above, the third means giving the evaluated point for the prescribed time when sum total of running average value of the above positive parts during the prescribed time exceeds the threshold value, and the fourth means and the fifth means based on respectively solution loss carbon quantity and nitrogen content in the top furnace gas, are provided. And, the threshold value of the above means is changed in every moment based on the prescribed value gone back to the past.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Field of Application) The present invention relates to a method for predicting heat drop in a blast furnace 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 furnace thermal changes, especially since hot metal may solidify due to temperature drop and may no longer flow out of the blast furnace.

As a method for predicting blast furnace furnace heat, Japanese Patent Application Laid-Open No. 60-39107
There are some that have been disclosed. This method is based on the viewpoint that the charge temperature around the furnace belly has a strong correlation with the hot metal temperature.
As shown in FIG. 14, the relationship between the temperature around the furnace belly area detected by the sensor (belly sonde) 2 installed in the blast furnace 1 and the hot metal temperature is subjected to linear regression as shown in FIG. 15. Based on this linear equation, from the temperature around the furnace belly to the hot metal temperature, 1.
It predicts.

However, with this method, it is necessary to insert the furnace probe 2 in order to measure the temperature near the inner wall of the furnace, which means that temperature measurements can only be carried out intermittently, which naturally deteriorates the accuracy of hot metal temperature prediction. There was a problem.

Further, even if the hot metal temperature is the same, the furnace temperature may change due to changes in the production plan, raw material charging conditions, etc. Therefore, the linear equation based on the absolute value of the furnace wall shown in FIG. 15 has the problem that accurate prediction cannot necessarily be made.

On the other hand, another prediction method has conventionally been used to predict blast furnace heat based on the increase or decrease in the amount of solution loss carbon (hereinafter referred to as "Tsuru loss Cff1J"), which indicates the quality of the reduction state of the blast furnace. shows that the so-called Truroth reaction shown below is promoted.

C+002-+2GO: Since this reaction is an endothermic reaction, it can be predicted that the blast furnace heat will decrease.

The amount of turret loss C is determined by adjusting the proportions of CO, CO2, N2, etc. in the top gas, air blowing conditions, and raw material charging conditions every analysis cycle (about 3 degrees) of gas chromatography, which usually analyzes the top gas composition. Traditionally, it is calculated based on the hourly tsuru loss C.
The decrease in furnace heat was controlled by the average value of the amount.

In FIGS. 16(a) and (b), FIG. 16(a) shows the change over time of the crane loss Cf1 (11) every 3 minutes and the average value of the crane loss C amount (12) every hour; b) is a graph showing changes in hot metal temperature over time. In the figure, although the threshold value ε1 is exceeded at time 17, no action is taken to raise the temperature, and thereafter the hot metal temperature drops significantly as shown in (■). N
FIG. 17 shows the respective changes over time when the heating action A was initiated at time 13:00 when the threshold value ε2 was exceeded. Note that, similarly to FIG. 16, 11 in the figure indicates an instantaneous value every 3 minutes, and 2 indicates an hourly average trail loss Cff1.

By comparing Fig. 17, it can be seen that due to heating action A, the temperature of hot metal decreases as shown in ■ in Fig. 16(b).
As shown in the figure, it can be seen that this can be avoided to some extent.

However, the problem with predicting the average value of the amount of vine loss C every hour is that when there is a sudden increase in the amount of vine loss C, in the worst case, there is a delay in predicting the decrease in furnace heat by about an hour. There was a point. For example, in the case of Figure 17,
Due to the prediction delay, heating action A is taken because the melt R temperature is the control temperature T.

It has become a problem since it has fallen below a certain level.

Therefore, in order to avoid this problem, when predicting the decrease in furnace heat using the instantaneous value of the amount of truss loss C measured at intervals of about 3 degrees, the individual variations as shown in 11 in Figs. 16 and 17 are reduced. Because the noise component is large, the data is not sustainable. Therefore, it is almost impossible to predict the decrease in furnace heat using the instantaneous value of the truss loss C amount.

(Object of the Invention) The object of the present invention is to provide a blast furnace furnace heat drop prediction method that can solve the problems of the prior art described above, obtain predictions as quickly as possible, and accurately predict the drop in hot metal temperature. It is to be.

