Converter gas recovery prediction method, device, equipment, storage medium and product

Abstract

The invention discloses a converter gas recovery prediction method, device and equipment, a storage medium and a product, and relates to the technical field of ferrous metallurgy, and the method comprises the steps: carrying out the feature extraction of converter gas data collected in real time, inputting the feature data in a first prediction time range into a short-term prediction model, and outputting a short-term prediction value; determining a mid-term prediction value in a second prediction time range according to the production plan and a preset curve template; and performing weighted fusion on the short-term prediction value and the medium-term prediction value according to the overlapping time region, and outputting a total gas recovery amount prediction value and a mixed heat value prediction value in future preset time. According to the method, real-time acquisition and feature extraction are performed on the converter gas data, so that the data are accurately obtained; short-term prediction is carried out depending on high-precision real-time data, medium-term prediction is carried out based on a production plan, and the prediction reliability of a full time scale is ensured through model fusion. And multi-time-scale accurate prediction of gas output is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of steel metallurgy, and particularly relates to a converter gas recovery prediction method, device, equipment, storage medium and product. BACKGROUND

[0002] Converter gas is an important secondary energy source for a steel enterprise, and efficient recovery and utilization thereof is crucial for energy saving and emission reduction. At present, a commonly used technical solution in the industry is to install a flow meter and a calorific value meter on a converter gas recovery pipeline, and to control recovery or diffusion according to carbon monoxide and oxygen content in the gas through a switching station. The existing system is a passive response, lacks foresight, and can only measure and control the generated gas, and cannot predict the trend of gas generation in the future. Moreover, energy scheduling lacks data foresight support and mainly relies on artificial experience, and the stability and economy need to be improved. This makes downstream users such as gas tank scheduling and generator unit deployment in a passive state, which is easy to lead to excess gas diffusion or shortage. SUMMARY

[0003] The main purpose of the present application is to provide a converter gas recovery prediction method, device, equipment, storage medium and product, aiming to solve the technical problem that the existing converter gas recovery mode cannot accurately predict the converter gas recovery amount and calorific value.

[0004] To achieve the above-mentioned purpose, the present application provides a converter gas recovery prediction method, which comprises the following steps: characteristic extraction is performed on real-time collected converter gas data to obtain characteristic data; the characteristic data in a first prediction time range is input into a short-term prediction model to output a short-term prediction value; a medium-term prediction value in a second prediction time is determined according to a production plan and a preset curve template, the second prediction time is greater than the first prediction time; the short-term prediction value and the medium-term prediction value are weighted and fused according to an overlapping time region, the overlapping time region is an overlapping region of the first prediction time range to the second prediction time range, to output a total converter gas recovery amount prediction value and a mixed calorific value prediction value in a future preset time.

[0005] In an embodiment, the converter gas data comprises switching station state information, oxygen content and carbon monoxide content, flow and calorific value, and the characteristic data comprises a mixed calorific value; the step of performing characteristic extraction on real-time collected converter gas data to obtain characteristic data comprises: an effective recovery state value of the converter gas data is determined according to the switching station state information, oxygen content and carbon monoxide content, flow and calorific value; determining a total instantaneous gas recovery amount according to the effective recovery state value and the instantaneous flow rate measurement of each converter; determining a mixed heat value according to the effective recovery state value, the instantaneous flow rate measurement, the total instantaneous gas recovery amount and the gas heat value measurement of each converter.

[0006] In an embodiment, the feature data includes a current smelting stage and a flow stability index; after the step of determining a mixed heat value according to the effective recovery state value, the instantaneous flow rate measurement, the total instantaneous gas recovery amount and the gas heat value measurement of each converter, the method further comprises: determining the current smelting stage according to the oxygen lance height and the oxygen blowing time, different current smelting stages correspond to different gas amounts and heat values; determining the flow stability index according to the contribution proportion of each converter and the Gini coefficient.

[0007] In an embodiment, the step of determining a medium-term prediction value in the second prediction time range according to the production plan and the preset curve template comprises: determining a target steel grade to be smelted in the second prediction time range according to the production plan; determining a target curve in the preset curve template according to the target steel grade, the preset curve template being an average gas generation curve template of historical same-type steel heats; aligning and superimposing the target curve in time to obtain the medium-term prediction value.

[0008] In an embodiment, the step of weighting and fusing the short-term prediction value and the medium-term prediction value according to the overlapping time region, and outputting a total gas recovery amount prediction value and a mixed heat value prediction value in a future preset time comprises: determining a short-term weight and a medium-term weight according to a prediction time and a transition time, the transition time being a transition time point from the first prediction time range to the second prediction time range; weighting and fusing the short-term prediction value and the medium-term prediction value according to the short-term weight and the medium-term weight, and outputting a total gas recovery amount prediction value and a mixed heat value prediction value in a future preset time.

[0009] In an embodiment, after the step of weighting and fusing the short-term prediction value and the medium-term prediction value according to the overlapping time region, and outputting a total gas recovery amount prediction value and a mixed heat value prediction value in a future preset time, the method further comprises: updating a model parameter and a coupling coefficient according to a total gas recovery amount actual value, a mixed heat value actual value, the total gas recovery amount prediction value and the mixed heat value prediction value, the coupling coefficient reflecting the influence intensity of flow rate change on heat value.

