A furnace condition analysis method, terminal, medium and product

Abstract

This invention discloses a furnace condition analysis method, terminal, medium, and product, relating to the field of blast furnace production technology. By collecting process parameters from the blower side, identifying and extracting hot blast stove changeover disturbance data, constructing a steady-state operation benchmark database for the blower, and comprehensively diagnosing real-time parameters based on multiple decision rules of an expert system, the blast furnace operation status can be inverted from the air supply side, providing real-time and accurate decision-making basis for blast operation, effectively improving the response speed and operational accuracy of abnormal furnace conditions, and reducing surge risk and the probability of unplanned shutdowns.

Description

Technical Field

[0001] This invention relates to the field of blast furnace production technology, and in particular to a furnace condition analysis method, terminal, medium, and product. Background Technology

[0002] In actual blast furnace production, when abnormal furnace conditions occur (such as suspended charge, pipe travel, hearth buildup, etc.), blower operators often find it difficult to respond promptly to rapid changes in process parameters. Due to the lack of effective decision support, operational adjustments often lag and deviate, leading not only to decreased energy efficiency but also potentially causing the unit to deviate from safe operating thresholds, inducing surge protection or even unplanned shutdowns.

[0003] Current blast furnace condition analysis technologies primarily focus on blast furnace body parameters, such as gas flow distribution, burden surface temperature field, and burden descent rate, to guide blast furnace operation process optimization. Related research largely revolves around blast furnace condition analysis, establishing furnace condition assessment models and developing targeted optimization schemes to improve furnace conditions. However, existing technologies lack a systematic understanding of the dynamic coupling characteristics between the blower and the blast furnace system. The blower side can only passively respond to changes in furnace conditions, making it difficult to provide effective guidance for the real-time operation of the blast furnace blower. Summary of the Invention

[0004] To address the above problems, this invention provides a furnace condition analysis method, which aims to invert the blast furnace operating status through the process parameters on the blower side, and provide real-time and accurate decision-making basis for blast furnace operation.

[0005] In a first aspect, the present invention provides a furnace condition analysis method, comprising: S1 collects process parameters from the blower side, including blower outlet pressure, throat differential pressure, stator blade angle, ambient temperature, and safety margin. S2, based on the hot blast stove changeover start signal and changeover duration, identify and remove disturbance data during the changeover phase; S3, Under the condition of smooth furnace operation, based on the process parameters and the data after the furnace replacement condition, construct a benchmark database for steady-state operation of the blower; S4. Based on the expert system, perform process rationality diagnosis and analysis. The expert system contains multiple decision rules. The decision rules use the benchmark database to judge the real-time collected process parameters on the blower side in order to evaluate the current furnace condition.

[0006] Furthermore, S2 specifically includes: Collect the hot blast stove changeover start signal, calculate the changeover duration based on historical changeover data, and calculate the confidence interval [Ta, Tb] for the changeover duration; If the current furnace replacement duration exceeds (Ta+Tb) / 2+T0, where T0 is a preset constant, the furnace replacement operation is considered complete, and the data of the furnace replacement stage is stripped.

[0007] By introducing the hot blast stove replacement start signal and combining it with the statistical confidence interval of the historical replacement duration, the accurate identification and data stripping of the replacement stage are achieved. This effectively eliminates non-furnace condition disturbance data caused by the replacement (such as drastic fluctuations in air pressure), avoiding its interference with the subsequent construction of the benchmark database and diagnostic model, thereby significantly improving the accuracy and reliability of furnace condition analysis.

[0008] Furthermore, S3 specifically includes: The process parameters under the furnace operation conditions are sampled by a sliding window of preset duration. The average value of the throat differential pressure, the average value and standard deviation of the exhaust pressure, and the average value and standard deviation of the safety margin within the window are calculated to form a sample dataset. A dataset matrix was constructed based on the sample dataset, taking into account different ambient temperature ranges and still leaf angle ranges. Using the centroid method, the range of variation of exhaust pressure and safety margin is obtained based on the position of the real-time throat differential pressure in the dataset matrix.

[0009] By employing a sliding window sampling and dataset matrix construction method, steady-state process parameters can be dynamically extracted under continuous furnace conditions to form a benchmark database. Combined with the center-of-gravity method for estimation, the reasonable range of variation of exhaust pressure and safety margin can be accurately calculated under different ambient temperatures and stationary blade angles. This provides a scientific and adaptive judgment benchmark for subsequent expert systems, overcoming the poor adaptability of traditional fixed threshold judgment methods.

