A method for judging the tapping state of a blast furnace ladle

Abstract

The application discloses a judgment method of the tapping state of a blast furnace ladle, and belongs to the technical field of steel production. The method comprises the following steps: step one, building a liquid level automatic monitoring and early warning system; step two, performing Kalman filtering processing on the obtained liquid level data, performing real-time denoising and prediction, and judging the tapping state according to the processed liquid level data; and step three, performing early warning when the liquid level of the molten iron is too high and reaches the set filling amount warning point, and recording the alarm time under the conditions of the tapping start, the tapping end and the high liquid level at the same time of monitoring the tapping state. The application realizes automatic monitoring and automatic judgment of the tapping state, can monitor the liquid level in the ladle in real time and accurately, effectively solves the problems of time waste and low production efficiency caused by judgment errors and interference compared with the traditional manual judgment method, and improves the full ladle rate and use efficiency of the ladle.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of steel production, and more particularly relates to a judgment method for the tapping state of a blast furnace ladle. BACKGROUND

[0002] The ladle is a device for containing and transporting molten iron. During the tapping of the blast furnace, the liquid level of the ladle needs to be detected to control the filling amount of the tapping ladle within a safe range, and to improve the utilization rate of the molten iron ladle as much as possible under the premise of safety. At present, when the molten iron ladle of the steel blast furnace is filled with molten iron, the liquid level of the molten iron in the ladle is often judged by manual visual observation. However, due to the high temperature of the molten iron and the surrounding area of the blast furnace tapping hole, and the large amount of dust diffused around the blast furnace tapping hole, the visibility around the blast furnace tapping hole is low, which affects the accuracy of visual observation.

[0003] It can be seen that manual field observation not only brings dangerous operation risks to the work of field workers and hidden dangers to their health, but also is not conducive to manual close-range observation. In addition, the low visibility around the blast furnace tapping hole affects the accuracy of visual observation. Therefore, the method of manually observing the liquid level of the molten iron in the ladle is neither safe nor accurate, and it is easy to cause molten iron overflow or low utilization rate of the molten iron ladle.

[0004] To solve the above problems, Chinese patent application No. 202210668184.3, filed on June 14, 2022, and entitled "A liquid level monitoring device and system for a molten iron ladle", is provided. The liquid level monitoring device for the molten iron ladle includes a radar level meter, a dust removal device, and a cooling device. The radar level meter and the dust removal device are both installed inside the cooling device. The radar level meter is fixedly installed on the dust removal device, and the dust removal device is fixedly installed on the base inside the cooling device. The application monitors the liquid level by setting the radar level meter, which improves the accuracy and safety compared to manual visual observation. However, during the tapping of the blast furnace molten iron ladle, the molten iron is in a high-temperature state, and the molten iron in the ladle does not behave like water, but constantly rolls and fluctuates. Therefore, although the on-site setting of the level meter to measure the liquid level has the advantages of safety, accuracy, and convenience compared to manual observation, the liquid level data read on site still fluctuates greatly. The fluctuation of the data means that it is difficult for the system to automatically determine the current tapping state of the molten iron ladle based on the read liquid level. The tapping of the molten iron ladle is complex and variable in the field production. In addition to the typical process of starting tapping from an empty ladle to a full ladle, there are also various situations such as stopping tapping halfway, waiting for several minutes after tapping halfway, and continuing tapping. In addition, the liquid level fluctuates greatly, and it is necessary to develop a new method to automatically analyze the current tapping state of the molten iron ladle based on the real-time reading of the liquid level without human intervention. SUMMARY

[0005] 1. Problem to be solved

[0006] The purpose of the present application is to provide a method for judging the tapping state of a blast furnace ladle, which can process and analyze the detected tapping liquid level data to judge the production required information such as tapping start, tapping end, and tapping alarm time, so as to solve the problems in the background art such as inaccurate liquid level height judged by artificial, complex on-site production, large fluctuation of liquid level data read by instrument, and difficulty in judging the current tapping state according to the fluctuating liquid level data.

[0007] 2. Technical solution

[0008] In order to solve the above problems, the technical scheme adopted by the present application is as follows:

[0009] The method for judging the tapping state of a blast furnace ladle comprises the following steps:

[0010] Step one, a liquid level monitoring and early warning system is built to monitor and early warn the liquid level height of molten iron in the ladle in real time;

[0011] Step two, the obtained liquid level data is subjected to Kalman filtering processing for real-time denoising and prediction, and the tapping state is judged according to the processed liquid level data, and the liquid level lower than the low liquid level setting value is empty ladle state, the liquid level continuously rising is tapping state, and the liquid level reaching the high liquid level setting value is full ladle state;

[0012] Step three, early warning is performed when the molten iron liquid level is too high and reaches the set filling amount alarm point, and the tapping start, tapping end, and alarm time in the case of high liquid level are recorded while monitoring the tapping condition, and the current molten iron ladle receiving iron weight is predicted by the industrial computer according to the current real-time liquid level data, and the molten iron flow in the blast furnace tapping process is recorded.

