JPH04329809A - Blast furnace operation support equipment - Google Patents

Abstract

PURPOSE:To provide an assisting apparatus for blast furnace, in which information from various kinds of sensors in the blast furnace are processed and evacuated to obtain the operational control standard values. CONSTITUTION:The assisting apparatus for blast furnace is provided with a blast furnace condition evaluation means 5 for processing data from each sensor 1 in the blast furnace body and evaluating the blast furnace condition, a data storing means 6 for temporarily storing various kinds of processed data and the blast furnace condition evaluated values, a means 7 for calculating correlation between various kinds of data and the blast furnace condition evaluated values, and a means 8 for preparing the operational control standard values by using the blast furnace condition evaluated values, an the basis of the values obtd. from both of an average value calculating means, where the values have a correlation comparable to or beyond the standard values with the blast furnace condition evaluated values, and the various kinds of data showing the blast furnace condition. By this method, as the data in sensor informations are renewed every day, the control standard values corresponding to variation of the used raw material and variation of the operational condition can be obtd.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a blast furnace operation support system that processes and evaluates information from various sensors of a blast furnace and obtains operational control reference values.

0002

[Prior Art] Conventionally, as a method for diagnosing and managing the condition of a blast furnace, blast furnace operators qualitatively judge information from various sensors installed in the blast furnace to evaluate the condition of the blast furnace. , a method of optimally adjusting operational factors has been adopted, but the results of this evaluation vary depending on the ability and experience of the operator, making it difficult to standardize operational actions, and the evaluation is not quantitative. Therefore, there was a problem that it was difficult to perform operational analysis.

[0003] For this reason, for example, Japanese Unexamined Patent Application Publication No. 1983
A method for detecting the condition of a blast furnace is proposed in Japanese Patent No. 64705. This involves determining indices (burden charge descent status index, pressure fluctuation status index, etc.) based on information from sensors installed in the blast furnace, comparing these index values with limit values, graded evaluation into several categories, weighting, and After calculating by addition, the obtained index is added after weighting to obtain the furnace condition diagnosis index, and the long-term information of multiple factors corresponding to each index such as the gas flow distribution index is converted into an index. The long-term furnace condition diagnosis index is determined by comparing the value with the target value, graded evaluation into several categories, weighting, and addition, and calculating the long-term furnace condition diagnosis index by weighting and adding these indices and the above-mentioned furnace condition diagnosis index. This is to detect the long-term condition of the reactor.

0004

[Problems to be Solved by the Invention] However, the method for detecting the blast furnace status disclosed in Japanese Patent Application Laid-Open No. 59-64705 does not detect changes in the raw materials used in the blast furnace or changes in operating conditions (for example, changes from oil coke operation to pulverized coal injection operation). It is necessary to maintain the limit value of the index value due to changes in the furnace life (changes in There is a problem.

[0005]

[Means for Solving the Problems] The present invention attempts to solve the above-mentioned problems, and provides a data input means for taking in data from various sensors installed in a blast furnace at a predetermined time, and means for calculating an average value of the sensor data; means for creating various data indicating the status of the blast furnace such as unloading speed, pressure loss, gas utilization rate, etc. based on the sensor data; and a value obtained by the average value calculation means. and a blast furnace condition evaluation means for evaluating the blast furnace condition from various data indicating the blast furnace condition, and a value obtained from the average value calculation means,
Means for temporarily storing various data indicating the blast furnace condition and the evaluation value obtained from the blast furnace condition evaluation means; means for calculating the correlation between the two, a means for extracting data having a correlation greater than a reference value with respect to the blast furnace condition evaluation value from the obtained correlation, and an average having a correlation greater than the reference value with the blast furnace condition evaluation value. The blast furnace operation support device is characterized by comprising means for creating an operation management reference value using a blast furnace condition evaluation value regarding the value obtained from the value calculation means and various data indicating the blast furnace condition.

[0006]

[Operation] Measured data from various sensors installed in the blast furnace is processed by average value calculation means and various data calculation means into data necessary for evaluating furnace conditions such as charge descent rate and pressure loss. The blast furnace condition evaluation means provides a furnace condition diagnosis index through stage evaluation, weighting based on empirical rules, etc. The correlation calculation means analyzes the sensor information (average value of measured values, processed data) and the furnace condition diagnostic index, and determines the cross-correlation and maximum correlation function between the sensor information and the condition index. The cross-correlation and maximum correlation function between the sensor information and the condition index are obtained by the control reference value creation means, and from these, the control reference value for each sensor information item for maintaining a stable reactor condition is obtained.