[0010] In addition, to achieve the above object, the application further provides a converter gas recovery prediction device, which comprises: a feature extraction module, configured to perform feature extraction on the converter gas data collected in real time to obtain feature data; a short-term prediction module, configured to input the feature data in a first prediction time range into a short-term prediction model to output a short-term prediction value; a medium-term prediction module, configured to determine a medium-term prediction value in a second prediction time range according to a production plan and a preset curve template, the second prediction time being greater than the first prediction time; a weighted fusion module, configured to perform weighted fusion on the short-term prediction value and the medium-term prediction value according to an overlapping time region, the overlapping time region being an overlapping region of the first prediction time range to the second prediction time range, to output a total converter gas recovery amount prediction value and a mixed heat value prediction value in a future preset time.

[0011] In addition, to achieve the above object, the application further provides a converter gas recovery prediction device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the converter gas recovery prediction method.

[0012] In addition, to achieve the above object, the application further provides a storage medium, which is a computer readable storage medium, and the storage medium stores a computer program, the computer program being executable by a processor to implement the steps of the converter gas recovery prediction method.

[0013] In addition, to achieve the above object, the application further provides a computer program product, which comprises a computer program, the computer program being executable by a processor to implement the steps of the converter gas recovery prediction method.

[0014] The application provides a converter gas recovery prediction method, which comprises the following steps: extracting features from real-time collected converter gas data to obtain feature data; inputting the feature data in a first prediction time range into a short-term prediction model to output a short-term prediction value; determining a medium-term prediction value in a second prediction time range according to a production plan and a preset curve template; and performing weighted fusion on the short-term prediction value and the medium-term prediction value according to an overlapping time region to output a total converter gas recovery amount prediction value and a mixed heat value prediction value in a future preset time. The application can accurately obtain data by real-time collection and feature extraction of converter gas data, perform short-term prediction depending on high-precision real-time data, perform medium-term prediction based on a production plan, and ensure the prediction reliability of the whole time scale through model fusion. The converter gas data and the production plan are deeply coupled for gas prediction, the multi-time scale accurate prediction of gas output is realized, and energy scheduling is changed from passive response to active intervention. A coupling prediction model of converter gas flow and heat value is proposed, the relationship between the two is quantified through a mathematical formula, and the flow and the heat value are simultaneously predicted. A hybrid architecture of real-time monitoring combined with AI prediction is adopted, the short-term prediction depends on high-precision real-time data, the medium-term prediction is based on production rules, and the prediction reliability of the whole time scale is ensured through model fusion. BRIEF DESCRIPTION OF DRAWINGS

[0015] The accompanying drawings, which are incorporated herein and form a part of the specification, illustrate embodiments consistent with the present application and, together with the description, further serve to explain the principles of the application.

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced as follows. Obviously, for those skilled in the art, other drawings can also be obtained based on these drawings without any creative work.

[0017] Figure 1 A flowchart is provided for the converter gas recovery prediction method embodiment one of the present application; Figure 2 A flowchart is provided for the converter gas recovery prediction method embodiment two of the present application; Figure 3 A flowchart is provided for the converter gas recovery prediction method embodiment three of the present application; Figure 4 A structure diagram of the converter gas recovery prediction system of the present application is provided; Figure 5 A data operation logic diagram of the converter gas recovery prediction system of the present application is provided; Figure 6 A pipeline instrument configuration diagram of the converter gas monitoring and prediction of the present application is provided; Figure 7A module structure schematic diagram of a converter gas recovery prediction device of an embodiment of the present application; Figure 8 A device structure schematic diagram of a hardware running environment involved in a converter gas recovery prediction method in an embodiment of the present application.

[0018] Explanation of reference numerals: 1, 2, 3: converter; 4, 8, 12: variable frequency fan; 5, 9, 13: gas detector; 6, 10, 14: switching station; 7, 11, 15, 17: flow meter; 16: calorimeter; 18: converter gas tank.

[0019] The purposes, functional features and advantages of the present application will be further explained in combination with embodiments and with reference to the drawings. DETAILED DESCRIPTION

[0020] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application, and are not used to limit the present application.

[0021] In order to better understand the technical solutions of the present application, the following will be described in detail in combination with the drawings of the specification and specific embodiments.

[0022] The main solution of the embodiment of the present application is: extracting features from the real-time collected converter gas data to obtain feature data; inputting the feature data in the first prediction time range into a short-term prediction model to output a short-term prediction value; determining a medium-term prediction value in a second prediction time range according to a production plan and a preset curve template, the second prediction time being greater than the first prediction time; weighting and fusing the short-term prediction value and the medium-term prediction value according to an overlapping time region, outputting a total gas recovery amount prediction value and a mixed heat value prediction value in a future preset time, the overlapping time region being an overlapping region of the first prediction time range to the second prediction time range.

