A method, system, device and medium for dynamically predicting residual thickness of a runner

Abstract

This invention discloses a method, system, equipment, and medium for dynamic prediction of residual thickness in molten iron troughs, belonging to the field of molten iron trough technology. It includes acquiring time-series data of the trough shell temperature and iron tapping conditions; extracting temperature response features based on the time-series data of the trough shell temperature and mapping them to multiple residual thickness prediction units; calculating the thermal flux coupling erosion amount of each residual thickness prediction unit in the current prediction update cycle based on the mapped temperature response features and the iron tapping conditions; updating the residual thickness of each residual thickness prediction unit based on the residual thickness of the previous prediction update cycle and the thermal flux coupling erosion amount, forming the residual thickness state field of the current prediction update cycle; back-calculating the trough shell temperature based on the updated residual thickness state field and comparing it with the measured trough shell temperature; correcting the erosion calculation parameters based on the back-calculation error, and outputting the residual thickness distribution and risk level of the molten iron trough. This invention solves the technical problem of molten iron trough thickness prediction.

Description

Technical Field

[0001] This invention relates to the field of iron trough technology, and in particular to a method, system, equipment and medium for dynamically predicting the residual thickness of iron troughs. Background Technology

[0002] The molten iron trough is a crucial channel in the blast furnace tapping system, used to receive and transport high-temperature molten iron and slag. The lining of the trough is typically composed of refractory castables, ramming mixes, or precast refractory components. During long-term tapping, it is continuously subjected to the combined effects of thermal shock from the high-temperature molten iron, mechanical erosion from the flowing molten iron, and chemical corrosion from the slag. With increasing tapping frequency and flow rate, the refractory material in the working layer of the trough gradually thins. If the remaining thickness falls below a safe range, it can easily lead to safety accidents such as burn-through of the trough, molten iron leakage, and equipment damage.

[0003] Currently, the determination of the residual thickness of molten iron channels mainly relies on manual experience, periodic channel shutdown inspections, infrared thermography, embedded temperature sensing elements, or simple estimation based on channel shell temperature. While these methods can reflect the service status of the molten iron channels to some extent, they still have many shortcomings. On the one hand, manual inspections usually need to be carried out when the channels are shut down or under maintenance, making it impossible to monitor the changes in residual thickness in different areas of the molten iron channels in real time during continuous tapping. On the other hand, the erosion of molten iron channels does not occur uniformly; the degree of damage varies significantly in locations such as the inlet impact zone, the bottom of the main channel, the slag line area, and the area near the skimmer. Relying solely on single-point temperature or local experience makes it difficult to accurately reflect the overall residual thickness distribution of the molten iron channels. Furthermore, existing temperature-based judgment methods mostly rely on alarms based on channel shell temperature or estimations based on a fixed relationship between temperature and residual thickness. These methods typically do not fully consider the temporal characteristics of temperature rise lag, temperature rise rate, peak temperature, and cooling decay during a single tapping process, and also fail to incorporate the influence of operating conditions such as molten iron flow rate, tapping duration, and slag-to-iron ratio on the erosion of refractory materials.

[0004] Therefore, it is necessary to provide a dynamic prediction method for the residual thickness of iron troughs that can combine the temporal changes in trough temperature and iron tapping conditions to dynamically update the prediction units for different residual thicknesses of iron troughs, and can correct the erosion calculation parameters based on back-calculation errors, so as to improve the real-time performance, accuracy and safety early warning capability of iron trough residual thickness prediction. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method, system, equipment, and medium for dynamically predicting the residual thickness of iron troughs, aiming to solve the problems of existing technologies.

[0006] To achieve the above objectives, this application provides the following solution: In a first aspect, this application provides a method for dynamically predicting the residual thickness of a molten iron ditch, comprising: acquiring time-series data of the ditch shell temperature and tapping conditions; extracting temperature response features based on the time-series data of the ditch shell temperature and mapping them to multiple residual thickness prediction units; calculating the thermal flux coupling erosion amount of each residual thickness prediction unit in the current prediction update cycle based on the mapped temperature response features and the tapping conditions; updating the residual thickness of each residual thickness prediction unit based on the residual thickness of the previous prediction update cycle and the thermal flux coupling erosion amount, forming a residual thickness state field for the current prediction update cycle; back-calculating the ditch shell temperature based on the updated residual thickness state field and comparing it with the measured ditch shell temperature; correcting the erosion calculation parameters based on the back-calculation error, and outputting the residual thickness distribution and risk level of the molten iron ditch. Secondly, this application provides a dynamic prediction system for the residual thickness of a molten iron trough, including a data acquisition module for acquiring time-series data of the trough shell temperature and molten iron tapping conditions, with one tapping process as a prediction update cycle; a temperature response feature extraction module for extracting temperature response features based on the time-series data of the trough shell temperature; a residual thickness prediction unit mapping module for dividing the molten iron trough into multiple residual thickness prediction units and mapping the temperature response features to the corresponding residual thickness prediction units according to the correlation between each residual thickness prediction unit and the temperature measurement location and the molten iron flow path; and an erosion calculation module for calculating the erosion amount based on the mapped temperature response features and the molten iron tapping conditions. The system uses the following modules: a residual thickness prediction module, a residual thickness state update module, and a risk output module. The residual thickness prediction module calculates the thermal flux coupling erosion amount of each residual thickness prediction unit in the current prediction update cycle. The residual thickness state update module updates the residual thickness of each residual thickness prediction unit based on the residual thickness of the previous prediction update cycle and the thermal flux coupling erosion amount, forming the residual thickness state field for the current prediction update cycle. The reverse calculation correction module calculates the trench shell temperature based on the residual thickness state field and corrects the erosion calculation parameters of the corresponding residual thickness prediction unit in the next prediction update cycle based on the deviation between the reverse calculated trench shell temperature and the measured trench shell temperature. The risk output module outputs the residual thickness distribution and risk judgment results of the molten iron trench based on the corrected residual thickness state field.