[0010] Furthermore, the specific calculation process for the range of change is as follows: At ambient temperature T k Below, the real-time differential pressure at the throat is ΔP k Determine ΔP in the corresponding dataset matrix. k The interval [μ] k (ΔP x ), μ k (ΔP x+1 )]; The average value of the real-time exhaust pressure μ is calculated using the following formula. k (P k ): ; Where, μ k (ΔP k () represents the average real-time differential pressure in the throat; μ k (ΔP x ), μk (ΔP x+1 The value represents the average laryngeal differential pressure collected by the dataset matrix. Standard deviation σ of real-time exhaust pressure k (P k Take σ k (P x ) and σ k (P x+1 The maximum value in ); Based on the confidence interval of the exhaust pressure, obtain the range of real-time exhaust pressure variation. .

[0011] By clearly defining the location and interpolation calculation method of the real-time throat differential pressure in the dataset matrix, the continuity and accuracy of the exhaust pressure estimation values ​​are ensured. Simultaneously, methods such as taking the maximum standard deviation and calculating confidence intervals ensure the conservatism and safety of the estimation range.

[0012] Furthermore, S3 also includes: The surge points under different stator blade angles were obtained through unit surge experiments, and the unit surge boundary line function was obtained by fitting. Under normal furnace conditions, the coordinate changes of the operating point before and after the stator blade adjustment are calculated. Combined with the stator blade adjustment angle deviation, the range of exhaust pressure change after the stator blade action is estimated, and an exhaust pressure deviation dataset under different stator blade deviations is established.

[0013] By fitting the surge boundary function through surge experiments and combining it with the changes in the coordinates of the operating points before and after stator blade adjustment, dynamic prediction of the exhaust pressure change after stator blade operation was achieved. This provides operators with a quantitative basis for stator blade adjustment, avoids blind operation that could cause the operating point to approach the surge zone, and improves operational safety and equipment protection capabilities.

[0014] Furthermore, S4 also includes: Assign weights to each alarm level of the multiple decision rules; Calculate the comprehensive evaluation weight based on the judgment results of multiple decision rules and the alarm level weight; Based on the value range of the comprehensive evaluation weights, the overall furnace condition is determined and the corresponding operation strategy is generated.

[0015] By assigning weights to multiple decision-making rules and calculating a comprehensive evaluation weight based on the judgment results of each rule, a multi-dimensional and quantitative evaluation of the furnace condition is achieved. This can comprehensively reflect the complex changing trends of the furnace condition, avoid operational misguidance caused by misjudgment of a single rule, and automatically match response strategies based on the comprehensive weight, thereby improving the intelligence level of the decision-making system.

[0016] Secondly, the present invention also provides a computer terminal, comprising: Memory, which stores executable programs; A processor for running the program, wherein the program executes the furnace condition analysis method during runtime.

[0017] Thirdly, the present invention also provides a computer-readable storage medium comprising a stored executable program, wherein the executable program, when running, controls the device where the computer-readable storage medium is located to execute the furnace condition analysis method.

[0018] Fourthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the furnace condition analysis method.

[0019] Compared with the prior art, the beneficial effects of the present invention are: This invention breaks through the limitations of traditional blast furnace condition analysis, which focuses solely on blast furnace body parameters. Instead, it takes the process parameters on the blower side as the starting point. By collecting key data such as blower outlet pressure, throat differential pressure, stator blade angle, ambient temperature, and safety margin, it constructs an analytical framework for retrieving the blast furnace operating status from the air supply side. Based on this, the invention ensures the purity of the analysis sample by identifying and removing disturbance data during the hot blast stove replacement phase. It also provides a scientific and adaptive quantitative benchmark for subsequent diagnosis by constructing a steady-state operating benchmark database for the blower under normal furnace conditions. Furthermore, based on multiple decision rules included in the expert system, the benchmark database is used to comprehensively judge real-time process parameters, achieving an accurate assessment of the current furnace condition. This invention incorporates the dynamic coupling characteristics of the blower and blast furnace system into a unified analysis model, providing blower operators with real-time and accurate decision-making basis, significantly improving response speed and operational accuracy under abnormal furnace conditions, and effectively reducing surge risk and the probability of unplanned shutdowns. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this drawing or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this drawing. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0021] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the embodiments provided by this invention without inventive effort are within the scope of protection of this invention.