(Means for Achieving the Object) In order to achieve the above object, the blast furnace heat drop prediction method according to the present invention includes installing an inner wall thermometer at a predetermined location of the blast furnace,
A first method that measures the inner wall temperature difference at predetermined time intervals using the inner wall thermometer, and gives an evaluation point for a predetermined period when the total value of the portion showing a positive value of the inner wall humidity difference at a certain time exceeds a threshold value. a second prediction means for giving an evaluation point for a predetermined period when the total value of the portion showing a negative value of the inner wall temperature difference at a certain time exceeds a threshold; a third prediction means that gives an evaluation point for a predetermined period when the sum of the moving average values of the time width of the portion showing a positive value exceeds a threshold; a fourth prediction means that gives an evaluation point when the moving average value in a predetermined time width exceeds a threshold value; and a fifth prediction means for giving an evaluation point when the moving average value is less than a threshold value, and according to the comprehensive evaluation of the evaluation points of at least one of the first to fifth prediction means. When predicting a decrease in furnace heat, the threshold values of the first to third prediction means are determined based on the threshold value of each corresponding response during a predetermined time period extending backward from the time when a decrease in furnace heat was observed within a certain period of time in the past. The fourth. The threshold value of the fifth prediction means is configured to change from time to time based on the average value and standard deviation of the solution loss carbon amount or the nitrogen amount for each response within a certain past period.

(Example) C. First reason for the decrease in furnace heat The following factors can be considered as a cause of 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　+CO The endothermic reaction of is promoted and the furnace heat decreases.

By the way, when the gas flow 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. If this rapid temperature rise is detected, a decrease in furnace heat can be predicted.

B. Second Reason for Decrease in Furnace Heat The following may be considered as a cause of the decrease in furnace heat in the blast furnace.

If the unloading speed in the blast furnace increases for the same reason as Nyu,
The level of the cohesive zone in the blast furnace decreases due to so-called raw ore descent, causing 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. If this rapid temperature drop is detected, a decrease in furnace heat can be predicted.

C6 Third Reason for Decrease in Furnace Heat Furthermore, the following may be considered as a contributing factor to the decrease in furnace heat in 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, for reasons similar to those related to flies and specific gravity, a portion of the gas flow may flow to the periphery of the furnace. If this state continues for a long time, heat dissipation from the gas flow from the blast furnace wall becomes greater than during normal operation, resulting in a decrease in furnace heat.

By the way, when a part of the gas flow steadily flows into the periphery of the furnace, the temperature of the furnace wall gradually increases. If such a gradual temperature rise over a relatively long period of time is detected, a decrease in furnace heat can be predicted.

D. First to Third Predicting Means FIGS. 1(a) and 1(b) are a side sectional view and a plan sectional view, respectively, showing the arrangement of an inner wall thermometer used in an embodiment of the present invention. As shown in the figure (a), the inner wall thermometer 3
7 pieces in the height direction of blast furnace 1 (3 pieces 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 the 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. Temperature sensitive part 6. are arranged in parallel so that they are placed at different locations in the length direction, and a roughly sheathed dummy rod 7 is connected to the tip of the temperature sensing section 6 to align the leading ends. Sheath type dummy rod 7 is 2
Two conductive wires 5 are insulated and buried in parallel, and have substantially the same thermal conductivity as the sheath type temperature sensing element 4. FM sensor 3
is formed by keeping the sheath type temperature measuring element 4 in a non-contact manner with an insulating material 8 and storing it in a 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 port into which 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", "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 i-th FM sensor 3 at time j is T.
・If closing, 1 sample J at time j, 1 ring time Δ, and the previous inner wall temperature T, then J-
1, I T1. The inner wall temperature difference (difference value) between and T
l J-1,1... becomes J,1...(1) 8T...-"j,i"j-1,iJ,1. This state is shown in FIG.

This difference value ΔTj,i is multiplied by a weight Wi in consideration of the height of each FM sensor 3 in two circumferential directions, etc. Roughly speaking, the difference value Δ
For Tj, where i is negative, vi = 01; for other cases, multiply by the positive/negative parameter V indicating v, = 1, and calculate the corrected difference value at time j (positive difference value) CTj , , is obtained.