[0007] Thirdly, this application provides a computer device including one or more processors; and a memory storing computer program instructions, which, when executed, cause the processor to perform the steps of the method described above.

[0008] Fourthly, this application provides a computer-readable medium having a computer program / instructions stored thereon, which, when executed by a processor, implement the steps of the method described above.

[0009] Through the above technical solutions, the beneficial effects of this invention are as follows: Using a single tapping process as the prediction update cycle, and combining time-series data of the trough shell temperature and tapping conditions, the residual thickness prediction units of each trough are dynamically updated, enabling timely reflection of residual thickness changes in different areas. By extracting temperature response characteristics and calculating the thermal flux coupling erosion, the prediction results simultaneously consider the effects of heat conduction, molten iron scouring, and slag erosion, improving the accuracy of residual thickness prediction. Furthermore, the trough shell temperature is calculated back from the residual thickness state field, and the erosion calculation parameters are corrected using the back-calculation error, giving the model self-correcting capabilities. This allows for the output of residual thickness distribution, risk level, and remaining tapping times, providing a basis for trough maintenance, repair, and safety early warning. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention, and the embodiments in the accompanying drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 A flowchart of a method for dynamically predicting the residual thickness of a molten iron ditch according to an embodiment of this application; Figure 2 This is a schematic diagram of a prediction unit and temperature measurement mapping provided in an embodiment of this application; Figure 3 This is a schematic diagram of temperature response feature extraction provided in an embodiment of this application; Figure 4 This is a schematic diagram illustrating the erosion calculation and residual thickness update provided in an embodiment of this application; Figure 5 This is a schematic diagram of the back-calculation correction and risk output closed loop provided in an embodiment of this application; Figure 6 This is a structural block diagram of a dynamic prediction system for residual thickness of molten iron ditch provided in an embodiment of this application; Figure 7 This is a schematic diagram of the structure of a computer device provided in one embodiment of this application; The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0012] It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of the application. Rather, these embodiments are provided to make the disclosure more thorough and complete, and to fully convey the scope of the disclosure to those skilled in the art.

[0013] The foregoing and other technical contents, features and effects of the present invention are described in conjunction with the appendix below. Figure 1-7 The detailed description of the embodiments will make this clear. All structural details mentioned in the following embodiments are based on the accompanying drawings.

[0014] Exemplary embodiments of the present invention will now be described with reference to the accompanying drawings.

[0015] In one exemplary embodiment, such as Figure 1 As shown, a method for dynamically predicting the residual thickness of molten iron troughs is provided, including: S101, acquire the time-series data of the trough shell temperature and the iron tapping condition data; S102, extract temperature response features based on the time series data of the trench shell temperature, and map them to multiple residual thickness prediction units; S103, Based on the mapped temperature response characteristics and the iron tapping condition data, calculate the thermal flux coupling erosion amount of each residual thickness prediction unit in the current prediction update cycle; S104. Based on the residual thickness of the previous prediction update cycle and the thermal flux coupling erosion amount, update the residual thickness of each residual thickness prediction unit to form the residual thickness state field of the current prediction update cycle. S105, calculate the trench shell temperature based on the updated residual thickness state field and compare it with the measured trench shell temperature. S106, corrects the erosion calculation parameters based on the back-calculation error, and outputs the residual thickness distribution and risk level of the iron ditch.

[0016] For step S101, taking one iron tapping process as a prediction update cycle, the system first acquires the time-series data of the trough shell temperature and the iron tapping condition data. The time-series data of the trough shell temperature can be obtained through thermocouples, infrared thermometers, or infrared thermal imaging devices placed at different locations on the trough shell. The iron tapping condition data may include molten iron temperature, tapping duration, molten iron flow rate, iron throughput, slag-to-iron ratio, etc. The system synchronizes the above data according to a unified time reference, ensuring that the trough shell temperature change process matches the corresponding iron tapping process.

[0017] For step S102, temperature response features are extracted based on the time-series data of the trough shell temperature. Specifically, during a single tapping process, the time it takes for the trough shell temperature to rise from its initial temperature is recorded, using the start of tapping as a reference, to obtain the temperature rise lag time; the temperature rise rate is obtained based on the slope of the temperature curve during the tapping stage; the peak temperature is obtained based on the highest temperature in the temperature curve; and the cooling decay time is obtained based on the time required for the temperature to drop to a preset proportion or preset temperature after tapping. These features can reflect the rate of heat transfer from the molten iron side to the trough shell side and the heat storage state of the trough lining.

[0018] like Figure 2 As shown, the molten iron channel is divided into multiple residual thickness prediction units. These units can be further divided along the molten iron flow direction into the inlet impact zone, the front section of the main channel, the middle section of the main channel, the rear section of the main channel, the area near the skimmer, and the outlet zone. Each area can be further subdivided according to the location of the channel bottom, sidewalls, and slag line. For each residual thickness prediction unit, a correlation is established between it and the temperature measurement location and the molten iron flow path. This correlation can be determined based on the distance between the temperature measurement point and the prediction unit, the direction of molten iron flow, the location of the channel structure, and historical erosion patterns. Through this correlation, the temperature response characteristics are mapped to the corresponding residual thickness prediction unit.