[0013] As a further improvement of the present application, in step one, the liquid level monitoring and early warning system comprises a radar level meter and an audible and visual alarm, the radar level meter is installed on the ladle and faces the ladle below the tapping hole, the audible and visual alarm is arranged beside the tapping hole, and the radar level meter and the audible and visual alarm are electrically connected to the industrial computer through the controller, and the industrial computer is further connected to the display screen.

[0014] As a further improvement of the present application, in step two, the method for judging the tapping state is as follows:

[0015] Step S1, the radar level meter detects the liquid level data in the ladle in real time, transmits the collected liquid level data to the industrial computer, subjects the liquid level data to Kalman filtering processing, and records the liquid level data corresponding to the time after processing;

[0016] Step S2, according to the liquid level data corresponding to the time in the record, real-time calculating the liquid level difference of two points in interval At time;

[0017] Step S3, comparing the liquid level difference of two points in interval At time calculated in real-time with the rate mark point v start of the beginning of the tapping state and the rate mark point v end of the end of the tapping state, to determine the beginning state and the end state of the tapping, when the ladle begins to tap, the calculated liquid level rising rate reaches v start , indicating that the tapping state is entered, and the detection is continued until the liquid level rising rate is equal to v end , i.e. the tapping is ended.

[0018] As a further improvement of the present application, the method for Kalman filtering processing of the liquid level data is as follows:

[0019] Taking the liquid level data as the state quantity, the state prediction formula is as follows:

[0020]

[0021] Among them, is the state estimation value at k time, indicating the liquid level estimation value at k time, is the state estimation value at k-1 time, i.e. the liquid level value at the last time; F k is the state transition matrix; B k is the input control matrix; u k is the input at k time, representing the current tapping amount in actual production;

[0022] The updating formula of Kalman filtering is as follows:

[0023]

[0024] P k =(I-K k H k )P k|k-1

[0025] Among them, is the state estimation value at k-1 time, which is the correction of the prediction value at k time, representing the liquid level correction value at k time; y k is the liquid level measurement value at k time, H k is the measurement matrix; P k|k-1 is the covariance matrix of state estimation error, K k is the Kalman gain.

[0026] As a further improvement of the present application, after the Kalman filtering denoising, the liquid level data after the Kalman filtering is normalized by maximum and minimum value, and the formula is as follows:

[0027]

[0028] In the formula, x is the liquid level data after Kalman filtering, x' is the liquid level data after Kalman filtering and further normalization, min(x) is the minimum value in the liquid level data, and max(x) is the maximum value in the liquid level data.

[0029] As a further improvement of the present application, according to the historical liquid level data, the iron starting rate marker point v start and the iron ending rate marker point v end are determined.

[0030] (1) Select n pieces of historical liquid level data in the historical data, each piece of historical liquid level data records the liquid level data corresponding to the time of a complete iron tapping process of a certain ladle, and then calculate the corresponding rate according to the liquid level difference of two points at an interval Δt time. The rate is the liquid level rising rate of the selected ladle from an empty ladle, to the start of iron tapping, and then to the end of iron tapping.

[0031] (2) From the historical liquid level data, select the i-th piece of liquid level data, determine the region with a liquid level less than 0.2 m as a non-iron tapping state region, calculate the corresponding liquid level rising rate in the state region, find the maximum value diff Li of the liquid level rising rate; at the same time, when the liquid level continues to rise and the liquid level rising rate rises to a higher value and is in stable fluctuation, take the time corresponding to the maximum liquid level rising rate as the starting point and extend 5 minutes as the ending point, and the time period region is determined as an iron tapping state region. Find the minimum value diff Hi of the liquid level rising rate in the state region, i = 1, 2, 3, 4, … n.

[0032] (3) Repeat the operation of step (2) to find all diff L1 , diff L2 , diff L3 , … diff Ln , diff H1 , diff H2 , diff H3 , … diff Hn in the n pieces of historical liquid level data, find the maximum value diff L1 from diff L2 , diff L3 , … diff Ln , find the maximum value diff Lmax from diff H1 , diff H2 , diff H3 , … diff HnThe minimum value diff is found in the middle Hmin ,

[0033] (4) When the taping-out end rate flag point v end = 0, it is judged that the taping-out ends.

[0034] As a further improvement of the present application, the collected n pieces of historical liquid level data are respectively subjected to Kalman filtering denoising processing, and the liquid level data after denoising are subjected to normalization processing to obtain normalized liquid level data, and then the normalized liquid level data are used to determine the taping-out start rate flag point v start and the taping-out end rate flag point v end .