[0023] Since converter gas is an important secondary energy source for steel enterprises, its efficient recovery and utilization is crucial for energy saving and emission reduction. At present, the technical solution commonly used in the industry is: installing a flow meter and a calorimeter on the converter gas recovery pipeline, and controlling recovery or diffusion according to the carbon monoxide and oxygen content in the gas through a switching station. The existing system responds passively, lacks foresight, can only measure and control the generated gas, and cannot predict the trend of gas generation in the future. And the energy scheduling lacks data foresight support, mainly relying on manual experience, and the stability and economy need to be improved. This makes the downstream users such as gas tank scheduling and generator unit deployment in a passive state, which is easy to lead to excess gas diffusion or shortage.

[0024] The application provides a solution, by extracting features from real-time collected converter gas data to obtain feature data; inputting the feature data in a first prediction time range into a short-term prediction model to output a short-term prediction value; determining a medium-term prediction value in a second prediction time range according to a production plan and a preset curve template; weighting and fusing the short-term prediction value and the medium-term prediction value according to an overlapping time region to output a total converter gas recovery amount prediction value and a mixed heat value prediction value in a future preset time. The application accurately acquires data by real-time collection and feature extraction of converter gas data; performs short-term prediction relying on high-precision real-time data, performs medium-term prediction based on a production plan, and ensures the prediction reliability of the whole time scale through model fusion. The converter gas data and the production plan are deeply coupled for the first time for gas prediction, realizing multi-time scale accurate prediction of gas output, and changing energy scheduling from passive response to active intervention.

[0025] It should be noted that the execution subject of the method of the embodiment can be a computing service device with converter gas recovery prediction, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone and the like; or a converter gas recovery prediction device with the same or similar functions. The embodiment and the following embodiments will be described taking the converter gas recovery prediction device as an example.

[0026] Based on this, the embodiment of the application provides a converter gas recovery prediction method, referring to Figure 1 , Figure 1 is a flowchart of the first embodiment of the converter gas recovery prediction method of the application.

[0027] In the embodiment, the converter gas recovery prediction method includes steps S10-S40: Step S10, extracting features from real-time collected converter gas data to obtain feature data.

[0028] It can be understood that the converter gas data generated by each converter can be collected in real time by installing flow meters, heat value meters and the like on the converter gas recovery pipeline, for example, oxygen lance features, flow, heat value and the like. Then extract time sequence features, intensity features and other feature data related to gas generation and consumption rules.

[0029] Step S20, inputting the feature data in a first prediction time range into a short-term prediction model to output a short-term prediction value.

[0030] It should be understood that the embodiment proposes multi-model collaborative prediction for different time sequence dimensions. For short-term prediction, the lance feature, flow, and heat value sequence in the first prediction time range (i.e., the current and a period of time in the past, for example, 30 minutes) is input into a pre-trained short-term prediction model, and the short-term prediction value in the future 30 minutes is output. The short-term prediction model can be an LSTM model, which is modeled as:

[0031] where X_t = [Q(t), CV(t), Phase(t), Stability(t),...] is the input feature vector, and T_s is the first prediction time range (for example, 30 minutes).

[0032] In step S30, the medium-term prediction value in the second prediction time range is determined according to the production plan and the preset curve template, and the second prediction time is greater than the first prediction time.

[0033] It can be understood that the medium-term prediction is the prediction in the second prediction time range greater than the first prediction time, for example, in the time range of 30 minutes-24 hours. In the embodiment, the production plan data is obtained from the production management system in advance, and based on the production plan, the historical typical coal gas generation curve is matched according to the steel grade and the planned output by using a template matching algorithm, the coal gas output condition in the future is predicted, and the medium-term prediction value is obtained.

[0034] In step S40, the short-term prediction value and the medium-term prediction value are weighted and fused according to the overlapping time region, and the total coal gas recovery prediction value and the mixed heat value prediction value in the future preset time are output, and the overlapping time region is the overlapping region of the first prediction time range to the second prediction time range.

[0035] It should be understood that there is an overlapping region when the first prediction time range transitions to the second prediction time range, and the importance of the short-term prediction value and the medium-term prediction value in the overlapping region is different, so the weights of the two can be set (for example, the short-term model is given a higher weight), and the weighted fusion is performed according to the respective weights, so that the prediction data of the total coal gas recovery prediction value and the mixed heat value prediction value in the future hours with the granularity of minutes / ticks is output.

[0036] In a feasible implementation, step S30 can include steps S301-S303: In step S301, the target steel grade to be smelted in the second prediction time range is determined according to the production plan.

[0037] It can be understood that, for the medium-term prediction, the production plan can be obtained in advance, and then the target steel grade planned to be smelted in the second prediction time range (e.g., several hours in the future) is found according to the production plan.

[0038] In step S302, a target curve is determined in a preset curve template according to the target steel grade, the preset curve template being an average gas generation curve template of historical same-type steel grades.

[0039] It should be understood that the target curve can be determined in the preset curve template according to the target steel grade. For a planned heat j, the predicted target curve is:

[0040] wherein the template curve Template_Matching is an average gas generation curve based on historical same-type heats, SteelGrade_j is the steel grade of the heat j, and Volume_j is the gas amount generated by the heat j.

