JPH01205011A - Apparatus for controlling furnace heat in blast furnace - Google Patents

Abstract

PURPOSE:To control furnace heat in a blast furnace at high accuracy corresponding to the actual furnace condition by executing the prescribed inference as artificial intelligence from processed data prepared from the data and the knowledge data based on the experiences in the blast furnace. CONSTITUTION:The various data in the blast furnace 1 are stored into filing means 14 from a scanner 11 through collecting function 12. These data are operated with operating means 16 to store the obtd. processed data into a common data buffer 24. Based on these processed data and the knowledge bases from the knowledge base storing means 22, the furnace heat level and the furnace heat shifting in the blast furnace 1 are inferred and operated with an inference engine 26, to decide an action rate to the blast furnace 1. The above knowledge bases contain various judging knowledge base, correcting knowledge base, synthetic judging knowledge base and rules deciding the action rates. The above decided action rate is outputted into a digital measuring instrument 32 from blast furnace heat control device 10 through output means 18 and the furnace heat control in the blast furnace is executed under considering the actual operational result.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to a blast furnace furnace heat control device that controls the temperature of hot metal tapped from a blast furnace.

[Prior Art] Conventionally, as a method for estimating, managing and controlling the temperature of hot metal in a blast furnace, a blast furnace operator qualitatively judges information from various sensors installed in the blast furnace. The method used is to evaluate the blast furnace situation and make optimal adjustments to operating factors.

However, there are individual differences in the evaluation results depending on the ability and experience of the operator, which makes it difficult to standardize operational actions, and the evaluation is not quantitative, making it difficult to estimate the hot metal temperature. there were.

For this reason, a process control method for a blast furnace has been proposed, for example, as disclosed in Japanese Patent Publication No. 51-30007. This process control method maintains the air blowing temperature constant and determines the amount of direct reduction in the furnace from a constant value of ∆ which can be manually determined continuously during operation. Depending on the difference from the smoothed value, the S
The air humidity is determined by an equation to which a correction term is added to prevent long-term fluctuations in the i content, and the heat balance in the furnace is controlled by this air humidity determination value.

For this reason, accurate operation has been achieved by correcting the large wave changes (long-period changes) that occur when the blast furnace status is controlled by calculation.

[Problems to be Solved by the Invention] In the conventional process control method disclosed in the above-mentioned Japanese Patent Publication No. 51-30007, information from sensors is analyzed and input into a model to perform predetermined calculations. ing. For this reason, the computer that executes the calculation uses, for example, Fortran as the language, but the calculation capacity is extremely large. Furthermore, since blast furnaces change over time, the analysis model itself must be changed for maintenance, but since the analysis model itself is complex, changing the conditions of the analysis model becomes an extremely troublesome task.

In addition, as a means to solve the above-mentioned problems, it is possible to improve maintainability by using a computer system that uses an artificial intelligence language such as LISP, but here, sensor information (authenticity data, various sensor data) ,
In using the knowledge base to reason about the furnace thermal situation,
When using the production rule, for all related sensors, for example, (1) IF (the temperature of sensor i is in the range of T1 to T2).
THEN (CF value, which is a high fever level, is C1)...
..., (CF value, which is a low heat level, is Cn); ■IF (Temperature of sensor 1 is in the range of T1 to T3.) THEN
(CF value C'l of high fever level).

・・・・・・(CF value of low fever level is C'n);■・・・
. . . It is necessary to express the rules as follows, which results in a huge number of rules, which increases the inference time and makes the adjustment of the CF value extremely complicated.

For this reason, there is a problem in that the creation of a program for the above-mentioned rules or the amount of numerical values indicating the reliability described in the rules becomes enormous, making the manual work complicated.

Furthermore, the effects of actions taken in the past are not taken into account, and therefore the actual situation of the furnace is not necessarily understood, which also leads to the problem of insufficient furnace temperature control accuracy. there were.

The present invention was made to solve these problems, and it is possible to control the furnace heat of a blast furnace with high precision, improve the calculation capacity and calculation speed when realized by a computer, Another object of the present invention is to provide a blast furnace furnace heat control device that allows rules to be easily added and modified in response to new situations such as aging of the blast furnace, and also takes into account the effects of past actions.

[Means for Solving the Problems] FIG. 1 is a block diagram showing the concept of a blast furnace furnace heat control device according to the present invention. The blast furnace furnace heat control device according to the present invention includes:
A data input means that takes in data from various sensors installed in the blast furnace at a predetermined timing, and based on the data from the sensors, determines the siding temperature, unloading speed, pressure loss, furnace top temperature, gas utilization rate, and solutions. It has processing data creation means for creating various data indicating the situation of the blast furnace, such as the amount of loss, and comparing the various data with reference data to create difference data; A storage means for temporarily storing data (hereinafter referred to as processed data); a knowledge base storage means for storing various knowledge bases based on experience, results, mathematical models, etc. regarding blast furnace operations; and a knowledge base storage means for storing the processed data in the storage means. The blast furnace includes inference means for inferring the furnace heat level and furnace heat transition based on the knowledge base of the knowledge base storage means and determining an action amount for the blast furnace, and means for outputting the action amount.