[0035] 3. Beneficial effects

[0036] Compared with the prior art, the beneficial effects of the present application are:

[0037] The judgment method of the taping-out state of the blast furnace ladle of the present application automatically monitors the liquid level of the ladle through a radar level meter, preliminarily judges that the liquid level starts to rise when the liquid level obviously increases, continues to detect the liquid level rising rate, judges that the current state is a taping-out state when the liquid level rising rate is detected to be stable, and determines that the time when the liquid level starts to rise is the start of taping-out, continues to monitor the liquid level rising rate, and judges that the time when the liquid level rising rate is 0 is the taping-out end time.

[0038] The method can automatically monitor and judge the taping-out state in real time and accurately, and can monitor the liquid level in the ladle in real time and accurately. Compared with the traditional manual judgment method, the method can avoid the problems of time waste and low production efficiency caused by judgment errors and human interference, improve the full ladle rate of the ladle, and thus improve the use efficiency of the ladle.

[0039] In addition, the method can also avoid the heat loss caused by taping-out under the condition of not full ladle, and solve the problems of inaccurate manual judgment of liquid level height, complex on-site production conditions, and large fluctuation of liquid level data. At the same time, the method can improve production efficiency, save time and cost, improve safety, avoid safety accidents caused by negligence of workers, improve the accuracy of measurement and judgment, improve product quality and production efficiency.

[0040] The method can also save energy and resources, reduce environmental pollution, realize data management and intelligent manufacturing, reduce labor cost and labor intensity, reduce the rate of defective products, optimize production planning, and reduce energy consumption and waste gas emission. Not only can the production efficiency and quality of the taping-out link be improved, but also the production efficiency of the subsequent electric furnace and converter steelmaking links can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 A flow chart of a judgment method of a tapping state of a blast furnace ladle according to the present application;

[0042] Figure 2 A comparison chart of original liquid level data collected during tapping and liquid level data after Kalman filtering and noise reduction processing;

[0043] Figure 3 A comparison chart of liquid level data after noise reduction before and after Kalman filtering and normalization processing;

[0044] Figure 4 A comparison chart of an original liquid level rising rate broken line and a liquid level rising rate broken line after Kalman filtering and normalization processing;

[0045] Figure 5 A liquid level and corresponding liquid level rate broken line chart of a complete tapping process after Kalman filtering processing; DETAILED DESCRIPTION

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of the drawings shown.

[0047] The present application will be further described below in combination with specific embodiments.

[0048] Embodiment 1

[0049] As shown in the figure, the judgment method of a tapping state of a blast furnace ladle according to the present embodiment comprises the following steps: Figure 1

[0050] Step one, build a liquid level monitoring and early warning system for real-time monitoring and early warning of the liquid level height of molten iron in the molten iron tank; the monitoring and early warning system comprises a radar level meter, a display screen, an audible and visual alarm and an industrial computer, the radar level meter is installed above the molten iron tank and located above the tapping hole, the molten iron ladle is directly below the tapping hole, the radar level meter measures the liquid level height directly below, the measuring range is 0-30m, the error range is ±1mm, and the audible and visual alarm and the display screen are arranged beside the tapping hole.

[0051] ​The sound and light alarm and the display screen are connected with the radar level meter and the industrial computer respectively, the radar level meter starts to collect liquid level data continuously after being connected with power supply and transmits the data to the sound and light alarm, the display screen and the industrial computer through data line, the sound and light alarm, the display screen and the industrial computer are independent, independently receive the liquid level signal transmitted by the radar level meter, the sound and light alarm alarms automatically after receiving the liquid level data of the dangerous height of the molten iron ladle, the dangerous height of the molten iron ladle is determined according to different specifications of the molten iron ladle, and the display screen only displays the current liquid level height.

[0052] The radar level meter has the highest medium temperature resistance of 1450 DEG C and the environmental temperature resistance of 400 DEG C, the working power supply is AC220V, the display screen is an LED dot matrix display screen, the radar level meter of the taphole transmits the real-time collected liquid level signal to the field display screen and the sound and light alarm, and another signal is converted into a digital signal under the modbus bus protocol through an analog-digital converter, the digital signal is transmitted to the industrial computer located in the control room of the iron mill through optical fibers, network lines and the like, the java program in the industrial computer receives the signal through serial communication and carries out subsequent data processing and taphole state judgment.

[0053] Step two, the liquid level height of the molten iron in the molten iron tank is monitored in real time through the radar level meter, and the liquid level data collected by the radar level meter is read by the industrial computer.