[0041] In step S303, time alignment and superposition are performed according to the target curve to obtain a medium-term prediction value.

[0042] It can be understood that, after the target curve is determined, time alignment and superposition can be performed according to the target curve to obtain a medium-term prediction value. For example, the production plan shows that BOF2 will start the next smelting (steel grade Q235) after 40 minutes. The template matching module calls the typical curve of the Q235 steel grade, and predicts that BOF2 will start to generate gas after 60 minutes.

[0043] It should be noted that coupling correction of the flow rate and the heat value can also be performed. The corrected heat value prediction is:

[0044] wherein ΔQ(t) = [ - Q(t)] / Q(t) is the flow rate change rate, and a is a coupling coefficient learned through historical data and reflecting the influence strength of the flow rate change on the heat value.

[0045] In the embodiment, the typical gas generation curve template corresponding to the steel grade planned to be smelted in the next several hours is found according to the production plan, time alignment and superposition are performed. Thus, the oxygen lance signal and the production plan can be deeply coupled for gas prediction, and multi-time scale accurate prediction of the gas output is realized.

[0046] The embodiment provides a converter gas recovery prediction method, real-time collected converter gas data is subjected to feature extraction, and feature data is obtained; the feature data in a first prediction time range is input into a short-term prediction model, and a short-term prediction value is output; a medium-term prediction value in a second prediction time range is determined according to a production plan and a preset curve template; the short-term prediction value and the medium-term prediction value are weighted and fused according to an overlapping time region, and a total converter gas recovery amount prediction value and a mixed heat value prediction value in a future preset time are output. The embodiment realizes real-time collection and feature extraction of the converter gas data, and accurately obtains data; short-term prediction is performed by relying on high-precision real-time data, medium-term prediction is performed based on the production plan, and the prediction reliability of the full time scale is ensured through model fusion. The converter gas data and the production plan are deeply coupled for the first time for gas prediction, multi-time scale accurate prediction of gas output is realized, and energy scheduling is changed from passive response to active intervention.

[0047] Based on the first embodiment of the application, in the second embodiment of the application, the same or similar contents as the above-mentioned embodiment one can refer to the above introduction, and the subsequent will not be described in detail. On this basis, please refer to Figure 2 , the converter gas data includes switching station state information, oxygen content and carbon monoxide content, flow and heat value, and the feature data includes mixed heat value; step S10, the converter gas recovery prediction method further includes steps S101-S103: Step S101, determining the effective recovery state value of the converter gas data according to the switching station state information, the oxygen content and the carbon monoxide content, the flow and the heat value.

[0048] It should be noted that the collected converter gas data can be determined for effectiveness. The information required by the switching station refers to the state of the main valve of the switching station, for example, the recovery side sealing butterfly valve is opened, the burner or ignition diffusion valve at the top of the diffusion chimney is closed, and the three-way switching valve points to the recovery side channel, which indicates that the switching station is in the recovery state; otherwise, it is in the diffusion state. For each converter i, the effective recovery state S_i is determined by the following Boolean function: S_i = F_switch ∧ F_safety ∧ F_quality ∧ F_flow ∧ F_cv Wherein: F_switch = (switching station state == "recovery"), F_safety = (O2 content < threshold O2max), F_quality = (CO content > threshold CO min), F_flow = (flow ∈ [threshold flow_min, threshold flow_max]), F_cv = (heat value ∈ [threshold_cv_min, threshold_cv_max]) Step S102, determine the total instantaneous gas recovery amount according to the effective recovery state value and the instantaneous flow measurement value of each converter.

[0049] It should be understood that the total instantaneous gas recovery amount can be determined according to the effective recovery state value and the instantaneous flow measurement value of each converter, and the formula is as follows: Q_total = Σ(S_i×F_i)(i=1,2,3,……,N) Wherein, F_i is the instantaneous flow measurement value of the i-th converter.

[0050] Step S103, determine the mixed heat value according to the effective recovery state value, the instantaneous flow measurement value, the total instantaneous gas recovery amount and the gas heat value measurement value of each converter.

[0051] It can be understood that the mixed heat value is calculated according to the effective recovery state value, the instantaneous flow measurement value, the total instantaneous gas recovery amount and the gas heat value measurement value of each converter, and the formula is as follows: CV_mixed = [Σ(S_i × F_i × CV_i)] / Q_total(when Q_total>0) Wherein, CV_i is the gas heat value measurement value of the i-th converter.

[0052] Finally, the cumulative recovery amount integral calculation can also be performed, and the cumulative volume calculation formula is as follows: V_cumulative(t) = ∫Q_total(τ)dτ ≈ Σ[Q_total(t_k) × Δt_k] Wherein, Δt_k is the sampling time interval.

[0053] In a feasible implementation, the feature data includes the current smelting stage and the flow stability index; after step S103, steps S104-S105 can also be included: Step S104, determine the current smelting stage according to the oxygen lance height and the oxygen blowing time, and different current smelting stages correspond to different gas amounts and heat values.