The knowledge base storage means stores a “furnace heat level determination knowledge base J” used to estimate the furnace heat level, a “furnace heat transition determination knowledge base” used to estimate the furnace heat transition, and a “furnace heat level determination knowledge base J” used to estimate the furnace heat level. "Action Judgment Knowledge Base" used to determine the amount of action based on heat level and furnace heat transition
, an "action correction knowledge base" that determines the correction value of the action amount based on past actions and disturbances, and a "comprehensive judgment knowledge base" that determines the actual action amount from the action amount and the action correction amount. There is.

The above-mentioned "Furnace heat level judgment knowledge base" and the above-mentioned "Furnace heat transition judgment knowledge base" use machining data and furnace heat level as independent variables, and use the probability that these combinations will occur (hereinafter referred to as confidence) as a dependent variable. each has a confidence function and a rule group that determines how to apply this confidence function, and the "action determination knowledge base" has a rule that determines the amount of action based on the furnace heat level and the furnace heat transition,
The "comprehensive judgment knowledge base" has rules for determining the actual amount of action from the amount of action and the amount of action correction.

Further, the "action correction knowledge base" includes an action join correction rule that is used to determine an action correction amount by determining an action taken in the past and its influence amount.

[Operation] In the present invention, after the processed data creating means creates various data indicating the status of the blast furnace based on the blast furnace data from the data input means, the processed data is created based on the data. Based on the processed data and the knowledge base stored in the knowledge base storage means, inference calculations are performed as artificial intelligence, and the amount of action for the blast furnace is determined.

In the inference calculation, the furnace heat level is estimated using the furnace heat level determination knowledge base, and the furnace heat transition is estimated using the furnace heat transition determination knowledge base.

When inferring these furnace heat levels and furnace heat transitions, a confidence function and a set of rules that determine how to apply the confidence function are used to obtain confidence in the inference results.

Next, the amount of action is determined by applying the above-mentioned furnace heat level and furnace heat transition to the action determination knowledge base. Furthermore,
Action correction Knowledge-based action correction rules are used to determine the influence of past actions, and an action correction amount is determined based on this.

Next, the actual amount of action is determined from the amount of action and the amount of action correction using the comprehensive judgment knowledge base. Then, the blast furnace furnace heat is controlled by adjusting the operating factors based on the action amount.

[Example] Embodiments of the present invention will be described below based on the drawings.

FIG. 2 is a block diagram showing a blast furnace furnace heat control device and related equipment according to an embodiment of the present invention.

In the figure, (1) is a blast furnace to be controlled, (10) is a blast furnace furnace heat control device according to the present invention, and a data scanner (
11), sensor data collection means (12), file means (14), calculation means (16), output means (18), knowledge base (22), inference engine (2B), and common data buffer (24). Contains.

The sensor data collection means (12) collects data from various sensors, such as temperature sensors, pressure sensors, gas sensors, etc., using data scanners (lla), (llb), (l).
Input processing is performed in chronological order via lc).

The file means (14) performs a file data banking function. The calculation means (IB) performs exponential smoothing processing on the data from the sensor data collection means (12) stored in the file means (14), and then returns the data to the file means (14).
14). Then, at predetermined intervals, for example, every 20 minutes, the average value and the difference data between the average value and the reference value are sent as processed data to a common data buffer (24) which is processed data storage means. In addition to the above, this calculation means (16) also performs part of the calculation processing of the action correction rule, which will be described later.

The inference engine (26) uses its data and knowledge base (2
Perform predetermined inference calculations based on the knowledge of 2), calculate the amount of action to be taken next, and use the common data (offer (2)
4), and also stored in the file means (14).

(30) is a CRT on which the inference results of the inference engine (2B) are transmitted and displayed via the file means (14).

(32) is a digital instrumentation device, which is a blast furnace furnace thermal control device (l
It controls the temperature of the blast furnace based on the command of O), and adjusts the manipulated variables that can adjust the furnace heat, such as air humidity, air temperature, and liquid fuel such as heavy oil. At this time, the command of the blast furnace furnace heat control device (10) is sent to the digital instrumentation device (32) via the output means (18).

(40) is a hot air stove, (42), (44), (4B
) are each control valves.

An overview of the operation of this embodiment having the above configuration will be explained.

(1) First, data from various sensors are sequentially read by the sensor data collection means (12) via the data scanner (11) at predetermined timings, for example, at one-minute intervals;
It is stored in the file means (14).

, ) The data stored in the file means (14) is subjected to exponential smoothing processing by the calculation means (16). Here, data regarding unloading, temperature, gas utilization rate, tap slag, etc. are processed. The various data subjected to the arithmetic processing are stored again in the file means (14). Next, the average value of these various data at intervals of a predetermined period of time, for example, 200 minutes, and the difference between the average value and a predetermined reference value are calculated and transferred to the common data buffer (24) as processed data. .

(3) The inference engine means (2B) has a knowledge base (22
) and the processed data in the common data buffer (24), the situation inside the blast furnace is inferred and calculated.

Here, the knowledge base includes a furnace heat level judgment KS group, a furnace heat transition judgment KS group (KS; knowledge source), an action judgment KS, and an action correction judgment KS, as shown in Figure 3.
, comprehensive judgment KS group, and operating status judgment KS knowledge base units.