CT・・−W・・V・・ΔT・・・・・(2)
J, l l l J, primary, corrected difference value CTj, , for all FM sensors 3, and according to the following equation (4), the value of this difference value sum 5T1j becomes larger than a predetermined threshold value ε1. , it is assumed that there has been a sudden temperature rise, and an evaluation point of 1 is given for the predetermined period.

ST1·≧81 (4) The above is the first prediction means based on the reason for N1.

The second prediction means based on the reason B is as shown below.

In equation (2), the positive and negative parameters Vi are the difference values ΔK 1
．． For those where is positive, v,=O, that J, 1
For values other than 1, set v to -1, then corrected difference value CTj
For all FM sensors 3 of the absolute values of It is assumed that there has been a sudden temperature drop, and an evaluation point of 1 is given for the predetermined period.

ST2・≧82　　　　　　　　　　...(4).

The third prediction means based on the reason for C0 is as shown below.

As with the first prediction means, the positive/negative parameter V in equation (2) is -01 for negative difference values ΔTj, .
For other cases, it is set as ■i-1. Also time j
The corrected difference value before the sampling time (that is, before ΔtXk time) is set to CT j−.

Then, for all FM sensors 3 of the value at time j of the moving average of the predetermined time width nΔt of this corrected difference value, and the next (4
)", if the value of this moving average sum 5T3j becomes larger than a predetermined threshold value ε3, it is assumed that there has been a gradual increase in humidity for a long period of time, and an evaluation point of 1 is given for the predetermined period.

ST3・≧83...(4)"Since the above-mentioned first to third prediction means are each performed based on the furnace wall temperature difference (difference value), Accurate predictions can be made.Moreover, F
The M sensor 3 can be arranged to cover the entire circumference of the blast furnace due to its good workability and good temperature measurement response, and by being able to grasp continuous inner wall temperature differences, more accurate predictions can be made.

Further, the first to third prediction means described above can be realized by a computer. FIG. 6 is a flowchart showing the processing flow of the first prediction means. In the same figure,
In step S1, the furnace wall temperature T· of each FM sensor 3 is measured every J,1 sampling time Δ. Next, in step S2, the difference value of each FM sensor 3 is calculated based on equation (1).

Then, in step S3, (2), (calculates the total positive difference value 5T1j based on equation 31.Furthermore, in step S4, this positive difference value total ST1· is compared with a predetermined threshold value ε1. , if formula (4) is satisfied, it is assumed that a decrease in the furnace heat will occur in step S5 due to the rapid flow of gas flow around the furnace, and an evaluation point of 1 will be given for the predetermined period.On the other hand, if formula (4) is satisfied If not,
Assuming that there is no abnormality, return to step S1, and proceed to step 8 below.
Furnace heat reduction evaluation is performed by repeating steps 1 to S4.

FIG. 7 is a flowchart showing the process flow of the second prediction means. In the same figure, in step S11, the furnace wall temperature of each FM sensor is 9. is measured every 1 ng time Δ. Next, in step S12, the difference value of each FM sensor 3 is calculated based on equation (1).

Then, in step 813, the sum of negative difference values 5T2j is determined based on equation (3). Furthermore, step S
In step 14, this negative difference value sum 5T2j is compared with a predetermined threshold value ε2 in line 0, and if the formula (4) is satisfied, in step 815 the furnace heat decreases due to the increase in the unloading speed. It is assumed that this is the case, and an evaluation point of 1 is given for a predetermined period. On the other hand, if the formula (4) is not satisfied, it is assumed that there is no abnormality and the process returns to step S11, whereupon step 8
Furnace heat reduction evaluation is performed by repeating steps 11 to 814.

FIG. 8 is a flowchart showing the processing flow of the third prediction means. In the figure, in step S21, the furnace wall temperature Tj,i of each FM sensor 3 is measured at every sampling time Δ. Next, in step S22, the difference value of each FM sensor 3 is calculated based on equation (1).

Then, in step S23, the moving average sum ST3 in the time width nΔt of the positive difference value based on formula (3)
seek. Furthermore, in step S24-17-, this positive difference value moving average sum 5T3j is compared with a predetermined threshold value ε3, and if the equation (4) is satisfied, in step 825, the furnace heat due to the furnace body heat dissipation is It is assumed that a decrease will occur and a score of 1 is given for a predetermined period. On the other hand, if the formula (4) is not satisfied, it is assumed that there is no abnormality, and the process returns to step S21, whereupon steps 821 to S24 are repeated to evaluate the decrease in furnace heat.