[0019] For step S103, based on the mapped temperature response characteristics and the tapping condition data, the thermal flux coupling erosion amount of each residual thickness prediction unit within the current prediction update cycle is calculated. Specifically, the system uses the temperature response characteristics to characterize the heating state of the refractory material, uses the tapping condition data to characterize the molten iron operating state, and combines the two types of data to determine the degree of erosion experienced by each residual thickness prediction unit within the current cycle. The resulting thermal flux coupling erosion amount is used to represent the residual thickness loss of the corresponding residual thickness prediction unit within the prediction update cycle.

[0020] For step S104, based on the residual thickness of the previous prediction update cycle and the thermal flux coupling erosion amount, the residual thickness of each residual thickness prediction unit is updated to form the residual thickness state field of the current prediction update cycle. Specifically, the system reads the residual thickness data of each residual thickness prediction unit saved in the previous cycle and updates the residual thickness of each residual thickness prediction unit in combination with the thermal flux coupling erosion amount calculated in the current cycle. After the update is completed, the residual thickness of each residual thickness prediction unit together constitutes the residual thickness state field of the current cycle, which is used to represent the residual thickness distribution in different areas of the molten iron ditch.

[0021] For step S105, the trench shell temperature is calculated back based on the updated residual thickness state field and compared with the measured trench shell temperature. Specifically, the system calculates the theoretical trench shell temperature corresponding to the temperature measurement location based on the current residual thickness state field and the heat transfer relationship of the molten iron trench, and then compares the theoretical trench shell temperature with the measured trench shell temperature actually collected in the current cycle to obtain the back-calculation error between the two.

[0022] For step S106, the erosion calculation parameters are corrected based on the back-calculation error, and the residual thickness distribution and risk level of the tundish are output. Specifically, when there is a deviation between the theoretical tundish shell temperature and the measured tundish shell temperature, the erosion calculation parameters used in subsequent prediction update cycles are adjusted according to the deviation, so that the subsequent prediction results gradually approach the actual service state of the tundish. After the parameter correction is completed, the system outputs the current residual thickness distribution of the tundish and gives the corresponding risk level based on the residual thickness distribution.

[0023] Through the above implementation method, this embodiment can dynamically update the residual thickness according to the temperature change of the trough shell and the iron tapping conditions during the continuous operation of the molten iron trough, and use the back calculation error to correct the prediction process, thereby improving the real-time performance and reliability of the prediction of the residual thickness of the molten iron trough.

[0024] In one specific embodiment, the temperature response characteristics include at least two of the following: temperature rise hysteresis time, temperature rise rate, peak temperature, and cooling decay time; Among them, the temperature rise lag time is used to characterize the degree of delay in heat conduction, and the cooling decay time is used to characterize the rate at which the temperature of the lining of the molten iron trough decays.

[0025] like Figure 3 As shown, the temperature response characteristics are obtained based on the temperature change curve of the trench shell within a prediction update cycle. Using the start time of the current prediction update cycle as the time reference, the process of the trench shell temperature changing over time is recorded, and the time-series data of the trench shell temperature is smoothed. The temperature rise lag time is the time difference between the start of the prediction update cycle and the initial sustained rise in the trench shell temperature reaching a preset temperature rise threshold. This temperature rise lag time reflects the degree of delay in heat transfer from the inside of the molten iron trench to the outside of the trench shell. When the temperature rise lag time becomes shorter, it usually indicates that the heat transfer path in the corresponding area is shortened, and the residual thickness may be reduced.

[0026] The temperature rise rate can be determined based on the temperature change per unit time during the heating phase of the trench shell. For example, a stable heating range before the trench shell temperature rises from the initial temperature to the peak temperature can be selected, and the ratio of the temperature change to the time change within this range can be calculated. The larger the temperature rise rate, the stronger the thermal response in that region.

[0027] The peak temperature can be the highest temperature value in the time series data of the trench shell temperature within the current prediction update cycle, used to characterize the maximum degree of thermal impact on the trench shell within that cycle.

[0028] The cooling decay time is the time required for the trench shell temperature to drop from the peak temperature to a preset proportional temperature or a preset temperature threshold after the predicted update cycle ends. The cooling decay time characterizes the heat storage and dissipation state of the molten iron trench lining. A prolonged cooling decay time indicates stronger heat storage or slower heat dissipation in the corresponding area, and can be used as a basis for judging residual thickness changes.

[0029] By extracting at least two of the above features, namely, temperature rise lag time, temperature rise rate, peak temperature, and cooling decay time, temperature response information that reflects the heat transfer state of the molten iron lining can be obtained from the time series data of the trench shell temperature, providing a data basis for the calculation of erosion amount in the subsequent residual thickness prediction unit.

[0030] In one specific embodiment, the thermal flow coupled erosion amount includes at least two of the following: thermal penetration factor, flow erosion factor, and slag erosion factor; Among them, the heat penetration factor is determined by the temperature rise lag time, peak temperature and cooling decay time, the flow scouring factor is determined by the molten iron flow rate and the tapping duration, and the slag erosion factor is determined by the slag-to-iron ratio and the relative position of the residual thickness prediction unit.

[0031] like Figure 4 As shown, for the i-th residual thickness prediction unit, its thermal flux coupling erosion amount can be calculated as follows: ΔH i =K i ×A i ×B i ×C i ; Where, ΔH i K represents the thermal flux coupling erosion amount of the i-th residual thickness prediction unit in the current prediction update cycle. i Let A be the basic erosion coefficient corresponding to the remaining thickness prediction unit. i B is the heat penetration factor. i C is the flow scouring factor. i It is a slag erosion factor.