Furnace heat level determination KS Group uses an inference engine (2G) to determine at what level the furnace heat level of a blast furnace is.
"Hot metal temperature - Furnace heat level KS", which determines the furnace heat level using the hot metal temperature as the main determining factor, "Hot metal temperature - furnace heat level KS", which determines the furnace heat level using the measured quantities of other sensors as the main determining factor. Includes sensor furnace heating bell KSJ, etc.

Each of these KS uses each measured quantity and the furnace heat level as independent variables, and the probability (confidence) that a combination of these will occur.
It is made up of a confidence function (hereinafter referred to as CF function) whose dependent variable is , and a set of rules that determine the procedure for using the CF function.

Furnace heat transition determination KS Group uses an inference engine (
26), which determines the transition of furnace heat using the transition of hot metal temperature as the main determining factor [hot metal temperature - furnace heat transition KSJ, Determine the thermal transition [Includes sensor furnace thermal transition KSJ, etc.]

These KSs also include a CF function that has each measured quantity and furnace heat transition level as independent variables and the probability (confidence) that a combination of these will occur as a dependent variable, and a set of rules that determine the procedure for using the CF function. It is established from

The action determination KS is made up of a group of rules for determining the amount of action based on a combination of the furnace heat level and the furnace heat transition level. The action correction amount KS group is made up of a group of rules for correcting the current action amount based on information on actions taken in the past and disturbances that occurred in the past. The comprehensive judgment KS is made up of a group of rules for determining the final action amount based on the action judgment result and the action correction amount judgment result.

Then, the operating state determination KS determines whether or not the furnace is operating normally based on, for example, the furnace heat level,
It consists of a set of rules that sends the above action amount as is to the control system if it is normal, and if it is abnormal, it is displayed and provides guidance to the operator.

The inference engine means (26) executes each knowledge base, and as shown in the flowchart of FIG. 3, first determines the furnace heat level and furnace heat transition, and then determines the amount of action based on these determination results. Determine. This amount of action is corrected in a predetermined manner, and the result is temporarily stored in the common data buffer (24) and then stored in the file means (14).
) and the digital instrumentation device (3) via the output means (18).
2).

Then, the control valve (42) is controlled by the digital instrumentation device (32).
), (44), and (4G) are appropriately controlled to perform an action operation, and the temperature of the blast furnace (1) is controlled.
As a result, the hot metal temperature is controlled to a desired value.

Next, an outline of the structure of the knowledge base and its specific inference will be explained based on FIG. 3.

(A) Furnace heat level determination K S (Knowl
ege 5 source); knowledge source) group; This furnace heat level judgment KS group is a knowledge base for determining the state of furnace heat at the inference start time. It consists of KS groups such as Bell KSJ, and as shown below,
The CF value distribution is determined by the method described later for the furnace heat level, which is divided into seven levels from high to low for each KS group, and the level of maximum certainty is taken as the furnace heat level at the current time.

Furnace heat level evaluation 7 Large fever 6 Medium fever 5 Normal 4 Small cold 3 Medium Cold 2 people cold 1 Extra large cold Here, an example of the hot metal temperature-furnace heat level KS will be explained.

The KS is composed of an IF section in which conditions are set and a THEN section in which instruction contents are set when the conditions are met. Examples are as follows.

Rule Nα1 [IF section] Fully satisfied -1 NOT (There is a lot of residue) NOT (6180 minutes elapsed after the end of wind reduction) Judgment of Si and S is "low" [THE N section] Hot metal temperature - by normal three-dimensional function Find the CF value of the furnace heat level.

This rule No. 1 indicates that when the operating state of the blast furnace is in a steady state, the CF value of molten metal temperature - furnace heat level is determined by a normal three-dimensional function.

That is, the CF value function of hot metal temperature - furnace heat level usually has ~f
1 Slightly high ~ f 1 High ~ fMTE) IMTN
There are three types of MTH, and the inference engine (2B) selects which function to use based on the logic described above.

In other words, the hot metal temperature - furnace heat level KS is determined by the number of degrees (the number indicating the position of the ladle where the hot metal temperature was measured compared to the ladle used since the start of tapping) and the amount of slag after the end of wind reduction. Rules for determining which of the three types of functions to use are stored in accordance with the elapsed time. Other examples of this rule are as follows: Rule No. 2 [IF section] Pot head - INOT (a lot of residue) NOT (5180 minutes elapsed after the end of wind reduction) St and S are judged as "slightly high" [THEN CF of hot metal level by three-dimensional function when “slightly high”
Find the value.

Rule No. 3 [IF part] Pot head-1 NOT (Many residue) NOT (Elapsed time 5180 minutes after wind reduction ended) Si, S judgment is "high" [THEN part] Three-dimensional function when "high" Find the CF value of the hot metal level.

These rules NCL 1 to No. 3 indicate that when there is not much residue and sufficient time has elapsed since the end of wind reduction, the type of function is selected based on the determination of Si and S in the hot metal.

Further, the three CF functions of hot metal temperature-furnace heat level are all functions of pot head LN, hot metal temperature MT, and furnace heat level FHL.

That is, f - f (LN, MT, FHL) MTN
MTN t −f (LN, MT, FHL) MT
II M T 11 t -f (LN, MT, FHL) M T
E II M T E 11 Inference Engine (26)
is for the CF value function selected by the above rule,
The actual measured values are applied to LN and MT, and the numerical values 1 to 7 mentioned above are applied to FHL to determine the CF value corresponding to each furnace heat level.