[0007] Since the sensor information is updated every day, even if there are changes in the raw materials used, operating conditions, etc., management reference values corresponding to the changes in the situation can be obtained.

[0008]

EXAMPLES Examples of the present invention will be described in detail below. FIG. 1 is a block diagram of the apparatus of the present invention. In the figure, measurement data from each sensor (sonde, thermometer, pressure gauge, etc.) 1 installed in the blast furnace is sent to average value calculation means 3 via data input means 2. The average value calculation means 3 calculates a predetermined time period (for example, 1 minute, 1 hour, 1 day, etc.) of each measurement data.
The average value for each time is calculated and stored in the temporary storage means 6. On the other hand, the measurement data of each sensor is collected by various data creation means 4.
The data is processed into necessary data such as charge descending speed, pressure loss, gas utilization rate, etc., and is stored in the temporary storage means 6.

[0009] The blast furnace condition evaluation means 5 is a means for standardizing the blast furnace condition and quantitatively evaluating the furnace condition, and it calculates the average value of each measurement data obtained by the average value calculation means 3 and various data creation means. The processing data obtained in step 4, such as charge descent speed, pressure loss, and gas utilization rate, are compared with predetermined limit values or target values, and each data is divided into stages according to the comparison results. Perform an evaluation (e.g., 5-level evaluation). Then, each evaluation index is weighted based on empirical rules, and added together for each factor (type of data) that represents the burden descent situation, pressure fluctuation situation, etc., and is evaluated on a scale of 10 points, for example. . Through this evaluation, a charge descent condition index, a pressure fluctuation condition index, a gas flow fluctuation condition index, a circumferential balance condition index, and a furnace heat fluctuation condition index, which are evaluation factors for furnace condition diagnosis, are obtained. Then, the burden drop condition index, pressure fluctuation condition index, gas flow fluctuation condition index, circumferential balance condition index, and furnace heat fluctuation condition index are weighted and added, and for example, the furnace condition is diagnosed by a 50-point evaluation. Calculate the index (blast furnace condition evaluation value). Then, the obtained furnace condition diagnostic index (blast furnace condition evaluation value) is stored in the temporary storage means 6. An example of classification of factors (types of data) that serve as the basis for furnace condition diagnosis and each evaluation is as follows.

[0010] Furnace condition diagnosis Evaluation of charge descent status ・Charge descending speed σ ・Charging interval σ ・Slip strength Pressure fluctuation status evaluation ・Blowing pressure σ ・Shaft pressure σ ・Abnormal furnace top pressure fluctuation Gas flow fluctuation status evaluation ・Gas composition σ (ηco) ・Below bell sonde σ ・Furnace top temperature σ Circumferential balance status evaluation ・Stock line level σ ・Furnace top temperature σ ・Si concentration σ in hot metal Furnace heat fluctuation evaluation ・Siσ in hot metal ・Tueyer embedding temperature σ

The correlation calculation means 7 takes out the average value of each measurement data for a predetermined period, the processed data, the condition index, and the furnace condition diagnosis index (blast furnace condition evaluation value) from the temporary storage means 6, and performs a trend analysis and a correlation analysis. I do.

(1) Trend analysis Trend analysis is performed individually on all sensor information (average value of each measured value, processed data), and the analysis is performed using the least squares method using data from the last n days (setting). A linear regression line (Y=a+bX) is obtained. In addition, the correlation coefficient (γ) is determined for each analysis.

(a) Least squares method Given data xi: past 10 days from the current day,
An equation of the form Y=a+bX is obtained from the actual value corresponding to yi:xi using Equations 1, 2, and 3.

[0014]

[Math 1]

[0015]

[Math 2]

From equations 1 and 2,

[Math 3]

(b) Correlation coefficient: γ It is determined by Equation 4, Equation 5, and Equation 6.

[0018]

[Math 4]

[0019]

[Math 5]

[0020]

[Math 6]

(c) Output method All data are displayed on the CRT screen 9 or printer 10 in the format shown in FIG. The title and Y-axis value are items for each data analysis, and after plotting 10 days' worth of data, 1
Next regression line Y=a+bX, γ value is displayed.

(2) Correlation analysis sensor information and six status indexes (burden charge descent status index, pressure fluctuation status index, gas flow fluctuation status index, circumferential balance status index, furnace heat fluctuation status index, furnace status diagnosis index) The cross-correlation and maximum correlation function are calculated for each set value of the number of data n and the number of deviations (difference in days) τ set by the operator.

