JPH05156328A - Device for analyzing abnormal furnace condition in blast furnace - Google Patents

Abstract

PURPOSE:To provide an analizing device for abnormal furnace condition in a blast furnace analyzing precisely a developing cause of the abnormal furnace condition, such as blowing-off, slippage CONSTITUTION:The device is provided with an abnormal furnace condition detecting means 12 for detecting the development of the abnormal furnace condition in the blast furnace based on the furnace top pressure and sounding level and a data collecting means 11 for collecting and storing in a prescribed period by optionally selecting a sensor data related to the abnormal furnace condition from various kinds of the sensors set in the blast furnace and stopping the collection of the data from the detection of the development of the abnormal furnace condition with the abnormal furnace condition detecting means and outputting the collected sensor data before detection and a data analyzing means 13 for inputting the sensor data from the data collecting means and analyzing the furnace condition from these.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a blast furnace abnormal furnace condition analyzing apparatus for analyzing a furnace condition when an abnormal furnace condition such as blow-through or slip occurs.

[0002]

2. Description of the Related Art Abnormal furnace conditions in a blast furnace, such as blow through and slip, are phenomena that occur in a very short period of time. Moreover, the duration of these phenomena is several seconds, and the time from the appearance of a sign in the sensor information to the occurrence of these phenomena is also several minutes.
Up to now, as a method of predicting the occurrence of these, JP-A-59-64705 or JP-A-62-270712.
There is a method disclosed in the publication. JP-A-59-64
According to the method disclosed in Japanese Patent No. 705, sensor data that is known to be particularly useful for predicting abnormal reactor conditions from operating experience is selected from the sensor data, and these are associated with phenomena in the reactor to organize and quantify them. At the same time, the summarized and quantified results are detected from both short-term and long-term aspects of the furnace condition of the blast furnace for furnace condition management. Also, JP-A-62-27071
In the method disclosed in Japanese Patent Laid-Open No. 59-64705, instead of the statistical method used for analyzing sensor data in Japanese Patent Laid-Open No. 59-64705, AI technology is used to utilize the knowledge of the operator to determine the state of the furnace condition of the blast furnace. Is detected to control the furnace condition.

[0003]

Problems to be Solved by the Invention JP-A-59-6470
In the method of Japanese Patent Laid-Open No. 5, the sensor data obtained by collecting the sensor data in 1-minute cycles is statistically processed, and in the method of Japanese Patent Laid-Open No. 62-270712, the sensor data is knowledge processed to detect abnormalities in the blast furnace. It is intended to predict the occurrence of the furnace condition, and therefore, the sensor data of one minute cycle does not help to clarify the abnormal reactor condition such as blow-through and slip that occur in a short period of time. Therefore, it has been desired to develop an apparatus for collecting and accumulating sensor data before occurrence of abnormal furnace conditions at high speed and analyzing it over time.

The present invention collects sensor data relating to abnormal furnace conditions such as blow-through and slip at high speed and cyclically stores them, and stops the collection of sensor data when it is detected that these phenomena occur. By analyzing the collected sensor data, we investigated the cause of abnormal furnace conditions and established operating techniques that would prevent abnormal reactor conditions.
It is an object of the present invention to provide a blast furnace abnormal furnace condition analysis method capable of maintaining and improving stable operation of the blast furnace.

[0005]

A blast furnace abnormal furnace condition analysis apparatus according to the present invention detects an abnormal furnace condition of a blast furnace such as blow-through or slip based on the furnace top pressure and sounding level. Means and various sensors installed in the blast furnace to collect sensor data related to abnormal furnace conditions arbitrarily and collect them in a predetermined cycle, and when abnormal furnace condition detection means detects the occurrence of abnormal furnace conditions, data collection And a data collection means for outputting sensor data collected before the detection,
And data analysis means for inputting sensor data from the data collection means and analyzing the reactor condition accordingly.