[0054] It can be understood that the current smelting stage Phase can be determined based on the oxygen lance height H(t) and the oxygen blowing time T_elapsed: Phase = f(H(t), dH / dt, T_elapsed, T_remaining) Where: T_remaining = T_planned - T_elapsed, f is a classification function trained based on historical data, outputting {charging period, pre-conversion period, peak conversion period, post-conversion period, tapping period}. The charging period (0-2 minutes): gas generation amount ≈ 0; the pre-conversion period (2-8 minutes): gas generation amount gradually increases, with low calorific value; the peak conversion period (8-15 minutes): gas generation peak, with the highest calorific value; the post-conversion period (15-25 minutes): gas amount decreases, with fluctuating calorific value; the tapping period (25-30 minutes): gas generation amount ≈ 0.

[0055] Step S105, determining the flow stability index according to the contribution proportion of each converter and the Gini coefficient.

[0056] It should be understood that the flow stability index can also be determined according to the contribution proportion R_i of each converter and the Gini coefficient Gini: Stability = 1 - Gini(R_1, R_2, R_3) Where: R_i = F_i / Q_total, Gini is the Gini coefficient, used to measure the degree of imbalance of contribution distribution.

[0057] In this embodiment, by first judging the effective recovery of data, the total gas instantaneous recovery amount and the mixed calorific value are calculated after judging the effectiveness. Thus, real-time monitoring and metering of data are realized. Then, the oxygen lance signal is analyzed in real time, the oxygen blowing time and oxygen supply intensity are calculated, and the current smelting stage and the flow stability index of the converter are identified. Thus, it can be determined that different current smelting stages correspond to different gas amounts and calorific values, and the degree of imbalance of contribution distribution of each converter is measured.

[0058] Based on the first embodiment of the present application, in the third embodiment of the present application, the same or similar contents as the above-mentioned first embodiment can be referred to the above introduction, and will not be described in detail. On this basis, please refer to Figure 3 , step S40, the converter gas recovery prediction method further comprises steps S401-S402: Step S401, determining the short-term weight and the medium-term weight according to the prediction time and the transition time, and the transition time is the transition time point from the first prediction time range to the second prediction time range.

[0059] It can be understood that the short-term weight and the medium-term weight can be determined according to the prediction time and the transition time. Specifically, for the prediction time t_pred, the short-term weight and the medium-term weight are calculated as follows: W_short(t_pred) = max(0, 1 - t_pred / T_transition), W mid(t pred) = 1 - W short(t pred) (when t pred≤ T transition), 0≤ W short(t pred)≤ 1 wherein, T transition is a transition time point from the first prediction time range to the second prediction time range.

[0060] Step S402, weighting and fusing the short-term prediction value and the mid-term prediction value according to the short-term weight and the mid-term weight, outputting a total coal gas recovery amount prediction value and a mixed heat value prediction value in a future preset time.

[0061] It should be understood that the final weighting and fusing result is:

[0062] wherein, represents a [flow prediction, heat value prediction] vector.

[0063] In a feasible implementation, after step S40, step S50 can also be included: Step S50, updating the model parameters and the coupling coefficient according to the total coal gas recovery amount actual value, the mixed heat value actual value, the total coal gas recovery amount prediction value and the mixed heat value prediction value, the coupling coefficient reflecting the influence strength of flow change on heat value.

[0064] It is worth noting that an adaptive learning mechanism is also proposed in the embodiment to perform prediction error feedback and coupling coefficient optimization. First, the model parameter update is:

[0065] wherein η is a learning rate, θ is a model parameter, and the prediction error in the sliding window is used for adjustment.

[0066] Secondly, the coupling coefficient optimization is: α_optimal = argmin Σ[CV_actual(t_k) - CV^_corrected(t_k)]² The coupling coefficient α is adjusted online by the gradient descent method.

[0067] In the embodiment, the short-term weight and the medium-term weight are determined according to the predicted time and the transition time; the short-term predicted value and the medium-term predicted value are weighted and fused according to the short-term weight and the medium-term weight, and a total coal gas recovery amount predicted value and a mixed heat value predicted value in a future preset time are output. The model parameters and the coupling coefficient are updated in feedback according to an actual value of the total coal gas recovery amount, an actual value of the mixed heat value, the total coal gas recovery amount predicted value and the mixed heat value predicted value, and the coupling coefficient reflects the influence strength of the flow change on the heat value. In the embodiment, the short-term weight and the medium-term weight are determined according to the transition time from the short term to the medium term, so that the weighted fusion can be more accurate. The prediction result is used for coal gas tank site prediction, generator set load pre-adjustment and the like, and the actual recovery data is compared with the predicted value to continuously optimize the prediction model.

[0068] The converter coal gas recovery prediction method is described as a whole.