[0023] This invention provides a furnace condition analysis method, such as... Figure 1 As shown, it includes: S1 collects process parameters from the blower side, including blower outlet pressure, throat differential pressure, stator blade angle, ambient temperature, and safety margin.

[0024] S2, based on the hot blast stove changeover start signal and changeover duration, identifies and removes disturbance data during the changeover phase.

[0025] Specifically, the hot blast stove replacement start signal is collected, the replacement duration is calculated based on historical replacement data, and the confidence interval [Ta, Tb] is obtained with a 95% confidence level. If the current replacement duration exceeds (Ta+Tb) / 2+T0, where T0 is a preset constant, the replacement operation is determined to be completed, and the data of the replacement stage is stripped.

[0026] S3. Under the condition of smooth furnace operation, based on the process parameters and the data after the furnace replacement operation, construct a benchmark database for steady-state operation of the blower.

[0027] Under normal furnace operating conditions—that is, when the furnace throat temperature gradient is less than 3℃ / min, the permeability index fluctuation does not exceed 5%, and the daily output reaches more than 90% of the highest output record—the blower is in normal operating condition (anti-surge valve fully closed, instrument detection normal, and air supply pipeline unobstructed). Under these conditions, by monitoring and collecting data on key process parameters such as blower-side air pressure, stator blade angle, and throat differential pressure, a basic information database is established using big data statistical analysis methods.

[0028] A. Under normal furnace operation, the rational estimation process for exhaust pressure is as follows: Statistical analysis was performed on historical data under healthy furnace conditions. At a certain ambient temperature T and a certain stationary blade angle α (±0.25°), the correlation data between throat differential pressure ΔP and exhaust pressure P were calculated.

[0029] Specifically, a sliding window W1 with a duration of 30 seconds is set up, containing historical data such as ambient temperature, throat differential pressure, exhaust pressure, and stator angle. A time period with a constant stator angle and relatively stable ambient temperature (±0.5℃) is selected, and obviously abnormal data points are removed. Statistical analysis is then performed on the data within this window. The average throat differential pressure μ(ΔP), the average exhaust pressure μ(P), and the standard deviation σ(P) within the window are calculated, and these data are divided into a sample dataset. Multiple sample datasets are obtained using the same method.

[0030] At an ambient temperature of T m The angle of the stationary blade is α m The acquired data can be used to construct the following exhaust pressure dataset matrix. : ; Where n is the total number of samples, μ(ΔP1) < μ(ΔP2) < ... < μ(ΔP n Furthermore, the signal detection circuit showed no abnormalities and the detection data was intact.

[0031] By employing the centroid method, the reasonable range of exhaust pressure variation is obtained based on the position of the real-time throat differential pressure in the exhaust pressure dataset matrix.

[0032] Specifically, at temperature T k Below, the real-time differential pressure at the throat is ΔP k Determine the throat differential pressure ΔP at this time in the exhaust pressure dataset matrix. k Falling in the interval [μ k (ΔP x ), μ k (ΔP x+1 )], x∈[1, n-1].

[0033] The average value of the real-time exhaust pressure μ is calculated using the following formula. k (P k ): ; Where, μ k (ΔP k () represents the average real-time differential pressure in the throat; μ k (ΔP x ), μ k (ΔP x+1 The value represents the average laryngeal differential pressure collected by the dataset matrix. Standard deviation σ of real-time exhaust pressure k (P k Take σ k (P x ) and σ k (Px+1 The maximum value in ).

[0034] The formula for calculating the confidence interval CI of the exhaust pressure is: Taking a confidence level of 95%, then z 0.05 / 2 =1.96. Taking the historical trend calculation of Siemens WinCC host computer as an example, with an archiving period of 500ms and a sample size of 60, the confidence interval is 1.96. .

[0035] Therefore, the reasonable range of variation for the exhaust pressure is estimated to be: This range is abbreviated as .

[0036] B. Under normal furnace operation, the rational estimation process for the safety margin is as follows: The safety margin is defined as the difference between the blower's operating point and the surge adjustment line, expressed in kPa. Using the same method as for exhaust pressure estimation, the corresponding safety margin value can be obtained at a given ambient temperature T and a given throat differential pressure ΔP.

[0037] At an ambient temperature of T v At that time, the following safety margin dataset matrix is ​​established. : ; Where, μ v (ΔP n ), μ v (SM n ), σ v (SM n The values ​​are the average value of the differential pressure in the larynx, the average value of the safety margin, and the standard deviation of the safety margin, respectively.