[0006]

In the present invention, the data collecting means collects preselected sensor data at high speed and cyclically saves it in a file. When the abnormal furnace condition detecting means detects the occurrence of the abnormal furnace condition, the data collecting means stops collecting the sensor data. Then, the data analysis means performs processing such as statistical analysis on the sensor data at this time to elucidate the cause of occurrence of an abnormal furnace condition such as blow through or slip. By accumulating this analysis, the cause of the abnormal reactor situation is clarified,
This can lead to the establishment of operating technology that prevents abnormal furnace conditions.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram showing the configuration of a blast furnace abnormal reactor condition analysis apparatus according to an embodiment of the present invention. This blast furnace abnormal furnace condition analyzer collects sensor information at high speed and stores it cyclically, a data collecting device 11, an abnormal furnace condition detecting device 12 for detecting an abnormal furnace condition, a cause of the abnormal furnace condition, etc. The data analysis device 13 for analysis is provided. A CRT terminal (man-machine interface) 14 is attached to the data analysis device 13, and each has the following functions.

The data collection device 11 is a data analysis device 1.
According to the command from 3, the sensor data regarding the preselected abnormal furnace conditions such as the shaft pressure, the blast pressure, the furnace top gas pressure, the sounding level, etc. are started to be collected in a cycle of 100 ms and are cyclically stored as the sensor data. When an abnormal furnace condition detection signal is input from the abnormal furnace condition detection device 12, the collection of sensor data is stopped and the collected sensor data is transmitted to the data analysis device 13. In addition, the reason why the data collection is stopped by the abnormal reactor condition detection signal is that the data collection for one cycle is performed for the purpose of high-speed data collection. This is because the time is shorter than the time for transmitting to the device 13. Further, the sensor data is selected in advance and the data analysis device 13 is separately provided in order to collect data at high speed.

Various sensors are installed in the blast furnace, and the sounding level meter 21 continuously measures the distance to the surface of the raw material at the furnace top at a plurality of points in the circumferential direction. The furnace top gas pressure gauge 22 continuously measures the gas pressure in the furnace at the top of the furnace. A plurality of fixed sonde thermometers 23 are installed at equal intervals in the circumferential direction on the upper part of the shaft to continuously measure the gas temperature in the furnace. A plurality of shaft pressure gauges 24 are installed at equal intervals in the circumference of the upper part of the shaft and in the vertical direction, and continuously measure the gas pressure in the furnace. Blast pressure gauge 25
Continuously measures the pressure of hot air when blowing for combustion of coke from the tuyere at the bottom of the furnace.

The abnormal furnace condition detection device 12 preselects a sensor for high-speed data collection via the data analysis device 13 by the setting of the operator. Then, at the same time when a high-speed data collection start instruction is output to the data collection device 11 according to an operator's instruction, high-speed monitoring of the output of the furnace top gas pressure gauge 22 is started, and when the pressure suddenly rises, a blow-through occurs. I judge that I did. Also, high-speed monitoring of the output of the sounding level meter 21 is started, and when the level suddenly drops, it is judged that slip has occurred. And
When the occurrence of blow-through or slip is detected, the data collection device 11 is notified as an abnormal furnace condition detection signal. Also, when an abnormal furnace condition is detected, monitoring of the furnace top gas pressure and sounding level is stopped.

When the abnormal reactor condition detecting device 12 notifies the data collecting device 11 of the abnormal furnace condition detecting signal, the data analyzing device 13 receives the sensor data collected by the data collecting device 11 at that time, and outputs the analysis file. To store. Then, the sensor data is used to perform data analysis by dialogue with the operator, and the analysis result is notified to the operator.