[0032] The heat penetration factor A i The heat penetration factor is determined based on the temperature rise lag time, peak temperature, and cooling decay time. Specifically, when the temperature rise lag time is shortened, the peak temperature is increased, or the cooling decay time is prolonged, it indicates that heat is more easily transferred from the inside of the molten iron channel to the channel shell side. This suggests that the refractory material in the corresponding residual thickness prediction unit may be thinned, thus increasing the value of the heat penetration factor. The heat penetration factor can be obtained by weighting the normalized temperature rise lag time, peak temperature, and cooling decay time. For example: A i = a1×L i + a2×P i + a3×D i ; Among them, L i P is the normalized value corresponding to the temperature rise lag time. i D is the normalized value corresponding to the peak temperature. i L represents the normalized value corresponding to the cooling decay time, with a1, a2, and a3 being weighting coefficients. It should be noted that the shorter the temperature rise lag time, the stronger the heat penetration, therefore L... i It can be determined by using the inverse normalized value of the temperature rise lag time.

[0033] The flow scouring factor B iThe flow erosion factor is determined based on the molten iron flow rate and the tapping duration. Specifically, when the molten iron flow rate increases or the tapping duration is prolonged, the scouring effect of the molten iron on the bottom, sidewalls, and exit area intensifies, thus increasing the value of the flow erosion factor. The flow erosion factor can be obtained by weighting the normalized values ​​of the molten iron flow rate and the tapping duration. For example: B i =b1×Q i +b2×T i ; Among them, Q i T is the normalized value corresponding to the molten iron flow rate. i b1 and b2 are the normalized values ​​corresponding to the duration of iron tapping, and the weighting coefficients are b1 and b2.

[0034] The slag erosion factor C i The slag erosion factor is determined based on the slag-to-iron ratio and the relative position of the residual thickness prediction unit. Specifically, when the slag-to-iron ratio is high and the residual thickness prediction unit is located in the slag line region, near the skimmer, or in a region where the slag has a long residence time, the erosion effect of the slag on the refractory material is enhanced, thus increasing the value of the slag erosion factor. For example: C i =c1×S+c2×R i ; Where S is the normalized value corresponding to the slag-to-iron ratio, and R i Here, c1 and c2 are the position correction coefficients for the residual thickness prediction unit, and c1 and c2 are the weighting coefficients. Specifically, R is the position correction coefficient for the residual thickness prediction unit located in the slag line region. i The position correction factor is greater than that of the residual thickness prediction unit located in the ordinary sidewall region or trench bottom region.

[0035] In practical applications, K i a1, a2, a3, b1, b2, c1, c2, and R i The settings or corrections can be made based on the structure of the molten iron channel, the type of refractory material, historical iron tapping data, historical maintenance data, and measured residual thickness data. Through this method, the thermal flux coupling erosion amount can simultaneously reflect the effects of heat conduction, molten iron scouring, and slag erosion on the residual thickness of the molten iron channel.

[0036] In one specific embodiment, taking one iron tapping process as the prediction cycle, the update formula for the residual thickness includes: Current cycle residual thickness = previous cycle residual thickness - current cycle thermal flux coupling erosion amount + repair compensation thickness; The repair compensation thickness is determined by the repair area, repair time, and repair thickness in the iron trough repair history. For the i-th residual thickness prediction unit, its current cycle residual thickness can be determined according to the following formula: H i (n)=Hi (n-1)-ΔH i (n)+R i (n); Among them, H i (n) represents the current cycle residual thickness of the i-th residual thickness prediction unit after the n-th prediction cycle ends; H i (n-1) represents the residual thickness of the residual thickness prediction unit after the end of the previous prediction cycle; ΔH i (n) represents the thermal flux coupling erosion amount calculated within the current prediction period; R i (n) represents the repair compensation thickness corresponding to the current prediction period.

[0037] Specifically, after each tapping cycle, the system reads the residual thickness state field saved from the previous prediction cycle and deducts the residual thickness of each residual thickness prediction unit based on the thermal flux coupling erosion amount calculated for the current cycle. When a residual thickness prediction unit has not been repaired, the repair compensation thickness R of that unit is reduced. i (n) takes the value of zero.

[0038] When there are repair records for the molten iron ditch, the system determines the repair compensation thickness based on the repair history. The repair history includes the repair area, repair time, and repair thickness. If the repair area coincides with a certain residual thickness prediction unit, then the repair thickness is used as the repair compensation thickness for that residual thickness prediction unit; if the repair area only covers a part of the residual thickness prediction unit, then the repair compensation thickness can be determined based on the coverage ratio.

[0039] For example, if the residual thickness of a certain residual thickness prediction unit was 180 mm in the previous cycle, and the thermal flux coupling erosion in the current cycle is 4 mm, and no repair has been performed in this area, then the residual thickness in the current cycle is 176 mm. If there is repair in this area before or after the current cycle, and the repair compensation thickness is 20 mm, then the residual thickness in the current cycle is updated to 196 mm.

[0040] Through the above update method, the residual thickness state field can simultaneously reflect the refractory material loss caused by the thermal flux coupling effect during the ironing process, as well as the compensation effect of artificial repair on the residual thickness, making the predicted residual thickness of the iron trough more consistent with the actual service condition.