[0038] Then at ambient temperature T v Below, when the real-time throat differential pressure ΔP v Falling in the interval [μ v (ΔP t ), μ v (ΔP t+1 Similarly, within the range of )], the reasonable range of variation for the safety margin can be estimated as follows: .

[0039] C. Under normal furnace operation, the rational estimation process for adjusting the wind pressure corresponding to the stationary blades is as follows: Through unit surge tests, the surge point coordinates (F) under different stator blade angles (i.e., different operating conditions) were obtained. i Ep iThe unit's surge boundary function equation was obtained by curve fitting using Matlab. Combined with the addition and subtraction of the stator blades, the reasonable range of exhaust pressure change after the stator blades operate can be estimated under the premise that other conditions on the air supply side remain unchanged.

[0040] Let the fitting function for the surge boundary line be: Y=Ax 3 +Bx 2 +Cx+D; Where A, B, and C are constants, and A ≠ 0; With the vent valve fully closed, the coordinates of the operating points before and after reducing the stationary vane angle are H1(x1, y1) and H2(x0, y1), respectively. ), from y1-y =(x1-x0)×tanθ can be used to obtain y =y1-(x1-x0)×tanθ, where θ is the angle between the line formed by points H1 and H2 and the x-axis, y The exhaust pressure value after reducing the stationary vane angle is y1, the exhaust pressure value before reducing the stationary vane angle is x0, the throat differential pressure after reducing the stationary vane angle is x1, and the throat differential pressure before reducing the stationary vane angle is x1.

[0041] Taking the derivative of the fitted function Y, we obtain its tangent equation: Y'=3AX 2 +2BX+C; From this, the following inequality can be derived: y1>y >y1 - Y H2 '(x1-x0); Among them, Y H2 ' is the derivative of the Y function at point H2.

[0042] Using the same method as described above, within a certain temperature range (±1℃) and a stationary blade angle range (e.g., 40°~45°), the exhaust pressure drop values ​​corresponding to 1°, 1.5°, 2°, 2.5°, and 3° of the reduced stationary blade were collected, as shown in Table 1, and a deviation dataset matrix was established.

[0043] Table 1. Correspondence between collected data on the depressurization vane and the decrease in exhaust pressure.

[0044] Wherein, β is the difference between the average exhaust pressure before and after the depressurization vane within the same data window. Under the same conditions (same temperature, same depressurization vane range, same depressurization vane adjustment amount), multiple sets of differences are obtained and the average value is taken. Theoretically, β5 > β4 > β3 > β2 > β1.

[0045] S4. Based on the expert system, perform process rationality diagnosis and analysis. The expert system contains multiple decision rules. The decision rules use the benchmark database to judge the real-time collected process parameters on the blower side in order to evaluate the current furnace condition.

[0046] Based on the aforementioned benchmark database, an expert system is established to diagnose the rationality of the blower process parameters through multiple decision rules, thereby achieving intelligent assessment of the furnace condition.

[0047] A. Compliance diagnosis of stator vane control based on exhaust pressure Based on the furnace operating conditions, the estimated range of exhaust pressure corresponding to a certain ambient temperature and throat differential pressure. The estimated upper limit of exhaust pressure is compared with the current real-time exhaust pressure P. 排气 By comparing the results, the furnace condition judgment rules are established as shown in Table 2: Table 2. Furnace Condition Judgment Rules Based on Static Air Pressure

[0048] B. Surge risk assessment based on safety margin Based on the furnace operating conditions, the estimated range of safety margin corresponding to a certain ambient temperature and throat differential pressure. The estimated lower limit of the safety margin is compared with the current real-time safety margin SM to establish the furnace condition judgment rules as shown in Table 3: Table 3. Furnace condition judgment rules based on safety margin

[0049] C. Verification of the direction of throat differential pressure change based on Bernoulli's equation According to fluid mechanics principles, the throat differential pressure is proportional to the square of the flow rate. During airflow regulation, increasing the stator opening causes the throat differential pressure and exhaust pressure to rise synchronously, increasing the airflow and shifting the operating point to the upper right; decreasing the stator opening causes the throat differential pressure and exhaust pressure to fall synchronously, decreasing the airflow and shifting the operating point to the lower left.