FIG. 4 shows the CF value function for one pot head. Since the pot head is fixed, the CF value is a function of the hot metal temperature and the furnace heat level.

For example, if f is selected according to the rule, then when the hot metal temperature is 1400°C, the CF value is highest at furnace heat level 4, which is CF-0,2, and the hot metal temperature is 1480°C.
When the furnace heat level is 7, the CF value is highest and CFO0.
4.

Next, the sensor furnace heat level (tuyere embedding temperature - furnace heat level K
S, solution loss Cff1-furnace heat level KS) will be explained.

This KS includes rules for determining whether to use tuyere embedding temperature-furnace heat level KS and solution loss Cff1-furnace heat level KS, as well as tuyere embedding temperature, furnace heat level, and solution loss C amount, respectively. HT SL f or-f HT (HT, F HL) fsL-, which consists of CF value functions f and f, with and furnace heat level as two independent variables.
fsL (SL, FHL) Tuyere embedment temperature HT used to calculate this function,
For the solution loss Cr;kSL, the actual all value (its exponential smoothed value or moving average value) is used.

Examples of rules for determining the use or non-use of these functions are shown below.

Rule Nα1 [IF section] NOT (many residues) [THE N section] (1) Calculate the CF value of tuyere embedding temperature - furnace heat level from the tuyere embedding temperature.

(2) Calculate the CF value of solution loss C1a - furnace heat level from solution loss Cm.

In this case, the inference engine (26) uses both openings.

Rule No. 2 [IF section (lots of residue) [THEN section (1) The CF value of tuyere embedding temperature-furnace heat level is set to "0".

(2) Solution loss C amount - Set the CF value of the furnace heat level to "0".

In this case, both the tuyere embedding temperature-furnace heat level CF value and the solution loss C'm-furnace heat level CFi are constant values regardless of the furnace heat level.

This means that these factors do not contribute to the determination of the furnace heat level, which will be described later, and the furnace heat level is determined only by the above-mentioned molten metal temperature - furnace heat level KS.

FIG. 5 is an explanatory diagram showing the operation of the inference engine (26), which calculates the CF value for each furnace heat level based on the hot metal temperature - furnace heat level KS, and also calculates the CF value for each furnace heat level. K
S and Solution Loss C Amount - Furnace Heat Level KS Based on S, the CF value for each furnace heat level is determined. Then, the CF value of each level based on the tuyere embedding temperature KS and the CF value of each level based on the solution loss (, tKs) are added.

The CF value of the sensor level thus obtained and the CF value of each level based on the hot metal level KS described above are added.

In this way, the CF value for each furnace heat level (7 to 1) is determined.

CB) Furnace heat transition determination KS group; This furnace heat transition determination KS group includes hot metal temperature-furnace heat transition KS and sensor furnace heat transition KS, and is used to evaluate the degree of change in furnace heat transition from the past to the present. Therefore, as shown below, the CF value is determined for each rank in five stages between rapid rise and constant rapid decline, and the stage position of the maximum value is taken as the furnace heat transition state at the current time.

Level content 5 Rapid rise ４　　　　　　Up　Up 3. Staying the same. 2　　　　　　Down　Downward 1. Sudden descent Here, the hot metal temperature-furnace heat transition KS will be explained.

This KS is a CF function f (ΔMT,
VFHL) and ΔMT pre-processing and post-processing rules for this CF function are stored.

The method of using this KS will be explained below.

Rule No. 0 [IF Department (Initial setting) [THE N section] The following values are set as the CF value of the molten metal transition.

Level 12345 CF value　oooo.

Rule No. 1 [IF Part (1) NOT (S t, S judgment is "slightly high") (
2) NOT (Judgment of Si, S is "high") (3) Stability flag is ON (furnace condition is stable) [THEN section (1) ΔMT - (current tap hot metal temperature - previous) The CF value is determined for each furnace heat transition level as the hot metal temperature at tap), and
This time, the hot metal temperature-furnace heat transition CF value is used.

(2) Next, the CF value is determined for each furnace heat transition level as ΔMT-(hot metal temperature of the previous tap - hot metal temperature of the front tap), and is defined as "previous hot metal temperature - furnace heat transition CF value".

(8) [Current hot metal temperature - Furnace heat transition CF value] and [Previous hot metal temperature - Furnace heat transition CF value] are each multiplied by a weighting coefficient, and
Add to "Hot metal temperature - Smooth furnace heat transition CF value".

(These calculations are performed separately for each level of the furnace heat transition.) In other words, in a stable furnace condition, up to the most recent data is used to estimate the furnace heat transition.

Rule No. 2 [IF section] (1) (Judgment of Si and S is "slightly high") (2) NOT
(Judgment of S i, S is "high") (3) Stability flag is OFF [THE N section] (1) For each furnace heat transition level as ΔMT - (previous tap hot metal temperature - pre-previous hot metal temperature) Find the CF value and use it as the "previous hot metal temperature - furnace heat transition CF value".

(2) Multiply the "previous hot metal temperature - furnace heat transition" by a weighting coefficient, add it to the "hot metal temperature - smooth furnace heat transition CF value", and set it again as the "hot metal temperature - smooth furnace heat transition CF value".

Here, since the furnace conditions are not stable, "current hot metal temperature - furnace heat transition" is not considered.

Next, the sensor furnace heat transition KS will be explained.