[0054] It should be noted that due to irregular taphole time, much dust on site and molten iron splashing and other factors, the liquid level data monitored by the radar level meter has problems such as large fluctuation, much interference, large noise and many abnormal values, which causes great difficulty for data analysis, therefore, the liquid level data is taken as a state quantity, the real-time liquid level signal of time series is estimated and predicted by using Kalman filtering, and the estimation is updated according to the latest measurement result, so that the purpose of denoising and prediction is achieved.

[0055] Specifically, the state prediction formula of Kalman filtering is as follows:

[0056]

[0057] Among them, is the state estimation value at k time, that is, the current liquid level value, for the liquid level data, it represents the liquid level estimation value at k time. is the state estimation value at k-1 time, that is, the liquid level value at the last time. k is a state transition matrix, which describes the change rule of the liquid level data in time. Since the liquid level data is one-dimensional, the state transition matrix here is a scalar 1; k is an input control matrix, which describes the influence degree of external control on the liquid level. kis the input at time k, representing the amount of external control on the liquid level, in the present application, u k is the current tapping amount in actual production, because the pouring of molten iron into the ladle will cause the liquid level to rise. This formula represents the prediction of the liquid level state at the next time according to the liquid level state at the last time and the current tapping amount.

[0058] The update formula of Kalman filtering is as follows:

[0059]

[0060] P k =(I-K k H k )P k|k-1

[0061] wherein, is the state estimation value at time k-1, that is, the liquid level value predicted according to the state estimation value at the last time and the current tapping amount, is the correction of the prediction value at time k, representing the correction value of the liquid level at time k. yk is the measurement value at time k, that is, the liquid level value obtained by the radar level meter. H k is the measurement matrix, describing the mapping relationship between the state vector and the measurement value. For liquid level data, the state vector and the measurement value are the same variable, so it is a scalar 1; P k|k-1 is the covariance matrix of state estimation error, describing the uncertainty degree of the prediction of liquid level data. K k is the Kalman gain, describing the weight distribution between the state prediction value and the state measurement value; the update formula represents the correction of the liquid level prediction value according to the current liquid level measurement value, and the calculation of the corrected state estimation error.

[0062] More preferably, in order to eliminate the dimension or scale influence of the liquid level data and improve the precision and comparability of the liquid level data, the present application further carries out maximum and minimum value normalization processing on the liquid level data after Kalman filtering, and the maximum and minimum value normalization formula is as follows:

[0063]

[0064] In the formula, x is the liquid level data after Kalman filtering processing, x' is the liquid level data after further normalization after Kalman filtering processing, min(x) is the minimum value of the liquid level data, and max(x) is the maximum value of the liquid level data. Through this formula, we can scale the value range of the liquid level data to [0, 1] in order to carry out subsequent analysis and processing.

[0065] In order to further better explain the embodiments of the present application, the comparison between the original liquid level data and the liquid level data after Kalman filtering noise reduction is provided, as shown in Figure 2 .

[0066] Image 1 shows the original liquid level-time curve monitored by the radar level gauge, and image 2 shows the liquid level-time curve after Kalman filtering for noise reduction. As can be seen from the figure, after noise reduction, the abnormal and fluctuating data of the original liquid level data are weakened, and the overall liquid level growth pattern is more stable and clear, which is conducive to subsequent judgment and processing.

[0067] In addition, a comparison is provided between the liquid level-time curve after Kalman filtering and the liquid level-time curve after normalizing the filtered and denoised liquid level data, such as... Figure 3 As shown:

[0068] Label 1 shows the liquid level-time curve after Kalman filtering noise reduction, and label 2 shows the liquid level-time curve after further normalization of the Kalman filtering noise-reduced liquid level data. It can be seen that the data after normalization is easier to analyze and process.

[0069] Additionally, a comparison is provided between the original liquid level rise rate line graph and the liquid level rise rate line graph after Kalman filtering and normalization, as shown in the appendix. Figure 4 As shown:

[0070] The image labeled 1 is a line graph of the liquid level rate after Kalman filtering and normalization, while the image labeled 2 is the original line graph of the liquid level rise rate. It can be seen that the image labeled 1 is more regular and easier to judge.

[0071] The processed liquid level data is analyzed and judged. When a significant increase in liquid level is detected, it is initially judged that the liquid level has begun to rise. Next, the rate of liquid level rise is monitored. If the rate of liquid level rise is detected to be stable, then the current state can be judged as the iron tapping state, and the moment when iron tapping begins is the moment when the liquid level begins to rise.

[0072] In this invention, the original liquid level data is denoised using Kalman filtering to obtain denoised liquid level data. (See attached diagram) Figure 4 As shown, the denoised liquid level data was compared with the original liquid level data. Next, the denoised liquid level data was normalized to obtain the normalized liquid level data. (See figure) Figure 5 As shown, the normalized liquid level data was compared with the normalized Kalman filter-denoised liquid level data.