[0041] In one specific embodiment, the step of correcting the erosion calculation parameters based on the back-calculation error includes: When the deviation between the calculated trench shell temperature and the measured trench shell temperature exceeds the preset error threshold, the erosion coefficient of each residual thickness prediction unit is automatically adjusted according to the deviation. When the temperature back-calculation error exceeds the preset threshold, the weight of the influence coefficient corresponding to that temperature measurement point in the residual thickness update process is reduced.

[0042] like Figure 5 As shown, after forming the residual thickness state field for the current prediction update cycle, the trench shell temperature at the corresponding temperature measurement point is calculated based on this residual thickness state field. The calculated trench shell temperature is then compared with the actual measured trench shell temperature to obtain the temperature calculation error. For the j-th temperature measurement point, the temperature calculation error can be expressed as: E j =T Actual Measurement j -T reverse calculation j ; Among them, E j T represents the temperature back-calculation error at the j-th temperature measurement point, where T is the measured value. j This represents the measured temperature of the trench shell at that temperature measurement point, T is calculated in reverse. j This represents the trench shell temperature calculated from the current residual thickness state field.

[0043] When the absolute value of the temperature back-calculation error exceeds a preset error threshold, the system corrects the erosion calculation parameters of the corresponding residual thickness prediction unit based on the error direction and magnitude. Specifically, when the measured trench shell temperature is higher than the back-calculated trench shell temperature, it indicates that the actual heat transfer intensity is greater than the model calculation result, suggesting that the residual thickness of the corresponding area may be less than the predicted value, or that the erosion rate of the area is higher than the current model setting value. In this case, the erosion coefficient of the residual thickness prediction unit associated with the temperature measurement point is increased.

[0044] When the measured trench shell temperature is lower than the calculated trench shell temperature, it indicates that the model may have overestimated the heat transfer intensity or erosion degree in that region. In this case, the erosion coefficient of the corresponding residual thickness prediction unit should be reduced. The adjustment amount of the erosion coefficient can be determined based on the difference between the temperature calculation error and a preset error threshold. For example, it can be corrected as follows: K i (n+1)=K i (n)×[1+β×E j / E0] Among them, K i (n) represents the erosion coefficient of the i-th residual thickness prediction unit in the current prediction cycle, K i (n+1) is the erosion coefficient corrected for the next prediction period, β is the correction coefficient, and E j E0 represents the temperature back-calculation error, and E0 is the preset error threshold. In practical applications, upper and lower limits can also be set for the correction magnitude to avoid sudden parameter changes.

[0045] Furthermore, when the temperature back-calculation error at a certain temperature measurement point continues to exceed a preset threshold, or when the temperature time series data at that temperature measurement point shows signal loss, long-term constancy, or instantaneous abnormal changes, the system determines that the data reliability of that temperature measurement point is reduced, and reduces the weight of the corresponding influence coefficient of that temperature measurement point in the residual thickness update process.

[0046] For example, the original influence weight of a certain temperature measurement point on the adjacent residual thickness prediction unit was W. j If an anomaly occurs at this temperature measurement point, it can be corrected as follows: W j =γ×W j Among them, W j ' represents the corrected influence weight, and γ is the weight decay coefficient less than 1.

[0047] In this way, the system can use the deviation between the calculated trench shell temperature and the measured trench shell temperature to dynamically adjust the erosion coefficient and the weight of the temperature measurement influence, so that the residual thickness update results in subsequent prediction cycles are closer to the actual service state of the iron ditch, and reduce the interference of abnormal temperature measurement data on the prediction results.

[0048] In one specific embodiment, the risk level determination includes: Based on the updated residual thickness state field and the set safe residual thickness threshold, the risk level of each residual thickness prediction unit is determined; If the predicted residual thickness is less than the safe residual thickness threshold, a danger level is output; if it is greater than the safe residual thickness threshold, a normal level is output, and a prediction of the remaining iron tapping times is given.

[0049] The residual thickness state field includes the predicted residual thickness value corresponding to each residual thickness prediction unit. The system compares the predicted residual thickness value of each residual thickness prediction unit with a pre-set safe residual thickness threshold. For the i-th residual thickness prediction unit, its risk level can be determined as follows: When H i (n) <Hsafe i When this happens, the remaining thickness prediction unit is determined to be at a dangerous level; When H i (n)≥Hsafe i When the time is right, the residual thickness prediction unit is determined to be of normal level.

[0050] Among them, H i (n) represents the predicted residual thickness value of the i-th residual thickness prediction unit after the n-th prediction cycle, Hsafe i This indicates the safe residual thickness threshold corresponding to the residual thickness prediction unit. The safe residual thickness threshold can be preset according to the structural dimensions of the molten iron ditch, the type of refractory material, the usage conditions of the molten iron ditch, and safety production requirements.

[0051] In a further implementation, the remaining number of tapping operations can be predicted based on the difference between the predicted residual thickness value and the safe residual thickness threshold. Specifically, the system calculates the average single-cycle erosion amount based on the thermal flux coupling erosion amount of the residual thickness prediction unit in the most recent prediction cycles, and determines the remaining number of tapping operations before the residual thickness prediction unit reaches the safe residual thickness threshold based on the difference between the current predicted residual thickness value and the safe residual thickness threshold. For the i-th residual thickness prediction unit, its remaining number of tapping operations can be expressed as: N i =(H i (n)-Hsafe i ) / ΔHavg i Where, N i ΔHavg represents the remaining number of iron taps for the i-th residual thickness prediction unit. i This represents the average thermal flux coupling erosion amount of the remaining thickness prediction unit over the most recent prediction periods. In practical applications, the calculation result can be rounded to obtain the remaining number of tapping operations that can be used for production management.