The sensor furnace heat transition KS stores the number of CF values with the transition of measured values of a large number of sensors and the furnace heat transition as two independent variables, and the rules that determine how to use them. It is set in.

That is, f - f (AS t, VFHL) Δ
St Δ91 (i-1 to n: for each corresponding sensor) Among these, an example will be explained for a sensor whose sensor measures the tuyere embedding temperature.

Rule No. 1 [IF part] NOT (There is a lot of residue) [THEN part] The CF value of the tuyere embedding temperature - furnace heat transition is calculated as Δ5i - (tuyere embedding temperature - 60 minutes front feather embedding temperature) demand.

Rule No. 2 [IF section] There is a lot of residue CF value of wind transition level 4〉0 [THE N section] Overwrite the CF value of run transition level 4 with "0".

Level 12345 CF value *** 0 * (* indicates that the value remains as it was) Rule No.
3 CF value of texture transition level 5 with a lot of residue > 0 [THE N section] Overwrite the CF value of texture transition level 5 with "0".

Level 12345 CF value ***0 Figure 6 shows the tuyere embedding temperature-furnace heat CF function with the tuyere embedding temperature (difference from the reference value) and the furnace heat transition level as independent variables.

Note that it is also possible to create rules for the hot metal temperature-furnace heat transition KS by dividing it into "short-term transition" and "long-term transition,"
In addition to the tuyere-embedded KS, the sensor transition KS includes other KS,
For example, in addition to each KS such as unloading, blowing pressure, gas utilization rate, solution loss amount, etc., this information is also taken into consideration.

The inference engine (2B) calculates the CF value for each transition based on each rule of the hot metal temperature-furnace heat transition KS, and calculates the CF value for each sensor based on the rule of the sensor furnace heat transition KS. Find the CF value for each transition. Then, the CF values of these KSs are added for each furnace heat transition level to obtain the CF value for each furnace heat transition level.

(C) Action judgment KS.

This action determination KS is a knowledge base for determining the current furnace heat state on a matrix centered on the furnace heat transition and furnace heat level, and determining the action to be taken.

The inference engine (26) calculates the product of the CF value of the furnace heat level and the CF value of the furnace heat transition based on the above-mentioned action determination KS, and writes it into the matrix.

FIG. 7 shows an example of this, and this example shows that the apex (maximum value) of the CF value is a furnace heat level of -4 and a furnace heat transition of -3. Note that the action type and amount of action at each position on the matrix are stored in advance as knowledge in the frame.

FIG. 8 is a diagram showing an example of an action type, and FIG. 9 is a diagram showing an example of an action amount.

Note that when adopting an action-type action at a predetermined position, it is necessary that the CF value reaches a predetermined value. In principle, all action amounts are automatically controlled, but it is also possible to manually control some of them (for example, action amount G in FIG. 9).

(D) Action correction amount determination KS group; This action correction amount determination KS group determines the amount of corrected action by determining actions taken in the past or disturbances, and taking into account their influence at the current time. Contains various KS for The contents include changes in manipulated variables such as air humidity, air temperature, liquid fuel, coke ratio, etc.
It also consists of rules for detecting and responding to disturbances such as coke moisture and falling deposits.

For example, when the ventilation humidity is changed, the change time and amount are automatically detected by the "operation amount change detection" rule, and the subsequent influence amount is considered as a function of time by the "ventilation humidity" rule. In addition, when something attached to the furnace wall falls off, the "wall fall" rule automatically determines the falling place, the amount of influence on the furnace heat, and the time for the tuyere tip to fall, and the operation time and amount of the preliminary action are determined and correction calculations are made. be incorporated into.

The inference engine (2B) executes each of the above rules to obtain the necessary correction action amount and operation time.

(E) Overall judgment KS.

This comprehensive determination KS is a knowledge base for comprehensively determining the amount of action to be taken based on the determination results of (C) and (D) above. Then, when this KS is inferred by the inference engine (26) and a judgment result is obtained, the judgment result is input to the operating state judgment KS, the operating state is judged, and the action to be taken is displayed on the CRT (30). The quantity is instructed and guided to the operator, and at the same time, feedback is given to the digital instrumentation device (32) for predetermined automatic control (see Fig. 2).

Next, a method of creating a multidimensional function used when determining the CF value using each of the above knowledge bases (A) and (B), such as a CF function of hot metal temperature-furnace heat level, will be explained.

FIG. 10 is a characteristic diagram showing the change in hot metal temperature within the tap (continuous measurement results). As shown in the figure, even if the thermal condition inside the furnace is stable, the influence of bottom cooling during the residence in the lower part of the furnace,
During the early stage of tapping, the hot metal is discharged at a relatively low temperature due to cooling in the tap trough, which is the flow path for the hot metal. As time passes, the temperature of the hot metal in the tap runner increases as it receives the sensible heat of the hot metal, and as the influence of retention in the furnace decreases, the temperature of the discharged hot metal gradually rises and stabilizes. , the information is representative of the furnace heat that the system is controlling.

Furthermore, the tendency of this temperature increase is not always constant and changes depending on the operating conditions, so the information on hot metal temperature is ambiguous information whose suitability as information representing the system's control target changes from moment to moment. becomes. Therefore,
In order to observe the hot metal temperature and link it to the estimation of the furnace thermal state, it is necessary to keep in mind how much time has passed since the start of tapping, and also to take into account the ambiguity peculiar to the process. It becomes necessary.