[0069] Firstly, the system structure of the present application is described. The system structure can refer to FIG. 1, the system data operation logic can refer to FIG. 2, and the system architecture can refer to FIG. 3. Figure 4 Figure 5

[0070] (1) Data acquisition layer: used for real-time acquisition of the flow, heat value, CO content, O2 content, switching station state, oxygen lance height, oxygen blowing state signal of each converter coal gas, and acquisition of production plan data from a production management system; (2) Real-time calculation engine: deployed in a PLC / DCS, used for real-time judgment of the "effective recovery state" of a single converter and calculation of the total coal gas recovery amount and the mixed heat value of three converters; (3) Prediction engine: deployed in a server, is the core of the present application, including: (A) feature extraction module: based on the oxygen lance signal, the converter smelting stage (such as charging period, pre-smelting period, peak period, post-smelting period, tapping period) is identified, and time sequence features and intensity features related to the coal gas generation and consumption law are extracted.

[0071] (B) Multi-time scale prediction model: a. Short-term prediction model (0-30 minutes): based on the current oxygen lance state and real-time data, a time sequence model such as LSTM (Long Short Term Memory Network) is used to predict the change trend of the converter coal gas production and heat value; b. Medium-term prediction model (30 minutes-24 hours): based on the production plan, a historical typical coal gas generation curve is matched according to the steel grade and planned output through a template matching algorithm, and the coal gas output condition of future multiple heats is predicted; c. Model fusion module: the prediction results of different time scales are weighted and fused to generate a final comprehensive prediction curve; (4) Business application layer: provides a visual monitoring interface, displays real-time data and prediction results, and supports decision-making operations of dispatchers; ​​(5) Communication interface layer: realize data exchange with underlying PLC / DCS control system, production execution system (MES), energy management system (EMS), and ensure the collaborative work between systems.

[0072] Take a steel plant equipped with three converters (BOF1, BOF2, BOF3) as an example to illustrate the flow of the method, as shown in Figure 6

[0073] Data acquisition: the system acquires the data of BOF1 through 7-flow meter F1 = 25000 Nm³ / h, analyzes the results through 5-gas detector and calculates the heat value CV_1 = 1800 kcal / Nm³, oxygen content = 0.8%, carbon monoxide content = 45%, reads the state of 6-switching station as "recovery", oxygen lance height = 2.5m, oxygen blowing state = on; similarly, acquire the data of BOF2 and BOF3; Real-time calculation: the real-time calculation engine judges that BOF1 meets all the valid recovery conditions, and its data participates in the calculation; assuming that BOF2 is in the charging period (flow is 0) and BOF3 oxygen content exceeds the standard (switches to dispersion), then the instantaneous total recovery amount Q_total = F1 = 25000 Nm³ / h, and the mixed heat value CV_mixed = CV_1 = 1800 kcal / Nm³; Feature extraction: the prediction engine analyzes the oxygen lance signal of BOF1 and identifies that it has been blown for 10 minutes, is in the blowing peak period, and the oxygen supply intensity is high; Short-term prediction: the LSTM model predicts that the gas flow of BOF1 will reach a peak value of 28000 Nm³ / h in the next 15 minutes according to the current peak period state of BOF1; Medium-term prediction: the production plan shows that BOF2 will start the next smelting (steel grade Q235) in 40 minutes; the template matching module retrieves the typical curve of Q235 steel grade and predicts that BOF2 will start to produce gas in 60 minutes.

[0074] Fusion output: the system fuses the above predictions and outputs the result: the total flow will reach a peak value in 15 minutes; the total flow will appear a second peak value in 60 minutes due to the addition of BOF2. The system can send an early warning to the energy center and suggest to prepare for gas consumption before the 15-minute peak.

[0075] It should be noted that the above example is only for understanding the present application and does not constitute a limitation on the converter gas recovery prediction method of the present application. More forms of simple transformation based on this technical concept are within the protection scope of the present application.

[0076] The present application also provides a converter gas recovery prediction device, please refer to Figure 7 , the converter gas recovery prediction device comprises: ​The feature extraction module 10 is configured to extract features from the real-time collected converter gas data to obtain feature data. The short-term prediction module 20 is configured to input the feature data in the first prediction time range into a short-term prediction model to output a short-term prediction value. The medium-term prediction module 30 is configured to determine a medium-term prediction value in a second prediction time range according to a production plan and a preset curve template, the second prediction time being greater than the first prediction time. The weighted fusion module 40 is configured to perform weighted fusion on the short-term prediction value and the medium-term prediction value according to an overlapping time region, the overlapping time region being an overlapping region of the first prediction time range to the second prediction time range, to output a total converter gas recovery amount prediction value and a mixed heat value prediction value in a future preset time.

[0077] The converter gas recovery prediction device provided by the application can solve the technical problem by using the converter gas recovery prediction method in the above embodiment. Compared with the prior art, the converter gas recovery prediction device provided by the application has the same beneficial effects as the converter gas recovery prediction method provided by the above embodiment, and other technical features in the converter gas recovery prediction device are the same as the features disclosed in the above embodiment method, which will not be repeated here.

[0078] The application provides a converter gas recovery prediction device, which comprises at least one processor and a memory in communication connection with the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the converter gas recovery prediction method in the above embodiment one.