FIG. 2 is a diagram showing a method of detecting blow-through and slip by the abnormal furnace condition detection device 12. Blow-through means a phenomenon in which the reducing gas pushes up more than the weight of the charging object and the reducing gas reaches the top of the furnace from the bottom of the furnace at a stroke. When the furnace top gas pressure exceeds a certain value as shown in 2 (a), it is defined as blow through. Generally, the time during which the gas pressure at the top of the furnace rises when the blow-through occurs is 2 to 6 seconds, and before the occurrence of the blow-through, there are signs of changes in the gas components at the top of the furnace and changes in the shaft pressure from several minutes before. In addition, slip means a phenomenon in which the force of pushing up the reducing gas weakens and the raw material falls locally. When a slip occurs, the sounding bell drops rapidly, so as shown in Fig. 2 (b), the sounding level Slip is defined as the amount of change (falling amount) in which the value drops by more than a certain value. In general, signs of fixed sonde temperature and shaft temperature also appear before slip occurs. In FIG. 2B, it is the charging timing of the raw material that the amount of change suddenly rises in a certain cycle.

FIG. 3 is a diagram showing an example of analysis of a blow-by phenomenon by the abnormal furnace condition detection device 12. It was found from this example that the blow-through phenomenon has a strong correlation with the fluctuation (σ) of the shaft pressure. That is, when the fluctuation (σ) of the shaft pressure is small, blow-through is unlikely to occur, and when the fluctuation (σ) of the shaft pressure grows in a certain cycle, blow-through easily occurs.

FIG. 4 is a diagram showing an example of analysis of the slip phenomenon by the abnormal furnace condition detection device 12. It was found from this example that the slip has a strong correlation with the fluctuation (σ) of the fixed sonde temperature. That is, when the variation (σ) of the fixed sonde temperature is small, the slip is unlikely to occur, and when the variation (σ) of the fixed sonde temperature is large, the slip is likely to occur.

FIG. 5 is a flow chart showing the operation of the apparatus of FIG. 1, and the operation will be described with reference to this figure. (S1) From the CRT terminal (man-machine interface) 14 of the data analyzer 13 to the data collector 11 at high speed (1
00 ms) to specify a sensor for collecting data, such as shaft pressure. (S2) The sensor code designated in (S1) above is sent to the data collection device 11, and a sensor for data collection is selected in the data collection device 11. (S3) The CRT terminal 14 of the data analysis device 13 gives an instruction to start data collection. (S4) The start instruction of (S3) is sent from the data analysis device 13 to the data collection device 11, and the data collection device 11 starts collecting sensor data designated in advance. (S5) Simultaneously with the above (S4), the abnormal furnace condition detection device 12 takes in the top gas pressure and the sounding level, and determines blow-through and slip. This determination process is performed, for example, in a 1-minute cycle.

(S6) The data collecting device 11 cyclically stores data such as shaft pressure collected at high speed.
For example, it is saved at a cycle of 100 msec. (S7) If the abnormal furnace condition detection device 12 detects the occurrence of blow-through or slip in the process of (S5), the data collection device 11
Instruct to stop the data collection. Further, the abnormal furnace condition detection device 12 stops taking in the furnace top gas pressure and the sounding level for determining blow-through and slip. (S8) Upon receiving the stop instruction in (S7), the data collection device 11 stops data collection, transmits the collected data to the data analysis device 13, and initializes the data saved in the cycle. (S9) The data analysis device 13 stores the data received from the data collection device 11 in the analysis file. (S10) The data analysis device 13 analyzes the data by interacting with the operator using the data stored in the analysis file.

FIG. 6 is a flow chart showing the data analysis processing in the above (S10). (S11) According to the designation from the CRT terminal 14, specific analysis data is taken out from the analysis file. At this time, the data (1
It is assumed that a plurality (10 points / 100 ms × 20 minutes) are stored. (S12) Specify the sensor, analysis method, etc. for data analysis from the CRT terminal (i) Specify sensor: shaft temperature, shaft pressure, furnace top pressure ... (ii) Analysis method: linear regression, exponential smoothing, average difference (Iii) Analysis unit: n seconds (m points) unit (iv) Calculated value: / X, σ ... (v) Displayed value: Raw data, primary regression, / X, σ ...