(a) For the set values of the number n of cross-correlation data and the number of deviations τ, read the sensor information and situation index for the past n days from the current day. Next, the comparison data obtained by changing (plus) the situation index from 0 to τ with respect to the set value of τ is plotted. An example of the format is shown in Figure 3. Using the least squares method for the plotted data, Y=a+
Find bX.

FIG. 4 shows an example of the correspondence between the sensor information and the situation index in the case of n=15 and τ=2.

[0025]

[Math 7]

[0026] γ(τ) from 0 to τ is determined using Equation 7, and the maximum value thereof is set as the maximum correlation function γmax.

(b) Output method Figure 3
Style graph, linear regression line and maximum correlation function γmax
Display. γmax of all data for each situation index name
, linear regression line, best 1 of γmax in sensor information
0 ranking is performed and printed out using the printer 10 in the format shown in FIG. The management reference value creation means 8 determines the cross-correlation between the sensor information and the situation index, the maximum correlation function, and the stable operation management reference value.

(1) Processing details (difference analysis) (a) The operator inputs the data correspondence and situation index for which correlation is to be determined. (Example Figure 6) (b) Set the number of data (n) and number of deviations (τ) for the input situation index item and sensor information. (c) Perform correlation analysis on the items set in (a) and (b).

(2) Output contents (a) For one item of screen output sensor information of CRT9, γmax of selected item NO.
For example, if there are three items, the graph will be displayed on three sides. Display contents are data plot, linear regression line (Y=
a+bx), γmax, and the x value (sensor information) for the reference value (condition index for maintaining stable reactor conditions). (Example: Figure 7) (B) When displaying the form output screen on the printer 10, find the maximum x value for the reference value for each sensor information item, and use that value as the furnace condition stability control reference value, as shown in Figure 8. indicate.

[0030]

[Effects of the Invention] Since the present invention is configured as described above, a management reference value for maintaining a stable furnace condition can be obtained for each sensor information item for maintaining a stable furnace condition.
Sensor information is updated every day, so even if there are changes in raw materials used or operating conditions, the effect is that management standard values can be obtained that correspond to changes in the situation.

[Brief explanation of drawings]

FIG. 1 is a block diagram of the device of the present invention.

[Figure 2] CRT of trend analysis results by correlation calculation means
It is a figure which shows the example of a screen display.

[Figure 3] C of cross-correlation analysis results by correlation calculation means
It is a figure which shows the example of RT screen display.

FIG. 4 is a diagram showing an example of how sensor information and situation indexes correspond to days in a correlation calculation means.

FIG. 5 is a diagram illustrating an example of a printer's output of a cross-correlation analysis result for each situation index name by a correlation calculation means.

FIG. 6 is a diagram illustrating an example of inputting data correspondence and situation index reference values for determining correlation in management reference value creation means.

FIG. 7 is a diagram showing an example of a CRT screen display of the cross-correlation analysis results by the management reference value creation means.

FIG. 8 is a diagram illustrating an example of a printer output of a furnace condition stability reference value by a management reference value creation means.

[Explanation of symbols]

1 Furnace body sensors 2 Data input means 3 Average value calculation means 4 Various data creation means 5　Blast furnace condition evaluation means 6 Data storage means 7 Correlation calculation means 8 Management standard value creation means 9 CRT 10 Printer

Claims (1)

[Claims] 1. Data input means for taking in data from various sensors installed in a blast furnace at a predetermined time, means for calculating an average value of sensor data for each predetermined time, and calculating the unloading speed and pressure based on the sensor data. A means for creating various data showing the status of the blast furnace such as loss and gas utilization rate;
Blast furnace condition evaluation means for evaluating the blast furnace condition from the value obtained by the average value calculation means and various data indicating the blast furnace condition, and the value obtained by the average value calculation means, various data indicating the blast furnace condition, and blast furnace condition evaluation. means for temporarily storing the evaluation value obtained by the means; means for calculating the correlation between the values obtained from the stored average value calculation means and various data indicating the blast furnace condition and the blast furnace condition evaluation value; means for extracting data having a correlation greater than a standard value with respect to the blast furnace condition evaluation value from the correlation obtained, and means for calculating an average value having a correlation greater than the reference value with the blast furnace condition evaluation value; 1. A blast furnace operation support device comprising means for creating operation management reference values using blast furnace condition evaluation values regarding various data indicating conditions.