[0073] The processed liquid level data can be represented as a time series H = [h1, h2, ..., h n ], where n represents the number of sampling points. This invention uses the liquid level difference between every two sampling points with a time interval of Δt in the time series as the rate, as shown in the following formula:

[0074]

[0075] Where v is the rate of liquid level rise, Δh is the liquid level difference, and Δt is the time interval. The time interval Δt is generally taken as tens of seconds, which is sufficient to meet the factory's requirements for the delay in iron tapping time. The larger Δt is, the more obvious the change in the liquid level rate v is, and the better the judgment effect. Generally, 30 seconds is appropriate. The rate under the time series is expressed as V=[v1,v2,…,v n-Δt ],in

[0076] Further observation of the liquid level change curve over time revealed that the rate of increase in liquid level rise significantly when the liquid level begins to rise, and the rate of increase is stable before the end of iron tapping.

[0077] In this embodiment, the iron tapping state is determined by liquid level data. When the liquid level is very low, such as when the monitored liquid level value is between 0 and 0.2m, it is determined to be an empty ladle state, which is a non-iron tapping state. When the liquid level continues to rise and the rate of rise of the liquid level remains relatively stable, it is determined to be an iron tapping state. When the liquid level is high enough, such as above 2.5m, it is determined to be a full ladle state. The transition from the empty ladle state to the iron tapping state is determined to be the start of iron tapping. When the liquid level no longer rises during the iron tapping state and the rate of rise of the liquid level is 0, it is determined to be the end of iron tapping.

[0078] When the molten iron level is too high and reaches the warning point for the set loading amount, such as when the level is greater than 2.6m, the industrial control computer transmits an alarm signal to the on-site audible and visual alarm for iron tapping safety warning. While monitoring the iron tapping situation, it records the alarm time at the start and end of iron tapping and when the level is too high. The industrial control computer predicts the weight of iron to be received in the ladle from the current real-time level using a formula. At the same time, it records and displays the molten iron flow rate during the blast furnace tapping process. The formula algorithm is as follows: the industrial control computer calculates the amount of molten iron (m) in the ladle by using the real-time level data read by the radar level gauge and the size of the ladle itself. Then, it calculates the iron flow rate, i.e., the tapping speed (Q), by using the current amount of molten iron (m) and the current time (t).

[0079] Specifically, the volume V of the molten iron ladle is related to the volume v of the molten iron per unit liquid level. i The relationship can be represented as:

[0080] V = H·v i

[0081] Where H represents the total height of the molten iron ladle, usually in meters (m).

[0082] The physical meaning of this relationship is: molten iron stacks along the height direction in the ladle, forming the entire volume of the ladle. The volume v of the molten iron per unit liquid level is... iThe volume of molten iron increases per unit of liquid level height.

[0083] Therefore, the current mass m of molten iron can be calculated by the height h of the liquid level in the molten iron tank read by the liquid level gauge and the volume v of the molten iron tank i The calculation is as follows:

[0084] m = v i h

[0085] wherein v i represents the volume of molten iron per unit of liquid level height, and the unit is usually liters per meter (L / m) or cubic meters per meter (m 3 / m).

[0086] The tapping speed Q can be calculated by the mass m of molten iron and the time t taken:

[0087]

[0088] wherein t is usually in seconds (s). The unit of the tapping speed Q is usually liters per second (L / s) or cubic meters per second (m 3 / s). The tapping speed Q here is different from the rate v used for judging the tapping state in the method, and the rate v is a parameter for judging the state calculated by the processed liquid level data and Δt, i.e., the rate of the liquid level rising.

[0089] Further observation of the curve of the liquid level changing with time shows that when the liquid level starts to rise, the rate of the liquid level rising will significantly increase, and before the tapping ends, the rate of the liquid level rising is stable. At the beginning of tapping, the rate of the liquid level rising in the molten iron ladle will suddenly increase from 0 to a certain relative constant value, and will slightly fluctuate around the value. The tapping speed of the steelmaking plant is almost constant for each tapping, but the tapping speed of different batches may have slight changes. In order to identify the actual tapping state of the molten iron ladle, two rate markers need to be selected, one is the tapping start rate marker v start , and the other is the tapping end rate marker v end .

[0090] Due to the poor site environment of the ironmaking plant and the factors such as dust diffusion and molten iron splashing during tapping, even after Kalman filtering and normalization, the liquid level will still have certain errors and fluctuations, which will further cause errors and fluctuations in the real-time calculated rate. The selection of the tapping start rate marker v start needs to distinguish the highest rate in the marker 1 and marker 5 part regions in the non-tapping state, i.e. Figure 5 , from the lowest rate in the marker 6 part region in the tapping state, i.e. Figure 5 .