If we understand the temperature of hot metal during tapping from the above perspective,
The temperature distribution is expressed as shown in FIG. In the figure, each axis means the following: X axis; dimensionless time Y axis; hot metal temperature Z axis: appearance frequency (occurrence frequency rate).

FIG. 12 shows the relationship between the hot metal temperature Tt and the tap maximum hot metal temperature Tmi at tapping time X-X1 in FIG. 11. Based on this figure, the three-dimensional function shown in FIG. 13 is obtained by adjusting the relationship between the tap maximum hot metal temperature and the furnace heat, and the relationship between the appearance frequency and the CF value. Since the tapping temperature is measurement information that depends on the elapsed time from the start of tapping and the operating conditions, more than 30 types are prepared so that they can be used properly, and are automatically selected according to the conditions.

When the characteristic diagram of FIG. 13 is illustrated with the X axis replaced with sensor data, it is expressed as shown in FIG. 14.

In the figure, the straight line connecting points A1 and A4 and the straight line connecting points A3 and A6 each have a CF value of "0",
The CF value increases toward the middle of both straight lines. The straight line connecting points A2 and A5 has the maximum CF value of "1", indicating that it has the highest reliability.

As a result of controlling the furnace temperature based on such a three-dimensional function,
As shown in FIG. 15, it can be seen that the frequency of errors occurring is reduced compared to the conventional operator method.

The three-dimensional function shown in Fig. 14 is constructed by connecting the minimum and maximum values of a predetermined range of sensor data with a straight line.
In Figure 16, the sensor data area is expanded and divided into multiple parts depending on the size of the sensor data (in this example, it is divided into three parts).
, a three-dimensional function is constructed by connecting polygonal lines so that each divided region is continuous. The three-dimensional function shown in FIG. 16 shows a more realistic relationship between sensor data, furnace heat level, and CF value (confidence factor) than the three-dimensional function shown in FIG. 14. The maximum value of this CF value is "1" in each divided area, but it goes without saying that the maximum value may be different for each area.

In addition, the sensor data of the above three-dimensional function is the difference data from the reference value (-1TII constant value - target value) because
This is due to the following reasons.

b) The target value of hot metal temperature changes due to component adjustment, etc.

Regarding sensor data, the standards are also changed daily depending on the operational policy, for example, whether or not the system is oriented toward low fuel ratios.

c) Furthermore, depending on the degree of wear of the installed location of the bricks, etc., the temperature detected by the thermometer etc. will differ even if the situation inside the furnace is the same.

For the reasons mentioned above, the three-dimensional function uses the above difference data as a reference, so that it can respond to various changes.

FIG. 17 shows the relationship between action instructions, furnace heat level, and hot metal temperature. The action instruction using the 3-dimensional function in Figure 14 is performed at the tap of ■ at the timing shown by the solid line in the figure, and the action instruction using the 3-dimensional function (broken line) in Figure 16 is performed at the timing shown in A and B in the figure. be exposed.

As a result, in the case of using the three-dimensional function shown in FIG. 16, the molten metal temperature becomes the target value at the next tap. In comparison, the three-dimensional function shown in FIG. 14 deviates slightly from the target value.

Figure 18 shows the results regarding hot metal temperature, Si, and S when the three-dimensional functions in Figures 14 and 16 are used, and the results are better when the three-dimensional functions in Figure 16 are used. Characteristics superior to those shown in FIG. 14 are obtained.

As a result, in the case of using the three-dimensional function shown in FIG. 16, the molten metal temperature becomes the target value at the next tap. In comparison, the three-dimensional function shown in FIG. 14 deviates slightly from the target value.

Figure 18 shows the results regarding hot metal temperature, Si, and S when the three-dimensional functions in Figures 14 and 16 are used, and the results are better when the three-dimensional functions in Figure 16 are used. Characteristics superior to those shown in FIG. 14 are obtained.

By the way, since there is a certain relationship between an action instruction, its response, and its effect, a rule called an "action correction rule" is introduced into the above action correction judgment KS loop D to correct the amount of action. . This action correction rule is a rule used to judge actions taken in the past and their influence amounts, and to be taken into consideration when determining the next action amount. This action correction rule will be explained below.

Figure 19 shows the results of an action response test conducted during the furnace thermal stability period. Although the tuyere embedded temperature (T tu) immediately responds to the action of the air blowing temperature of 308C, there is a delay (dead time) in the response of the gas utilization rate (ηCO) and the air blowing pressure (pb). Both power is 80-90
It stabilizes in minutes. On the other hand, the sensor behavior in actual operation will be different from this example. In other words, it is customary to perform an operation to return to the standard level when a decrease in the furnace heat can be clearly predicted due to a decrease in the tuyere filling temperature or gas utilization rate, so the behavior of the sensor is also based on a certain standard value. controlled within a certain range.

Next, the processing of this "action correction rule" will be explained in more detail based on FIGS. 20 and 21.

FIG. 20 shows the change in air humidity over time, and FIG. 21 is a flowchart showing the operation of the "action correction rule". Note that the air humidity in FIG. 20 is data taken in through a data scanner (lie), and the calculations in the flowchart in FIG. 21 are performed by the calculation means (16).