E, 4th. Fifth prediction means The amount of trunnion C (Kg
/1-p) is calculated for each sampling time Δt. Here, the amount of trail loss C at time j is set to X, and time j
If the amount of true loss C before the sampling time (that is, Δtxk hours before) is Xj-8, then the moving average XH of the predetermined time width nΔt reaching the current time j can be calculated as follows.

Based on formula (5), <X is calculated for each sampling time Δt, and according to formula (6), ... εxn) is exceeded, the evaluation point i is determined by the maximum threshold εxi.
and evaluate it.

Xo>εx i(i-1~n)...(6
) The above is the fourth prediction means.

In addition, nitrogen mc% in the furnace top gas detected by gas chromatography (hereinafter referred to as "gas chromatography N2 amount") has a strong negative correlation with the amount of trundle loss C, and instead of increasing the amount of trundle loss C, A decrease in blast furnace heat can be predicted by a decrease in the amount of gas chromatin N2.

As a result, if the gas chromatographic N2 amount at the current time j is yj, and the gas chromatographic N2 amount at a sampling time before time j (that is, ΔtXk hours before) is V j-, then the predetermined time width at the current time j The moving average VH of nΔt can be calculated as follows.

Based on equation (7), <yH is calculated for each sampling time Δ, and using equation (8) below, y8 is set to a predetermined threshold ε ・(to j-1・m)(εy1>εy2>・・εy
m) Give an evaluation point j based on the minimum threshold value ε4 when it is less than J,
Evaluate furnace heat drop.

yH〈εy j (J = 1~m)...
(8) The above is the fifth prediction means.

Furthermore, when calculating the moving average of the amount of vine loss C, which is the fourth prediction means, the instantaneous value of the amount of vine loss C is shown in Fig. 9(a).
As shown in , the abnormal value E1. due to noise etc. E2 may occur. Here, the amount of truss loss C at time j is expressed as
, 1 sampling time Δ and the previous trailing loss C amount is X, then the absolute value of the difference value of the trailing loss C amount ΔX J is ΔX, = lX--X, 1... (9) J
It becomes J J-1. By comparing this ΔX・ with the threshold value ε7 as shown in FIG. Find E2 and draw the same figure (C
), a method of smoothing by replacing it with the immediately previous measured value can be considered. By applying this method, a more accurate moving average of the true loss CI can be obtained, and as a result, prediction with considerably high accuracy is possible.

A method for predicting a decrease in furnace heat by using a moving average of the amount of truss loss C, which includes such abnormal value correction, can be realized using a computer. FIG. 10 is a flowchart showing the flow of the process. In the figure, first, in step 831, εx1<εx2...εxn
The threshold values εx1 to ε are set with the magnitude of . Then, in step 832, the instantaneous value XJ of the twist n loss Cff1 is obtained at every sampling time Δ. Then, in step 833, the truss Cf
Find the absolute value ΔXJ of the difference value of f1, then step 83
4, if the absolute value ΔXJ of the difference value is larger than the threshold ε7, in step 835, this instantaneous value
is regarded as an abnormal value, and is replaced with the immediately previous measured value X. The process moves to step 836. On the other hand, if it is smaller than the threshold value ε7 in step 834, the process moves to step 836 without changing the instantaneous value XJ. In step S36, the moving average XH of the time width nΔt is calculated, and in the next step 8
In step 37, the evaluation point i is initialized to O.

Then, in step S38, a comparison is made between the moving average value of the amount of trailing loss C×8 and the threshold value εx1 (from i-O), and if X ≧ε, the value of i is increased by 1 from O→1 in step SHxi39. , step 840
The threshold εx(
Step 838 while increasing the value of i+1) step by step.
to 840 are repeated to calculate the evaluation score i, and step 341
to move to. Also, if XH x εx1, step 839
．． Step 840 is never executed, the evaluation point i is set to 0, and the process moves to step 841.

Finally, in step 841, steps 838 to S4
Output the evaluation point i obtained from 0.