[0091] The attached Figure 4and attached Figure 5 It can be seen that the minimum rate during tapping is much larger than the false maximum rate during non-tapping state, as shown by No. 6 in attached Figure 5 , which has been subjected to Kalman filtering and normalization processing, or the extreme fluctuation or noise, as shown by No. 1 in attached Figure 5 , so the selected tapping start rate marker v start is slightly larger than the minimum tapping rate during slow tapping, i.e. the historical minimum tapping rate v min , as shown by No. 7 in attached Figure 5 , so as to ensure that the tapping start state can be accurately determined regardless of the tapping speed, and there is almost no possibility of misjudgment by the present determination method. The specific method is to compare the rate calculated at the current liquid level h with v start , and as long as the rate is greater than v start , it is determined that the tapping starts.

[0092] The selection of the tapping end marker v end needs to accurately identify the zero moment of the rate during tapping end, i.e. v end ≈0. The specific method is to start real-time monitoring of the tapping end state of the ladle as soon as the ladle enters the tapping state and the liquid level is at a medium-high position, i.e. the liquid level height is higher than h middle . Compare the current liquid level rate with v end , and as long as the rate is less than v end , it is determined that the tapping ends.

[0093] As shown by No. 8 in attached Figure 5 , because during the tapping stage, the rate calculated from the liquid level data after Kalman filtering and normalization is much larger than 0, even considering the liquid level fluctuation, the rate will fluctuate around 0 during the non-tapping stage due to fluctuation and error value, and even the false maximum rate under extreme fluctuation or noise is still much smaller than the rate during tapping. Therefore, if the liquid level is at a medium-high position while the tapping state is monitored, and the rate is less than v end , it must be the tapping end state.

[0094] The selection and determination of the above two rate markers can ensure that each tapping can be accurately determined regardless of the tapping speed.

[0095] Determine the tapping start time: when the liquid level starts to rise, determine whether the current liquid level rate exceeds the rate marker v start . If so, it means that the current enters the tapping state, and the current time can be recorded as the tapping start time t start .

[0096] Determining the end time of tapping: In the tapping state, i.e., when a record of the current tapping start time already exists, and in the non-empty ladle state, i.e., the liquid level is at a medium-high level and the liquid level height is higher than h. middle Continue monitoring the liquid level rate, h middle Select the empty package liquid level h empty With the lowest iron tapping level h endmin The median value is used to accurately determine whether the package is in a high-level state or an empty state. When a stable liquid level rate is detected and below v... end When the current time is recorded, it indicates that the tapping process has ended, and this time can be recorded as the tapping end time t. end .

[0097] The formula for determining the start and end times of iron production is as follows:

[0098]

[0099] As attached Figure 5 As shown: Based on historical liquid level curves and verified by actual production data, this method can obtain the most characteristic liquid level rise rate curve when used in the blast furnace molten iron ladle tapping monitoring system to monitor the liquid level difference between two points at a time interval Δt.

[0100] As attached Figure 5 The short dotted line labeled "Kalman level" illustrates the liquid level change over time during the complete process of a molten iron ladle, from empty to full, and finally, its removal. As the liquid level begins to rise, the rate of increase significantly increases, and this rise remains stable until the end of the tapping process, without suddenly dropping below a certain threshold. This method analyzes historical rates calculated from historical liquid level data to identify the starting and ending tapping rate markers, thus accurately determining the start time of each tapping operation. The selected starting and ending tapping rate markers are universally applicable under various complex field conditions.

[0101] Specifically, in this invention, the starting rate marker v of iron tapping is determined based on historical liquid level data. start and the tapping end rate marker v end The method is as follows:

[0102] (1) Select n historical liquid level data in the historical data. Each historical liquid level data records the liquid level data at the corresponding time of a complete iron tapping process of a certain iron ladle. Then calculate the corresponding rate based on the liquid level difference between two points with an interval of Δt. This rate is the liquid level rise rate of the selected iron ladle from empty ladle to the start of iron tapping and then to the end of iron tapping.

[0103] (2) From the historical liquid level data, select the i-th liquid level data, determine the area with liquid level less than 0.2m as the non-tapping state area, calculate the corresponding liquid level rising rate in the state area, find the maximum value diff Li of the liquid level rising rate Hi ; at the same time, when the liquid level continues to rise and the liquid level rising rate rises to a higher value and is in stable fluctuation, take the time corresponding to the maximum liquid level rising rate as the starting point, and the time after 5 minutes as the ending point, the time period area is determined as the tapping state area, find the minimum value diff L1 of the liquid level rising rate in the state area L2 , i = 1, 2, 3, 4…n

[0104] (3) Repeat the operation of step (2) to find all diff L3 , diff Ln , diff H1 , ……diff H2 in n historical liquid level data H3 , diff Hn , diff L1 , ……diff L2 , diff L3 , diff Ln , ……diff Lmax , diff H1 , diff H2 , diff H3 , ……diff Hn , diff Hmin ,

[0105] (4) When the tapping end rate flag point v end = 0, it is judged that the tapping is ended.