[0079] Reference will now be made to the following description Figure 8 which shows a structural schematic diagram of a converter gas recovery prediction device suitable for implementing the embodiments of the application. The converter gas recovery prediction device in the embodiments of the application can include, but is not limited to, mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (Personal Digital Assistant), PADs (Portable Application Description), PMPs (Portable Media Player), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), and the like, and fixed terminals such as digital TVs, desktop computers, and the like. Figure 8 The converter gas recovery prediction device shown is only an example and should not impose any limitation on the functions and use range of the embodiments of the application.

[0080] As Figure 8As shown, the converter gas recovery prediction device can include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to programs stored in a read-only memory 1002 or programs loaded from a storage device 1003 into a random access memory 1004. In the random access memory 1004, various programs and data required for the operation of the converter gas recovery prediction device are also stored. The processing device 1001, the read-only memory 1002, and the random access memory 1004 are connected to each other through a bus 1005. An input / output interface 1006 is also connected to the bus. Generally, the following systems can be connected to the input / output interface 1006: input devices 1007 including, for example, a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; the storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the converter gas recovery prediction device to communicate wirelessly or wired with other devices to exchange data. Although the converter gas recovery prediction device with various systems is shown in the figure, it should be understood that all the shown systems are not required to be implemented or possessed. More or less systems can be alternatively implemented or possessed.

[0081] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing program codes for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network through the communication device, or installed from the storage device 1003, or installed from the read-only memory 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the methods of the embodiments disclosed in the present application are performed.

[0082] The converter gas recovery prediction device provided by the present application adopts the converter gas recovery prediction method in the above-mentioned embodiments, and can solve the technical problem of converter gas recovery prediction. Compared with the prior art, the converter gas recovery prediction device provided by the present application has the same beneficial effects as the converter gas recovery prediction method provided by the above-mentioned embodiments, and other technical features in the converter gas recovery prediction device are the same as the features disclosed in the previous embodiment method, which will not be repeated here.

[0083] It should be understood that various aspects of the disclosure can be implemented in hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any appropriate manner in any one or more embodiments or examples.

[0084] The above description is merely illustrative of the application and is not intended to limit the scope of the application. Any variations and modifications that can be made by any person skilled in the art within the spirit and scope of the application are intended to be encompassed by the application. The scope of the application is defined by the appended claims.

[0085] The application provides a computer readable storage medium having stored thereon computer readable program instructions (i.e., a computer program) for performing the converter gas recovery prediction method in the above-described embodiments.

[0086] The computer readable storage medium provided by the application may, for example, be a U disk, but is not limited to an electric, magnetic, optical, electromagnetic, infrared, or semiconductor system, system, or device, or any combination thereof. More specific examples of the computer readable storage medium can include, but are not limited to, an electric connection having one or more conductive wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any appropriate combination thereof. In the present embodiment, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer readable storage medium can be transmitted by any appropriate medium, including but not limited to an electric wire, an optical cable, an RF (Radio Frequency), etc., or any appropriate combination thereof.

[0087] The above-described computer readable storage medium can be included in the converter gas recovery prediction device; or can exist separately without being assembled into the converter gas recovery prediction device.

[0088] The computer readable storage medium described above carries one or more programs, when the one or more programs are executed by the converter gas recovery prediction device, the converter gas recovery prediction device is caused to: perform feature extraction on real-time collected converter gas data to obtain feature data; input the feature data in a first prediction time range into a short-term prediction model to output a short-term prediction value; determine a medium-term prediction value in a second prediction time range according to a production plan and a preset curve template; and perform weighted fusion on the short-term prediction value and the medium-term prediction value according to an overlapping time region to output a total converter gas recovery amount prediction value and a mixed heat value prediction value in a future preset time.

[0089] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++, Python, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0090] The flow diagrams and the block diagrams in the drawings are illustrations of architectures, functionalities, and operations of possible implementations of systems, methods, and computer program products according to various embodiments of present application. In this regard, each block in the flow diagrams or block diagrams can represent a module, a procedure, or a part of code, which comprises one or more executable instructions for implementing the specified functions. It should also be noted that, in some alternative implementations, the functions noted in the blocks can occur in a different order than that noted in the figures. For example, two blocks noted in succession can in fact be executed substantially concurrently or in the opposite order, depending on the functionality involved. It should also be noted that each block in the block diagrams and / or flow diagrams, and combinations of blocks in the block diagrams and / or flow diagrams, can be implemented by dedicated hardware-based systems that perform the specified functions or operations, or can be implemented by a combination of dedicated hardware-based systems and computer instructions.

[0091] The modules described in the embodiments of the present application can be implemented in the form of software or in the form of hardware. In some cases, the names of the modules do not constitute a limitation on the modules themselves.

[0092] The readable storage medium provided by the present application is a computer readable storage medium, which stores computer readable program instructions (i.e., a computer program) for executing the converter gas recovery prediction method described above, and can solve the technical problems. Compared with the prior art, the computer readable storage medium provided by the present application has the same beneficial effects as the converter gas recovery prediction method provided by the above embodiments, and will not be repeated here.

[0093] The present application also provides a computer program product, comprising a computer program, which, when executed by a processor, implements the steps of the converter gas recovery prediction method as described above.