[0052] When multiple residual thickness prediction units participate in risk assessment, the system can use the minimum remaining iron tapping count among all residual thickness prediction units as the total remaining iron tapping count of the molten iron trough. This method allows for the priority identification of hazardous areas within the molten iron trough that are closest to the safe residual thickness threshold, providing a basis for subsequent repair, trough shutdown for maintenance, or trough replacement arrangements.

[0053] In one specific embodiment, the method further includes: Based on historical iron tapping data and repair records, multi-cycle residual thickness prediction is performed to output the residual thickness of the iron trough after several future iron tapping cycles, and the iron tapping schedule or maintenance plan is optimized based on the prediction results.

[0054] When performing multi-cycle predictions, the system can determine the predicted operating conditions for future iron tapping processes based on the average iron tapping conditions of the most recent iron tapping processes or according to the planned iron tapping schedule. The predicted operating conditions may include data such as predicted molten iron flow rate, predicted iron tapping duration, predicted iron throughput, and predicted slag-to-iron ratio.

[0055] For the i-th residual thickness prediction unit, the system uses the residual thickness H at the end of the current prediction cycle. i Using (n) as the initial value, the corresponding thermal-fluid coupling erosion amount is calculated cycle by cycle according to the predicted future operating conditions, and the residual thickness after each future cycle is updated accordingly. The prediction process can be expressed as follows: H i (n+k)=H i (n+k-1)-ΔH i (n+k)+R i (n+k) Among them, Hi (n+k) represents the predicted residual thickness value of the ith residual thickness prediction unit after the (n+k)th prediction cycle; ΔH i (n+k) represents the thermal flux coupling erosion amount corresponding to the (n+k)th prediction cycle; R i (n+k) represents the repair compensation thickness corresponding to the (n+k)th prediction cycle; k is an integer greater than or equal to 1.

[0056] When there are no planned repairs in the future forecast period, R i (n+k) takes the value of zero; when planned repairs exist, the system determines the repair compensation thickness of the corresponding residual thickness prediction unit based on the planned repair area, repair time, and repair thickness. Through cycle-by-cycle iteration, the system can output the residual thickness distribution of the molten iron channel after several future tapping cycles and identify the residual thickness prediction unit that first approaches or falls below the safe residual thickness threshold. If the prediction results show that a certain area will reach a dangerous state within a few future tapping cycles, the system can generate suggestions for early repair, reducing tapping intensity, adjusting tapping rhythm, or arranging channel shutdown for maintenance; if the prediction results show that each residual thickness prediction unit is still within the safe range, the current tapping plan can continue.

[0057] By adopting the above methods, the maintenance of iron troughs can be transformed from post-event inspections to advance prediction and planned maintenance, which helps to reduce the risk of burn-through and improve the precision of the iron trough's service life management.

[0058] In an exemplary embodiment, the bottom area of ​​the middle section of the main channel of a blast furnace iron trough is taken as the i-th residual thickness prediction unit U_i. The residual thickness H_i(n-1) of the previous prediction cycle of this residual thickness prediction unit is 178 mm, the safe residual thickness threshold Hsafe_i is 150 mm, and the basic erosion coefficient K_i is 0.65 mm / cycle. This embodiment uses one iron tapping process as the current prediction update cycle.

[0059] The following data were collected during the current prediction update cycle: average molten iron flow rate was 5.0 t / min, molten iron tapping duration was 65 min, molten iron temperature was 1505℃, and slag-to-iron ratio was 0.11. The initial temperature of the trough shell at the corresponding temperature measurement point for this residual thickness prediction unit was 90℃, the peak temperature was 285℃, the temperature rise lag time was 10 min, and the cooling decay time was 38 min.

[0060] The system's preset benchmark parameters include: benchmark temperature rise lag time of 12 min, benchmark peak temperature of 260℃, benchmark cooling decay time of 35 min, benchmark molten iron flow rate of 4.8 t / min, benchmark tapping duration of 60 min, and benchmark slag-to-iron ratio of 0.10. The residual thickness prediction unit is located in the bottom region of the middle section of the main channel, and its position correction coefficient Rpos_i is set to 1.15.

[0061] First, the heat penetration factor A_i is determined based on the temperature response characteristics. Since a shorter temperature rise lag time indicates faster heat transfer, the temperature rise lag time is determined using inverse normalization: L_i = 12 / 10 = 1.20. The normalized value for the peak temperature is: P_i = 285 / 260 = 1.10. The normalized value for the cooling decay time is: Di = 38 / 35 = 1.09.

[0062] Let the weights for temperature rise lag time, peak temperature and cooling decay time be 0.4, 0.35 and 0.25 respectively, then the heat penetration factor is: A_i = 0.4×L_i + 0.35×P_i + 0.25×D_i = 1.14.

[0063] Then, the flow erosion factor B_i is determined based on the molten iron flow rate and the tapping duration. The normalized value of the molten iron flow rate is: Q_i = 5.0 / 4.8 = 1.04. The normalized value of the tapping duration is: T_i = 65 / 60 = 1.08.

[0064] Assuming the weights for molten iron flow rate and tapping duration are 0.6 and 0.4 respectively, the flow scouring factor is: B_i = 0.6 × Q_i + 0.4 × T_i = 1.06.

[0065] The slag erosion factor C_i is then determined based on the slag-to-iron ratio and the location of the residual thickness prediction unit. The normalized value of the slag-to-iron ratio is: S_i = 0.11 / 0.10 = 1.10. Assuming the weights corresponding to the slag-to-iron ratio and the location correction coefficient are 0.7 and 0.3 respectively, the slag erosion factor is: C_i = 0.7 × S_i + 0.3 × Rpos_i = 1.12.