[0106] In addition, the tapping speed of the iron mill is not constant, sometimes faster and sometimes slower, since the selection of the tapping start rate mark point in the present application considers the slowest tapping speed, and the average value is calculated, so that the start tapping at the slowest speed can be accurately captured, and the start tapping is not successfully judged; and the maximum rate in the rate fluctuation caused by the liquid level fluctuation when not in the tapping state is also not higher than the tapping start rate mark point, and the non-tapping state is not misjudged as the tapping start. The monitoring of the tapping end is the same as the monitoring of the tapping start, through the analysis of a large amount of liquid level and liquid level difference data at the corresponding Δt time interval, it can be found that under various complex tapping conditions and irregular liquid level fluctuation conditions, the liquid level difference at the lowest Δt time interval when the tapping ends is less than or equal to 0 (m / Δt), which is set as the tapping end rate mark point, so that the moment of tapping end can be captured in time, and there is almost no possibility of missing monitoring; at the same time, subsequent fluctuations will not have other effects on the judgment, avoiding misjudgment.

[0107] At the same time, the determination of the tapping start and the tapping end is not only the liquid level difference at the Δt time interval, but also the double determination according to the current liquid level value, when the liquid level is low, the liquid level difference at the Δt time interval can be determined as the tapping start. Only after successfully determining the tapping start, the continuous monitoring and determination of the tapping end will be carried out.

[0108] In combination Figure 5 The description will be described as follows: Figure 5 The mark 1 in the figure is the liquid level rate fluctuation before the tapping start, the mark 2 is the tapping start time, the mark 3 is the liquid level rate stable rising section, the time section is used to determine the tapping state, the mark 4 is the time section in which the liquid level rate is lower than the threshold value, the corresponding rate curve is the mark 5 in the figure. ​ The mark 6 is the tapping state, and the v start of the present application is the point of the mark 7, and the v end is the point of the mark 8.

[0109] The following will be described in combination with specific implementation cases:

[0110] A batch of tapping of the ladle, 10 historical data are collected, the 10 historical liquid level data are analyzed by the method of the present application, and the diff L of the 10 historical data in the non-tapping state is respectively 0.071, 0.069, 0.012, 0.035, 0.046, 0.070, 0.048, 0.039, 0.028, 0.055, wherein the maximum diff Lmax is 0.071, and the diff H0.278, 0.297, 0.268, 0.282, 0.253, 0.228, 0.237, 0.264, 0.244, 0.236, wherein the minimum value diff Hmin is 0.228, so that the calculated v start is

[0111] At this time, the 11th ladle begins tapping, and a radar level gauge is used for detection. When the liquid level is lower than 0.2 m, it is judged as non-tapping state, and continues to be detected. When t = 1020 s, the detected liquid level rising rate reaches 0.185, at this time, tapping begins, continues to be detected, and when t = 2550 s, the detected liquid level rising rate is 0, at this time, the tapping of the ladle is completed.

[0112] The application has been described in detail above with reference to specific exemplifying embodiments. However, it should be understood that various modifications and variations can be made to the application without departing from the scope of the application defined by the appended claims. The detailed description and the accompanying drawings are therefore to be taken as illustrative and not restrictive, and if any such modifications and variations are made to the application, they will fall within the scope of the application described herein. Furthermore, the background art is provided to illustrate the state of the art and its significance to the present technology, and is not intended to limit the application or the application field of the present application.

[0113] More specifically, although illustrative embodiments of the application have been described herein, the present application is not limited to these embodiments, but includes any and all embodiments having modifications, omissions, combinations (e.g., of aspects across various embodiments), adaptations and / or alterations as would be appreciated by those skilled in the art, in light of the foregoing description. The limitations in the claims are to be interpreted broadly based on the language employed in the claims and not limited to examples described in the foregoing description or during the prosecution of the application, which examples are to be construed as non-exhaustive. Any and all steps that are described in any of the foregoing examples can be employed in any order and / or in any combination, and are not limited to the specific order or combination described in the foregoing examples. Accordingly, the scope and spirit of the application should be judged in terms of the claims and the full extent of equivalents, instead of the foregoing description or the examples, which are to be construed in the broadest sense possible.