It goes without saying that the furnace heat drop prediction method including the abnormal value processing described above can be similarly applied to a computer in the case of furnace heat drop prediction based on the moving average value y8 of the gas chromatin N2 amount.

4. mentioned above. Since the fifth prediction means is based on a moving average for each sampling time Δ, prediction can be obtained quickly and the accuracy can be said to be sufficiently reliable.

F, threshold learning type, each threshold ε* of the first to fifth prediction methods described in EL
(Epsilon 1. ε2. ε3. εx1. εvj, hereinafter collectively referred to as "ε8") is determined in advance so as to obtain an optimal furnace heat reduction prediction rate. However, the initially optimal threshold value ε8 may become non-optimal due to changes in various conditions such as the production plan and raw material conditions during blast furnace operation. Therefore, the threshold ε1
It is necessary to change the threshold value ε1 to an optimal value during blast furnace operation, and learning of the threshold value ε1 is performed as shown below.

■Threshold value learning for first to third prediction means When measuring the furnace wall temperature of a blast furnace, FM, which is a furnace wall thermometer,
As mentioned above, the temperature measurement point of the sensor 3 changes with the erosion of the furnace wall. For this reason, not only the value of the measured furnace wall temperature itself but also the peak value (maximum! direct) of the furnace wall temperature may vary depending on the erosion of the furnace wall. Since the first to third prediction means make predictions based on the furnace wall temperature difference, the prediction accuracy will also vary in such a case.

In order to solve this problem, the following processing is performed. First, the time jk (k is 1.2.3.
...Numbered as r (r is the number of times the furnace heat decreases). ) is detected, and going backwards from the furnace heat drop time t to the past time span h8, each summation S 1 . ~S Ding 3. Find the maximum value of J J and set it to MAX1 k, MAX2, and MAX3, respectively.

Figure 11 shows an example of this, and the figures (a) and (b)
, (C), (d) are the sum of positive difference values, respectively, 5T1j
, the absolute value sum of negative difference values 8to2j, the moving average sum of positive difference values ST3. ．． Change over time of hot metal temperature below typical value
p+(] is shown. TC is the control temperature, and a decrease in furnace heat occurs at times t1 and t2. ■ to ■ are MAX1 and MAX1, respectively.
2. MAX21. MAX22, MAX31. MAX3
It is 2.

In this way, each maximum value MAXI is obtained.

MAX2, MAX3. Accordingly, the threshold values 81 to ε3 are set as follows (
Determined from equations 10) to (12).

ε1= al @ [(1/ r ) Σ MAX
Ik]k=l...(10) 6 =a - [(1/r) ΣMAX2k]2,
2
k-.

...(11) ...(12) Note that 81 to a3 (>O) are coefficients.

■No. 4. Threshold learning of the fifth prediction means - amount of trail loss C in a sampling time of about a week,
The average value of gas chromatography N21 changes with changes in operating conditions during blast furnace operation. Therefore, it is necessary to change the threshold values εxi and εvj accordingly.

In order to solve this problem, the following processing is performed. Average value SO of the amount of Tsuru loss C and gask mouth N2 amount that have increased in the past for a predetermined period (long term of about - weeks) from the current time t
1, N2 and standard deviation σS.

Find σ8 and use these values to obtain the following (13), (14)
The threshold values ε·, ε· are determined by the formula.

XI VJ ε,=Sol+a −i・σs・(13)XI
X ε ・−N −a −j・σN ・・・(14
)VJ 2 V Note that a, a, and (>O) are coefficients.

FIG. 12 is a flowchart illustrating the processing procedure for threshold erasure described in (1) and (2). In the same figure, first, in step 851, the furnace heat drop time tk (k=1 to r, r is the number of furnace heat drops) in the past fixed period hF from the current time t is detected, and then in step 852, the The total sum of positive difference values 5T1j during a predetermined time width h that is opposite to .

Negative difference value absolute value sum 5T2j, positive difference value moving average sum ST3. (7) Maximum value MAX1, MAX2J
k,, MAX3k is calculated. These maximum values MAXlk
- Using MAX3k, in step 853 (10)
Threshold values ε to ε3 are determined from equations ˜(12).