[0050] Based on the above patterns, the actual exhaust pressure deviation after reducing the stationary blades at a certain ambient temperature and stationary blade angle is compared with the predicted pressure drop in the database to establish the furnace condition judgment rules as shown in Table 4: Table 4. Furnace Condition Judgment Rules Based on Dynamic Air Pressure

[0051] D. Comprehensive furnace condition assessment and operation strategy generation Based on the above three diagnostic rules, the weights assigned to each marking level are shown in Table 5 below: Table 5. Weight Allocation

[0052] Based on the judgment results of each decision rule and their corresponding weights, the comprehensive evaluation weight ξ is calculated, and the overall furnace condition and corresponding response strategies are determined accordingly, as shown in Table 6 below: Table 6 Comprehensive Analysis of Furnace Conditions and Corresponding Strategies

[0053] It should be noted that the present invention is not limited to the above-described embodiments. The above embodiments are merely examples, and any embodiments that have the same structure and perform the same effects as the technical concept within the scope of the present invention are included within the scope of the present invention. Furthermore, various modifications that can be conceived by those skilled in the art to the embodiments, and other ways of constructing by combining some of the constituent elements of the embodiments, without departing from the spirit of the present invention, are also included within the scope of the present invention.

Claims

1. A furnace condition analysis method, characterized in that, include: S1 collects process parameters from the blower side, including blower outlet pressure, throat differential pressure, stator blade angle, ambient temperature, and safety margin. S2, based on the hot blast stove changeover start signal and changeover duration, identify and remove disturbance data during the changeover phase; S3, Under the condition of smooth furnace operation, based on the process parameters and the data after the furnace replacement condition, construct a benchmark database for steady-state operation of the blower; S4. Based on the expert system, perform process rationality diagnosis and analysis. The expert system contains multiple decision rules. The decision rules use the benchmark database to judge the real-time collected process parameters on the blower side in order to evaluate the current furnace condition.

2. The furnace condition analysis method according to claim 1, characterized in that, S2 specifically includes: Collect the hot blast stove changeover start signal, calculate the changeover duration based on historical changeover data, and calculate the confidence interval [Ta, Tb] for the changeover duration; If the current furnace replacement duration exceeds (Ta+Tb) / 2+T0, where T0 is a preset constant, the furnace replacement operation is considered complete, and the data of the furnace replacement stage is stripped.

3. The furnace condition analysis method according to claim 1, characterized in that, S3 specifically includes: The process parameters under the furnace operation conditions are sampled by a sliding window of preset duration. The average value of the throat differential pressure, the average value and standard deviation of the exhaust pressure, and the average value and standard deviation of the safety margin within the window are calculated to form a sample dataset. A dataset matrix was constructed based on the sample dataset, taking into account different ambient temperature ranges and still leaf angle ranges. Using the centroid method, the range of variation of exhaust pressure and safety margin is obtained based on the position of the real-time throat differential pressure in the dataset matrix.

4. The furnace condition analysis method according to claim 3, characterized in that, The specific calculation process for the range of change is as follows: At ambient temperature T k Below, the real-time differential pressure at the throat is ΔP k Determine ΔP in the corresponding dataset matrix. k The interval [μ] k (ΔP x ), μ k (ΔP x+1 )]; The average value of the real-time exhaust pressure μ is calculated using the following formula. k (P k ): ； Where, μ k (ΔP k () represents the average real-time differential pressure in the throat; μ k (ΔP x ), μ k (ΔP x+1 The value represents the average laryngeal differential pressure collected by the dataset matrix. Standard deviation σ of real-time exhaust pressure k (P k Take σ k (P x ) and σ k (P x+1 The maximum value in ); Based on the confidence interval of the exhaust pressure, obtain the range of real-time exhaust pressure variation. 。 5. The furnace condition analysis method according to claim 1, characterized in that, S3 further includes: The surge points under different stator blade angles were obtained through unit surge experiments, and the unit surge boundary line function was obtained by fitting. Under normal furnace conditions, the coordinate changes of the operating point before and after the stator blade adjustment are calculated. Combined with the stator blade adjustment angle deviation, the range of exhaust pressure change after the stator blade action is estimated, and an exhaust pressure deviation dataset under different stator blade deviations is established.

6. The furnace condition analysis method according to claim 1, characterized in that, S4 further includes: Assign weights to each alarm level of the multiple decision rules; Calculate the comprehensive evaluation weight based on the judgment results of multiple decision rules and the alarm level weight; Based on the value range of the comprehensive evaluation weights, the overall furnace condition is determined and the corresponding operation strategy is generated.

7. A computer terminal, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the computer-readable storage medium is located to perform the method according to any one of claims 1 to 6.

9. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 6.

Images