Claims

1. A method of judging the tapping state of a blast furnace ladle, characterized by: Specifically comprising the following steps: Step one, build a liquid level monitoring and early warning system for real-time monitoring and early warning of the liquid level of molten iron in the molten iron tank; Step two, the obtained liquid level data is processed by Kalman filtering, real-time denoising and prediction are performed, and the tapping state is judged according to the processed liquid level data, and the liquid level is lower than the low liquid level setting value is empty package state, the liquid level continues to rise is tapping state, and the liquid level reaches the high liquid level setting value is full package state; The method for judging the tapping state is as follows: Step S1, the radar level meter detects the liquid level data in the molten iron tank in real time, transmits the collected liquid level data to the industrial computer, processes the liquid level data by Kalman filtering, and records the liquid level data after processing corresponding to the time; Step S2, according to the liquid level data corresponding to the time in the record, the liquid level difference of two points in interval time is calculated in real time; Step S3, according to the comparison between the liquid level difference of two points in interval time calculated in real time and the rate mark point when entering the tapping state and the rate mark point when ending the tapping state, the starting state and the ending state of the tapping are determined; Step three, in the case that the liquid level of molten iron is too high and reaches the set filling amount warning point, early warning is performed, and the alarm time of the starting of tapping, the ending of tapping and the case of high liquid level is recorded while monitoring the tapping, the industrial computer predicts the weight of the molten iron received by the current molten iron ladle according to the current real-time liquid level data, and the flow of molten iron in the process of tapping of the blast furnace is recorded.

2. The method for determining the tapping state of molten iron from a blast furnace ladle according to claim 1, characterized in that: In step one, the liquid level monitoring and early warning system comprises a radar level meter and an audible and visual alarm, the radar level meter is installed on the molten iron tank and faces the molten iron ladle below the tapping hole, the audible and visual alarm is arranged beside the tapping hole, and the radar level meter and the audible and visual alarm are electrically connected with the industrial computer through the controller, and the industrial computer is further connected with the display screen.

3. The method of judging the tapping state of a blast furnace ladle according to any one of claims 1 to 2, characterized in that: The method for processing the liquid level data by Kalman filtering is as follows: The liquid level data is taken as a state quantity, and the state prediction formula is as follows: wherein, is k the state estimate value at the time instant t, represents k the liquid level estimate value at the time instant t, is k the state estimate value at the time instant t-1, i.e. the liquid level value at the previous time instant; is the state transition matrix; is the input control matrix; is k the input at the time instant t, characterized by the current tapping amount in the actual production; The update formula of Kalman filtering is as follows: wherein, is k the state estimation value at time -1, is a correction to the prediction value at time k characterized as a correction to the liquid level at time k ; is k the liquid level measurement value at time is the measurement matrix; is the covariance matrix of the state estimation error, is the Kalman gain.

4. The method of claim 3, wherein: the temperature of the molten iron is measured by a temperature sensor; the temperature of the molten iron is compared with the reference temperature; and the temperature of the molten iron is compared with the reference temperature. After Kalman filtering denoising, the liquid level data after Kalman filtering is normalized by maximum and minimum values, and the formula is as follows: In the formula, x is the liquid level data after Kalman filtering processing, is the liquid level data after Kalman filtering processing and further normalization, is the minimum value in the liquid level data, is the maximum value in the liquid level data.

5. The method for determining the tapping state of molten iron from a blast furnace ladle according to claim 3, characterized in that: Determine the starting point of the iron tapping rate based on historical liquid level data. and the end rate marker of tapping The method is as follows: (1) Select n pieces of historical liquid level data in historical data, each piece of historical liquid level data records the liquid level data of a certain complete tapping process of a hot metal ladle at the corresponding time, and then calculate the corresponding rate according to the liquid level difference of two points in time, which is the liquid level rising rate of the selected hot metal ladle from empty ladle to the start of tapping and then to the end of tapping. time. (2) From the historical liquid level data, select the first i liquid level data, determine the area with liquid level less than 0.2m as the non-tapping state area, calculate the corresponding liquid level rising rate in the state area, find the maximum value diff Li of the liquid level rising rate; at the same time, when the liquid level continues to rise and the liquid level rising rate rises to a higher value and is in stable fluctuation, take the time corresponding to the maximum liquid level rising rate as the starting point and extend 5 minutes as the endpoint, the time period area is determined as the tapping state area, find the minimum value diff Hi of the liquid level rising rate in the state area, i=1,2,3,4……n; (3) Repeat step (2) to find all diff values ​​in the n historical liquid level data. L1 diff L2 diff L3 ...diff Ln diff H1 diff H2 diff H3 ...diff Hn From diff L1 diff L2 diff L3 ...diff Ln Find the maximum value diff Lmax From diff H1 diff H2 diff H3 ...diff Hn Find the minimum value diff Hmin , ; (4) When the taping end rate flag point = 0, it is judged that the taping is ended.

6. The method of claim 5, wherein: The n pieces of collected historical liquid level data are respectively subjected to Kalman filtering denoising processing, and the liquid level data after denoising is subjected to normalization processing to obtain normalized liquid level data, and then the normalized liquid level data is used to determine the tapping start rate mark point and the tapping end rate mark point. .​ ​

Images