Next, in step 854, calculate the average values Sol, N2 and standard deviations σS, σN of the amount of truss loss C and the amount of gask opening N2 in the past fixed period p from the current time t,
In step S55, threshold values ε . and ε . are determined using equations (13) and (14). Thereafter, the process returns to step 851 and the threshold value is determined as needed by repeating the above-described process.

As explained in (■) and (■) above, by automatically determining the threshold value ε1 based on the changes in the sampling data in the first to fifth prediction means at the appropriate time, production plans and raw materials can be adjusted during blast furnace operation. Predictions can always be made with high accuracy regardless of changes in conditions and other conditions.

G. Comprehensive prediction means. By deriving the evaluation scores of the first to fifth prediction means described in E. while changing the threshold value ε8 from time to time as described in Dan-, a comprehensive prediction method is obtained as described below. Make predictions.

FIG. 13 is a flowchart showing the flow of the process, and will be described below with reference to the same figure.

First, in step 861, the evaluation points f −f of the first to third prediction means are determined. Evaluation points f to f3 are 1 degree O →
If it becomes 1, that value is held during a data hold period, which will be described later. Therefore, if the hold time (approximately 2 hours) is set at time t1, each of the evaluation points f1 to f3 will change from O to 1 during the time t to time t1+1hr (that is, the data hold period). , it never changes from 1 to O again. This data hold period is the FM sensor total value F (-f + f + f3)
When F=O−)F>O changes, the fluctuation time t.

This is set at the same time as initializing , and from then on, the time defined by is the data hold period as described above.

Next, in step S62, the FM sensor total value F is 0.
By determining whether or not, the data hold time h
If F=O, there is no setting of data hold time hr, so step 86
Move on to 5. On the other hand, if F≠O, since the data hold time has already been set, the variable time t is set in step S63.

By comparing with the hold time h, it is checked whether the current time is in the data hold period or not. If tC≦hr, it is in the data hold period, so there is no need to initialize the FM sensor total value F. Since there is no one, the process moves to step S65. However, in step 863 t. ＞hr
If so, it is determined that the data hold period has ended, and the FM sensor total value F is initialized to "O" in step S64.

In this way, the FM sensor total value F is determined while taking the data hold period into consideration. This means that the first to third prediction means are the fourth prediction means. Since it has higher foresight than the fifth prediction method (predictions can be obtained quickly), the prediction results of the first to third prediction methods must be held for a certain period of time in order to comprehensively evaluate the prediction results for the same point in the future. This is because it is necessary to keep it.

Then, in step S65, the evaluation score i by the fourth prediction means is calculated, and further in step 866,
An evaluation score j by the fifth prediction means is calculated. Next, in step 867, the overall evaluation C is determined according to the following formula.

C=WA F+WB i +wc j
・ (15) Here, wA, wB, wo are the first to third
．． 4th degree is the weight for each of the fifth prediction means. This comprehensive evaluation C is examined in step 868, and C-
If it is 0, it is assumed that there is no tendency for the furnace heat to decrease at all, and the process returns to step S61, and the comprehensive evaluation C is obtained again through steps 861 to 67. On the other hand, if C>0, step 8
At step 69, an alarm whose importance is changed according to the value of the comprehensive evaluation C is output.

Thereafter, the process returns to step S61, and the furnace heat reduction prediction continues.

In this way, by changing the degree of alarm according to the value of the comprehensive evaluation C, the degree of heating action can be finely changed. As a result, it becomes possible to select the necessary and sufficient heating action, which not only reliably prevents a drop in furnace heat, but also prevents unnecessary rises in furnace heat due to excessive heating action, resulting in stable Moreover, economical blast furnace operation became possible.

Furthermore, the prediction accuracy is further improved by automatically changing the threshold value ε1 of each of the first to fifth prediction means in accordance with changes in blast furnace operating conditions and the like.

1", Supplementary Note: In the first to third prediction means in this example, an FM sensor was used as the inner wall thermometer; however, even a normal temperature sensor (for example, a sheathed thermocouple) may have problems in terms of lifespan. Can be substituted. Furthermore, a stave thermometer or a brick-embedded thermometer can be used as a substitute, although the prediction accuracy will be slightly lower due to their reliability and low temperature response.