When calculating the manipulated variable at the current time point Xn in Fig. 20,
First, set the parameter i = 2, and change flags S , , W −
31), 3 points centered on i-2 (XI
,X2. The maximum value of the difference between X3) is determined (S2). The maximum value is compared with the allowable value Xf'M (S3), and if the maximum value is smaller than the allowable value number Xf'M, the average value of the three points is determined (S5).

Next, it is checked whether or not the change flag is 5W-O (S6), and since the change flag is 5W-O here, the above average value is stored as the operating value xb (S7). Then, the content of parameter i was increased by “1” and advanced by 1 minute (i
-i+1-3) (S8), the process returns to the step of finding the maximum value of the difference between the three points (S2).

As is clear from FIG. 20, the changes from X1 to X6 are small, so the above steps are repeated.

Therefore, the average value Xt6 is stored in the operating value xb.

The above calculation operation is repeated in the same way, but in the section where the humidity is changing, the maximum value of the three points becomes larger than the allowable value, and it is detected that the state is changing, so the change flag 5W- 1 (S4), time is advanced again to find the maximum value of the three points.

Then, entering the next unchanged interval, find the maximum value of the difference between the three points (Xi-1, Xi, Xi+1) in the same way as above, and if the maximum value is determined to be smaller than the allowable value,
After calculating the average value of the three points, it is determined that the change flag is not 5W-O (S6), and the absolute value of the difference between the above average value Xtt and the already calculated operating value xb (-X t 6) is the action value. Please check whether it is greater than the minimum value Xd.
9), if it is large, the manipulated variable Xz is changed to Xz−Xtl−
It is determined by Xb (S 10). Next, the current value Xti and change flag SW at this time are set as X b−X ti
, 5W-O (Sll), advance the time by 1 minute again (S8), return to the first calculation (S2), grasp the past operation amount of the ventilation humidity as described above, and take the next one. It is understood as a correction amount when calculating the correct operation amount. This calculation similarly determines past manipulated variables not only for the air humidity, but also for the air temperature, coke ratio, air flow rate, and the like.

The operation amount obtained in the above manner is used for the following comprehensive judgment KS.
It is used in calculation processing using , and the amount of action is determined. FIG. 22 shows a flowchart of the calculation process, and FIG. 23 is an explanatory diagram showing various cases regarding the correction order.

First, calculate the following equation.

Hvl-The amount of action for the 10th time in the correction ranking Here,
The first correction order is the past operation amount of the operating factor that should be controlled first, and if this corresponds to the air humidity in the example of case A in Figure 23, then the above operation m x z and the action amount (output of action determination KS). Then, this Hvl is 1 of the preset correction order.
If the number is smaller than the upper limit value, it is output as the current action amount A.

If Hvl exceeds the upper limit value of the first correction rank or the upper limit value, it is set as: Hvl - the upper limit value of the first correction rank Hv2 - the second number B of the current action amount of the correction rank. In addition, in case A, the second correction order is the past manipulated variable of the air temperature (value corresponding to Xz above).
The remaining B of the current action amount is the amount indicated by the remaining B of the current action amount - the current action amount - the first upper limit of the correction order.

If Hv2 obtained here is smaller than the upper limit value of the second correction order, Hvl and Hv2 are output as the current action amount.

If Hv2 exceeds the upper limit value or the upper limit value of the second correction order, set Hv2 - the upper limit value of the second correction order Hv3 - the remaining amount of action for the third time C of the correction order. In addition, in case A, the second correction order is the past manipulated amount of the coke ratio (the value corresponding to Xz above), and the remaining C of the current action amount is the remaining C- of the current action amount. This is the amount indicated by the remaining B of the current action amount - the second upper limit of the correction ranking.

If Hv3 obtained here is smaller than the upper limit value of No. 3 in the correction order, Hvl to Hv3 are output as the current faction.

If Hv3 is equal to or exceeds the upper limit of No. 3 in the correction order, set Hv3 - the upper limit of No. 3 in the correction order Hv4 - remaining D of the current action amount for No. 4 in the correction order. In addition, No. 4 in the correction order is the past operation amount of the air flow rate (value corresponding to Xz above) in case A.
The above-mentioned remaining amount D of the current action amount is the amount indicated by the remaining D of the current action amount - the remaining C of the current action amount - No. 3 in the correction order. And the above Hvl~
Hv4 is output as the current action amount.

Although the flowchart in FIG. 22 is based on the upper limit of each correction order, the lower limit is also processed using the same flowchart. The action amount obtained in this way is processed based on the operating state KS as described above, and then each control valve is connected to a digital instrumentation device (32
).

FIG. 24 is a characteristic diagram showing the operational results when fully automatic control is performed by controlling the blown water ratio. For example, the action Dl (water ratio -3 g/Nm3) is due to a decrease in the furnace heat level. Also, the action at 15 o'clock (water ratio -2g/Nm3) is C type (no action) at the current time.
However, this was taken in consideration of the fact that the coke ratio -2 kg/T would later fall to the tip of the tuyere.

Comparing the operational results according to this embodiment with the operational results by conventional operators, it can be seen that the accuracy of controlling the hot metal temperature has been improved as shown in FIG. 25. This means that operational management has been standardized.