[0066] Therefore, within the current prediction update cycle, the thermal flux coupling erosion amount of this residual thickness prediction unit is: ΔH_i(n)=K_i×A_i×B_i×C_i=0.88mm.

[0067] If the residual thickness prediction unit is not repaired within the current prediction period, then the repair compensation thickness R_i(n) = 0. Therefore, the residual thickness of the residual thickness prediction unit in the current period is: H_i(n) = H_i(n-1) - ΔH_i(n) + R_i(n) = 177.12 mm.

[0068] Therefore, the predicted residual thickness of the residual thickness prediction unit after the current prediction update cycle ends is 177.12 mm.

[0069] Furthermore, the system calculates the trench shell temperature at the measurement point based on the updated residual thickness state field. Assuming the trench shell temperature calculated from the current residual thickness state field is 278℃, while the measured trench shell temperature is 285℃, the temperature calculation error is: E_j = T_measured_j / T_calculated_j = 7℃.

[0070] Let the preset error threshold E_0 be 10℃. Since E_j is less than E_0, it means that the deviation between the current model back-calculation result and the measured result is within the allowable range. Therefore, no significant adjustment will be made to the basic erosion coefficient in this cycle, or only a small smoothing correction will be made.

[0071] For example, if the system is set not to correct the erosion coefficient when the error is less than the preset error threshold, then: Ki(n+1) = Ki(n) = 0.65 mm / cycle.

[0072] If the system is set to use a fine-tuning method for correction, and the correction coefficient β is 0.03, then the basic erosion coefficient for the next prediction cycle is: Ki(n+1) = Ki(n) × [1 + β × E_j / E_0] = 0.664 mm / cycle.

[0073] In practical applications, one can choose not to make corrections or make minor corrections based on the stability of the on-site model.

[0074] After completing the current cycle's residual thickness update, the system calculates the remaining number of iron tapping operations based on the average thermal flux coupling erosion over the most recent prediction cycles. Assuming that the average thermal flux coupling erosion ΔHavg_i of this residual thickness prediction unit is 0.82 mm / cycle over the last 10 prediction cycles, the remaining number of iron tapping operations before this residual thickness prediction unit reaches the safe residual thickness threshold is: N_i = (H_i(n) - Hsafe_i) / ΔHavg_i = 33.07.

[0075] Therefore, the system can output that the remaining iron tapping count for this residual thickness prediction unit is approximately 33. Since the current predicted residual thickness of 177.12 mm is greater than the safe residual thickness threshold of 150 mm, the system classifies this area as normal; however, since it has entered the continuous thinning stage, it can be marked as an area requiring monitoring in the risk output results.

[0076] If the remaining iron tapping counts for other residual thickness prediction units in the iron ditch are 45, 52, 39, and 58 respectively, the system uses the smallest remaining iron tapping count among all residual thickness prediction units as the overall remaining iron tapping count for the iron ditch. In this case, the overall remaining iron tapping count for the iron ditch is 33, and the system indicates that the middle section of the ditch bottom area is the current priority area.

[0077] In another scenario, if a localized repair of the main channel's bottom section is planned after the 20th tapping, with a planned repair thickness of 25mm, the system can incorporate a repair compensation thickness R_i(n+k) = 25mm into the corresponding prediction cycle during multi-cycle prediction. In this case, the predicted residual thickness state field after repair will be updated, allowing the system to further predict the available tapping times and risk changes of the repaired iron channel.

[0078] Through the above-described specific parameter calculation process, it can be seen that this embodiment can calculate the thermal flux coupling erosion amount of the current prediction cycle based on the time sequence characteristics of the trough shell temperature, the iron tapping condition data, and the location of the residual thickness prediction unit, and obtain the residual thickness of the current cycle by combining it with the residual thickness of the previous cycle; at the same time, the system can also use the deviation between the back-calculated trough shell temperature and the measured trough shell temperature to correct the erosion calculation parameters of the next cycle, thereby realizing the dynamic prediction of the residual thickness of the molten iron trough.

[0079] Based on the same inventive concept, this application also provides a dynamic prediction system for the residual thickness of iron troughs. The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations in one or more method embodiments provided below can be found in the limitations described above, and will not be repeated here.

[0080] In one exemplary embodiment, such as Figure 6 As shown, a dynamic prediction system for residual thickness of molten iron trenches is provided, comprising: The data acquisition module is used to acquire time-series data of the trough shell temperature and the tapping condition data, with one tapping process as a prediction update cycle. A temperature response feature extraction module is used to extract temperature response features based on the time-series temperature data of the trench shell. The residual thickness prediction unit mapping module is used to divide the molten iron trough into multiple residual thickness prediction units, and to map the temperature response characteristics to the corresponding residual thickness prediction units according to the correlation between each residual thickness prediction unit and the temperature measurement position and the molten iron flow path. The erosion calculation module is used to calculate the thermal flux coupling erosion of each residual thickness prediction unit in the current prediction update cycle based on the mapped temperature response characteristics and the iron tapping condition data. The residual thickness state update module is used to update the residual thickness of each residual thickness prediction unit based on the residual thickness of the previous prediction update cycle and the thermal flux coupling erosion amount, so as to form the residual thickness state field of the current prediction update cycle. The back-calculation correction module is used to back-calculate the trench shell temperature based on the residual thickness state field, and correct the erosion calculation parameters of the corresponding residual thickness prediction unit in the next prediction update cycle according to the deviation between the back-calculated trench shell temperature and the measured trench shell temperature. The risk output module is used to output the residual thickness distribution and risk assessment results of the molten iron trough based on the corrected residual thickness state field.