Furthermore, in the first to third prediction means in this example,
Although 28 FM sensors 3 were installed in 7 levels and in 4 directions, it goes without saying that they may be installed appropriately depending on the characteristics of the blast furnace.

Furthermore, as mentioned above, it is desirable to apply threshold learning to all of the first to fifth prediction means, but by applying it to at least one of the first to fifth prediction means, The prediction accuracy of the prediction means can be improved.

(Effects of the Invention) As explained above, according to the present invention, predictions can be made quickly, and the predictions can be made while changing the threshold values of the first to fifth prediction means in a timely manner. Accurate predictions can be made and heating actions can be taken as needed.

[Brief explanation of the drawing]

FIGS. 1(a) and 3(b) are a side sectional view and a plan sectional view, respectively, showing the arrangement of an FM sensor used in an embodiment of the present invention in the blast furnace wall, and FIGS. 2 and 3 are sectional views, respectively. FM sensor conceptual diagram, installation explanatory diagram, Figure 4 is a graph showing changes over time in the furnace wall temperature measured by the FM sensor, Figure 5 is a graph showing changes over time in the difference value of the furnace wall temperature measured by the M sensor, FIG. 6 is a flowchart showing the processing flow of the first prediction means, FIG. 7 is a flowchart showing the processing flow of the second prediction means, and FIG. 8 is a flowchart showing the processing flow of the third prediction means. , Figures 9(a), (b), and (C) are the instantaneous values of the amount of trailing loss C including abnormal values, and the trailing loss C, respectively.
A graph showing the absolute value of the difference value of ff1 and the instantaneous value of the amount of true loss C after removing abnormal values, FIG. 10 is a flowchart showing the process flow of the fourth prediction means, and FIG.
A graph showing a part of the threshold learning example of the third prediction means (
a), (b), (CL (d) are graphs showing changes over time in the sum of positive difference values, the sum of absolute values of negative difference values, the moving average sum of positive difference values, and the representative value of hot metal temperature, respectively, the 12th 13 is a flowchart showing the processing procedure of threshold learning, FIG. 13 is a flowchart showing the processing flow of the comprehensive prediction means,
Figure 4 is a side cross-sectional view showing the arrangement of the furnace belly sonde in the blast furnace in the prior art, Figure 15 is a graph showing the correlation between the hot metal temperature and the temperature in the local area of the furnace belly, and Figures 16 (a) and (b) are Graphs showing the hourly average value of vine loss C amount and changes in hot metal temperature over time, Figures 17 (a) and (b) show the 1 hour average value of vine loss C amount and the change in hot metal temperature when the heating action occurred, respectively. It is a graph showing changes over time.

Claims (1)

[Claims] (1) An inner wall thermometer is installed at a predetermined location in the blast furnace, and the inner wall temperature difference is measured at a predetermined time interval with the inner wall thermometer, and the sum of the portions that show a positive value of the inner wall temperature difference at a certain time. The first one gives an evaluation point for a predetermined period when the value exceeds the threshold.
and a second means for giving an evaluation point for a predetermined period when the total value of the portion indicating a negative value of the inner wall temperature difference at a certain time exceeds a threshold value.
a third prediction means for giving an evaluation point for a predetermined period when the sum of moving average values of a portion showing a positive value of the inner wall temperature difference at a certain time over a predetermined time width exceeds a threshold; A fourth prediction means that calculates the amount of carbon at predetermined time intervals and gives an evaluation point when the moving average value of the calculated values over a predetermined time width exceeds a threshold; a fifth prediction means that calculates for each interval and gives an evaluation point when the moving average value of the calculated values in a predetermined time width is less than a threshold, When predicting a blast furnace furnace heat drop according to the comprehensive evaluation of at least one evaluation point of the prediction means, the threshold values of the first to third prediction means are set from the time when a furnace heat drop was observed within a certain period of time in the past. The threshold values of the fourth and fifth prediction means change from time to time based on the total value of each correspondence or the maximum value of the total sum during a predetermined time span that has gone backwards in the past, and the threshold values of the fourth and fifth prediction means A blast furnace furnace heat drop prediction method, characterized in that the method changes from time to time based on the average value and standard deviation of the corresponding solution loss carbon amount or nitrogen amount.