[Effects of the Invention] As described above, according to the present invention, processed data is created from various data installed in a blast furnace, and predetermined inferences as artificial intelligence are made using the processed data and a knowledge base based on experience. As a result, conventional experience is fully utilized, standardizing operational management, preventing human errors in judgment, stably supplying high-quality hot metal to the next process with little temperature and composition fluctuations, avoiding furnace cooling, and saving labor. has been realized.

In addition, when the control device is implemented using a computer, it is based on a knowledge base, which improves calculation capacity and speed, and makes it easy to add and modify rules to accommodate new situations such as aging of the blast furnace. It has become.

Since a belief function is used to obtain each furnace heat level and each furnace heat transition CF value, the number of rules and the work of inputting their numerical values are reduced.

Furthermore, since the next action amount is determined in consideration of past operational results, control corresponding to the actual furnace can be performed, and therefore highly accurate control is possible. In particular, since it takes operational results into consideration, it is effective, for example, when offline control (manual control) is being performed and then switching to online control.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of the present invention, FIG. 2 is a block diagram of a blast furnace furnace heat control device according to an embodiment of the present invention,
FIG. 3 is a flowchart showing the knowledge base and its inference order, FIG. 4 is a characteristic diagram showing a three-dimensional function of the furnace heat level, and FIG. 5 is an explanatory diagram explaining the inference operation. Figure 6 is a characteristic diagram showing the three-dimensional function of the feather transition level.
FIG. 7 is an explanatory diagram showing an example of an action matrix, FIG. 8 is an explanatory diagram showing an example of an action matrix type, and FIG. 9 is an explanatory diagram showing an example of an action amount. FIG. 10 is a characteristic diagram showing the transition of hot metal temperature in the tap, and FIGS. 11 to 13 are explanatory diagrams showing a method of creating a three-dimensional function of the furnace heat level. Fig. 14 is an explanatory diagram showing an example of the three-dimensional function, Fig. 15 is a characteristic diagram showing the results of the three-dimensional function in Fig. 14, and Fig. 16 is an explanatory diagram showing another example of the three-dimensional function. , Fig. 17 is a characteristic diagram showing the relationship between control action instructions, furnace heat level, and hot metal temperature using the three-dimensional function shown in Fig. 14 and the three-dimensional function shown in Fig. 16, and Fig. 18 is the three-dimensional function shown in Fig. 14. FIG. 17 is a characteristic diagram showing the distribution of temperature difference, St, and S with respect to the target temperature, showing control results using the function and the three-dimensional function of FIG. 16; Figure 19 is a characteristic diagram showing the results of an action response test conducted when the furnace is stable, Figure 20 is a characteristic diagram showing changes in blast humidity, Figure 21 is a flowchart showing the operation of correction calculation, and Figure 22 is an action diagram. A flowchart showing the calculation operation of the output rule, and FIG. 23 is an explanatory diagram showing the priority order at the time of correction. FIG. 24 is a characteristic diagram showing the results of an operation example of fully automatic control using blowing water ratio side control, and FIG. 25 is a characteristic diagram showing the error occurrence rate of hot metal temperature. (11), (12); Data input means (1B), (
14); Processed data creation means (19), (24); Storage means (22), Knowledge base storage means (2B), Reasoning means agent Patent attorney Souji Sasaki, Souji Ward, Sou-ku, Sho Figure 9, 1Q Figure Time + rninl Figure 11 Figure 12 Figure 17 Figure 8 b +”7.1 Figure 19 □ Time (minl Figure 20 Figure 21 Figure 23 Case-A Case-B Case-〇
Case-D Figure 25 Error rectangle of molten iron temperature CT a, c Euo I-T 101) 1) L (l l) L9

Claims (1)

[Claims] (1) A data input means that takes in data from various sensors installed in the blast furnace at a predetermined timing; A processing data creation means that creates various data indicating the status of the blast furnace, such as utilization rate, solution loss amount, etc., and compares the various data with reference data to create difference data; and the various data and difference data. (hereinafter referred to as processed data)
a knowledge base storage means that stores various knowledge bases based on experience, results, mathematical models, etc. regarding blast furnace operation; and processed data in the storage means and knowledge in the knowledge base storage means. an inference means for inferring a furnace heat level and a furnace heat transition based on the base and determining an action amount for the blast furnace; and a means for outputting the action amount, and the knowledge base storage means estimates the furnace heat level. A furnace heat level determination knowledge base used for estimating the furnace heat level transition; a furnace heat transition determination knowledge base used for estimating the furnace heat transition;
An action judgment knowledge base that is used to determine the amount of action based on the furnace heat level and furnace heat transition, an action correction knowledge base that determines the correction value of the amount of action based on past actions and disturbances, and the amount of action and action. a comprehensive judgment knowledge base that determines the actual action amount from the correction amount, and the furnace heat level judgment knowledge base and the furnace heat transition judgment knowledge base use processing data and the furnace heat level as independent variables,
The action determination knowledge base includes a confidence function whose dependent variable is the probability that these combinations occur (hereinafter referred to as confidence), and a rule group that determines how to apply this confidence function. The comprehensive judgment knowledge base has a rule for determining the amount of action based on the level and the furnace heat transition, and the comprehensive judgment knowledge base has a rule for determining the actual amount of action from the amount of action and the amount of action correction, and the action correction knowledge base further includes: A blast furnace furnace heat control device characterized by having an action correction rule used to determine an action correction amount by determining an action taken in the past and its influence amount.