[0081] This invention, through joint processing of time-series data on trough temperature and iron tapping conditions, divides the prediction process of residual thickness in molten iron troughs into multiple functional modules, including data acquisition, temperature response feature extraction, residual thickness prediction unit mapping, thermal flux coupling erosion calculation, residual thickness status update, back-calculation correction, and risk output. Each module executes sequentially according to the prediction update cycle, and the back-calculation correction module feeds back the deviation between the measured and calculated trough temperature to the erosion calculation parameters, thus forming a dynamic correction closed loop for residual thickness prediction.

[0082] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 7 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and databases. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media to run. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection.

[0083] Those skilled in the art will understand that Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0084] In one exemplary embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0085] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0086] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0087] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (Read-Only Memory). Memory includes ROM, magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0088] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0089] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0090] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for dynamically predicting residual thickness of a runner, characterized by, include: Acquire time-series data of the trough shell temperature and iron tapping conditions; Temperature response features are extracted based on the time-series data of the trench shell temperature and mapped to multiple residual thickness prediction units. Based on the mapped temperature response characteristics and the iron tapping condition data, calculate the thermal flux coupling erosion amount of each residual thickness prediction unit in the current prediction update cycle. Based on the residual thickness of the previous prediction update cycle and the thermal flux coupling erosion amount, the residual thickness of each residual thickness prediction unit is updated to form the residual thickness state field of the current prediction update cycle. The trench shell temperature is calculated back based on the updated residual thickness state field and compared with the measured trench shell temperature. The erosion calculation parameters are corrected based on the back-calculation error, and the distribution of the residual thickness of the iron trough and the risk level are output.

2. The method of claim 1, wherein, The temperature response characteristics include at least two of the following: temperature rise lag time, temperature rise rate, peak temperature, and cooling decay time. Among them, the temperature rise lag time is used to characterize the degree of delay in heat conduction, and the cooling decay time is used to characterize the rate at which the temperature of the lining of the molten iron trough decays.

3. The method according to claim 2, characterized in that, The thermal flow coupled erosion amount includes at least two of the following: thermal penetration factor, flow scouring factor, and slag erosion factor; Among them, the heat penetration factor is determined by the temperature rise lag time, peak temperature and cooling decay time, the flow scouring factor is determined by the molten iron flow rate and the tapping duration, and the slag erosion factor is determined by the slag-to-iron ratio and the relative position of the residual thickness prediction unit.

4. The method according to claim 1, characterized in that, Taking one iron tapping process as the prediction cycle, the update formula for the residual thickness includes: Current cycle residual thickness = previous cycle residual thickness - current cycle thermal flux coupling erosion amount + repair compensation thickness; The repair compensation thickness is determined by the repair area, repair time, and repair thickness in the iron ditch repair history.

5. The method according to claim 1, characterized in that, The correction of erosion calculation parameters based on back-calculation error includes: When the deviation between the calculated trench shell temperature and the measured trench shell temperature exceeds the preset error threshold, the erosion coefficient of each residual thickness prediction unit is automatically adjusted according to the deviation. When the temperature back-calculation error exceeds the preset threshold, the weight of the influence coefficient corresponding to that temperature measurement point in the residual thickness update process is reduced.

6. The method according to any one of claims 1 to 5, characterized in that, The risk level assessment includes: Based on the updated residual thickness state field and the set safe residual thickness threshold, the risk level of each residual thickness prediction unit is determined; If the predicted residual thickness is less than the safe residual thickness threshold, a danger level is output; if it is greater than the safe residual thickness threshold, a normal level is output, and a prediction of the remaining iron tapping times is given.

7. The method according to claim 6, characterized in that, The method further includes: Based on historical iron tapping data and repair records, multi-cycle residual thickness prediction is performed to output the residual thickness of the iron trough after several future iron tapping cycles, and the iron tapping schedule or maintenance plan is optimized based on the prediction results.

8. A dynamic prediction system for residual thickness of molten iron trenches, characterized in that, The system includes: The data acquisition module is used to acquire time-series data of the trough shell temperature and the tapping condition data, with one tapping process as a prediction update cycle. A temperature response feature extraction module is used to extract temperature response features based on the time-series temperature data of the trench shell. The residual thickness prediction unit mapping module is used to divide the molten iron trough into multiple residual thickness prediction units, and to map the temperature response characteristics to the corresponding residual thickness prediction units according to the correlation between each residual thickness prediction unit and the temperature measurement position and the molten iron flow path. The erosion calculation module is used to calculate the thermal flux coupling erosion of each residual thickness prediction unit in the current prediction update cycle based on the mapped temperature response characteristics and the iron tapping condition data. The residual thickness state update module is used to update the residual thickness of each residual thickness prediction unit based on the residual thickness of the previous prediction update cycle and the thermal flux coupling erosion amount, so as to form the residual thickness state field of the current prediction update cycle. The back-calculation correction module is used to back-calculate the trench shell temperature based on the residual thickness state field, and correct the erosion calculation parameters of the corresponding residual thickness prediction unit in the next prediction update cycle according to the deviation between the back-calculated trench shell temperature and the measured trench shell temperature. The risk output module is used to output the residual thickness distribution and risk assessment results of the molten iron trough based on the corrected residual thickness state field.

9. An electronic device, characterized in that, The electronic device includes: One or more processors; and A memory storing computer program instructions, which, when executed, cause the processor to perform the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method according to any one of claims 1 to 7.

Images