[0018]

[Equation 1]

(S13) The sensor designated by the instruction from the CRT terminal 14 is analyzed by the designated method, and the analysis result is displayed as a transition diagram on the CRT. FIG. 7 shows an analysis procedure of the data analysis device 13 when designated as follows. Sensor designation: 5th stage average of shaft pressure Analysis method: Exponential average, (smoothing constant α) Analysis unit: 10 seconds (60 points) unit Calculated value: / X, σ display value: Raw data (however, 5th stage average) , Exponential average, / x, α (S14) First, of the data extracted from the analysis file, the average value of the fifth stage of the shaft pressure is obtained.

[0020]

[Equation 2]

(S15) Exponential smoothing is performed according to the instruction of the analysis method.

[0022]

[Equation 3]

(S16) 10 seconds (60 seconds) according to the analysis unit (10 seconds (60 points) unit) and calculation value (/ X, σ)
/ X and σ are obtained with the data of point) as a unit.

[0024]

[Equation 4]

FIG. 8 shows the state of the data at this time. From the data of 60 points for 10 seconds during the time Tm,
The value of one point of X and σ is created. (S17) The analysis result is displayed on the CRT according to the display value instruction. FIG. 9 is a diagram showing a display example at this time. (S18) Here, returning to the flowchart of FIG. 6 again, it is judged whether or not the analysis result of (S16) is good, and if necessary, the process returns to (S12) and repeats the above processing. Here, the determination of whether or not the analysis result is good is made based on an empirical rule. (S19) When the analysis result is good and the analysis result is necessary, the necessary data is output to the printer or the hard copy of FIG. 9 is taken according to the instruction from the CRT terminal 14.

FIG. 10 is a diagram showing the operation results. The cause of blow-through and slip has been clarified by the above-described embodiment, and the certainty factor used in Japanese Patent Laid-Open No. 62-270712, that is, blow-through, is the fluctuation of the shaft pressure (σ).
The slip combining weight of the certainty factor increases the prediction accuracy of the blow-through and the slip by increasing the weight connecting the certainty factor of the fixed sonde temperature fluctuation (σ), and as shown in FIG. The number of occurrences has decreased. This made the operation of the blast furnace more stable than before.

[0027]

As described above, according to the present invention, the cause of the abnormal furnace condition such as blow-through or slip is clarified based on the sensor data when the occurrence of the abnormal furnace condition is detected. The cause of the reactor situation is clarified, and
Prediction accuracy of abnormal furnace conditions is improved, the number of occurrences is reduced,
Stable blast furnace operation can be realized.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration diagram of a blast furnace abnormal furnace condition analysis apparatus according to an embodiment of the present invention.

FIG. 2 is a view showing a detection method of the abnormal furnace condition detection device of FIG.

3 is a diagram showing an example of blow-through analysis by the abnormal furnace condition detection device of FIG.

FIG. 4 is a diagram showing an example of slip analysis by the abnormal furnace condition detection device of FIG. 1.

5 is a flowchart showing the operation of the apparatus of FIG.

FIG. 6 is a flowchart showing a data analysis process in FIG.

7 is a flowchart showing a data analysis process in FIG.

8 is a diagram showing a data state of the analysis process of FIG.

FIG. 9 is a diagram showing an example of CRT display of analysis results.

FIG. 10 is an operation transition diagram of the blast furnace of the above embodiment.

[Explanation of symbols]

 11 data collection device 12 abnormal furnace condition detection device 13 data analysis device

Claims (1)

[Claims] 1. An abnormal furnace condition detection means for detecting the occurrence of an abnormal furnace condition of a blast furnace based on a furnace top pressure or a sounding level, and sensor data related to the abnormal furnace condition from various sensors installed in the blast furnace. Data collection means that arbitrarily collects data in a predetermined cycle, stops collection of data when the abnormal furnace condition detection means detects occurrence of abnormal furnace condition, and outputs sensor data collected before the detection. A blast furnace abnormal reactor condition analysis apparatus, comprising: a data analyzing device for inputting sensor data from the data collecting device and analyzing the reactor condition by the sensor data.