[0094] The computer program product provided by the present application can solve the technical problems. Compared with the prior art, the computer program product provided by the present application has the same beneficial effects as the converter gas recovery prediction method provided by the above embodiments, and will not be repeated here.

[0095] The above only describes some embodiments of the present application, and does not limit the protection scope of the present application. Any equivalent structural transformation, direct / indirect application in other related technical fields based on the technical concept of the present application, and the contents of the present application and the accompanying drawings are included in the protection scope of the present application.

Claims

1. A converter gas recovery prediction method characterized by, The method comprises: characteristic extraction is performed on the real-time collected converter gas data to obtain characteristic data; the characteristic data in the first prediction time range is input into a short-term prediction model to output a short-term prediction value; a medium-term prediction value in a second prediction time range is determined according to a production plan and a preset curve template, the second prediction time being greater than the first prediction time; the short-term prediction value and the medium-term prediction value are weighted and fused according to an overlapping time region, which is an overlapping region of the first prediction time range to the second prediction time range, to output a total converter gas recovery amount prediction value and a mixed heat value prediction value in a future preset time.

2. The method of claim 1, wherein, The converter gas data comprises switching station state information, oxygen content and carbon monoxide content, flow and heat value, and the characteristic data comprises a mixed heat value; The step of performing characteristic extraction on the real-time collected converter gas data to obtain characteristic data comprises: an effective recovery state value of the converter gas data is determined according to the switching station state information, oxygen content and carbon monoxide content, flow and heat value; a total gas instantaneous recovery amount is determined according to the effective recovery state value and an instantaneous flow measurement value of each converter; a mixed heat value is determined according to the effective recovery state value, the instantaneous flow measurement value, the total gas instantaneous recovery amount and a gas heat value measurement value of each converter.

3. The method of claim 2, wherein, The characteristic data comprises a current smelting stage and a flow stability index; After the step of determining the mixed heat value according to the effective recovery state value, the instantaneous flow measurement value, the total gas instantaneous recovery amount and the gas heat value measurement value of each converter, the method further comprises: a current smelting stage is determined according to an oxygen lance height and an oxygen blowing time, different current smelting stages corresponding to different gas amounts and heat values; a flow stability index is determined according to a contribution proportion of each converter and a Gini coefficient.

4. The method of claim 1, wherein, The step of determining a medium-term prediction value in a second prediction time range according to a production plan and a preset curve template comprises: a target steel grade to be smelted in the second prediction time range is determined according to the production plan; a target curve is determined in a preset curve template according to the target steel grade, the preset curve template being an average gas generation curve template of historical same steel grade heats; a time alignment and superposition are performed according to the target curve to obtain a medium-term prediction value.

5. The method of claim 1, wherein, The step of weighting and fusing the short-term prediction value and the medium-term prediction value according to an overlapping time region to output a total converter gas recovery amount prediction value and a mixed heat value prediction value in a future preset time comprises: a short-term weight and a medium-term weight are determined according to a prediction time and a transition time, the transition time being a transition time point from the first prediction time range to the second prediction time range; the short-term prediction value and the medium-term prediction value are weighted and fused according to the short-term weight and the medium-term weight to output a total converter gas recovery amount prediction value and a mixed heat value prediction value in a future preset time.

6. The method of claim 5, wherein, After the step of weighting and fusing the short-term prediction value and the medium-term prediction value according to an overlapping time region to output a total converter gas recovery amount prediction value and a mixed heat value prediction value in a future preset time, the method further comprises: The model parameters and coupling coefficients are updated according to the actual value of the total coal gas recovery amount, the actual value of the mixed heat value, the predicted value of the total coal gas recovery amount and the predicted value of the mixed heat value, and the coupling coefficients reflect the influence strength of the flow change on the heat value.

7. A converter gas recovery prediction device, characterized in that, The converter coal gas recovery prediction device comprises: The feature extraction module is configured to extract features from the real-time collected converter coal gas data to obtain feature data. The short-term prediction module is configured to input the feature data in the first prediction time range into a short-term prediction model to output a short-term predicted value. The medium-term prediction module is configured to determine a medium-term predicted value in a second prediction time range according to a production plan and a preset curve template, the second prediction time being greater than the first prediction time. The weighted fusion module is configured to weight and fuse the short-term predicted value and the medium-term predicted value according to an overlapping time region, output a total coal gas recovery amount predicted value and a mixed heat value predicted value in a future preset time, and the overlapping time region is an overlapping region of the first prediction time range to the second prediction time range.

8. A converter gas recovery prediction device characterized by comprising: The device comprises a memory, a processor and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the converter coal gas recovery prediction method according to any one of claims 1 to 6.

9. A storage medium, characterized by The storage medium is a computer readable storage medium, and the storage medium stores a computer program, and the computer program is executed by the processor to implement the steps of the converter coal gas recovery prediction method according to any one of claims 1 to 6.

10. A computer program product, characterised in that, The computer program product comprises a computer program, and the computer program is executed by the processor to implement the steps of the converter coal gas recovery prediction method according to any one of claims 1 to 6.

Images