Blast furnace burden distribution is one of the crucial aspects of blast furnace ironmaking. Due to the airtight nature of the top loading equipment , it is not possible to intuitiVely and accurately obserVe the actual distribution of furnace burden in the furnace throat , such as the distribution of bur- den layers and ore-to-coke ratio. During production , blast furnace operators typically rely on the distri- bution of cO2 , coal gas temperature , or gas floWrate in the coal gas at the furnace throat for upper ad- justments. With the adVancement of detection technology , operators can utilize infrared cameras , ther- mal imaging , laser technology , and radar technology to perform more precise upper adjustments. HoW- eVer , these technologies are only capable of capturing information pertaining to the surface of the fur- nace material , and they fail to proVide a detailed understanding of the distribution information regard- ing the material layer and ore-to-coke ratio. Based on the mathematical model of blast furnace burden distribution , the research comprehensiVely applies cutting-edge technologies such as numerical simula- tion , supercomputing , and artificial intelligence. Through big data cloud platforms and intelligent neu- ral netWorks , an intelligent monitoring system for the distribution of the entire blast furnace burden lay- er Was constructed. This system can achieVe online tracking and simulation of the material distribution process , assisting blast furnace operators in understanding the distribution of furnace materials. After the application of the system , Various economic and technical indicators of the blast furnace haVe been significantly improVed. compared With the past , the monthly output has increased by an aVerage of 4. 7% , the fuel ratio has decreased by 2. 37 kg/tFe , and the Vanadium recoVery rate is all aboVe 8o% .
As a key equipment in the hot-rolled Wide and thick plate production line , it is a complex system With multidimensional factors coupled With each other. Especially during operation , due to the strong coupling and nonlinear characteristics of Various influencing factors , the rolling mill operation process is in a“ black box ”state , Which makes it difficult to accurately grasp the rolling mill operation status in real time and formulate preVentiVe measures. This has become a bottleneck problem that re- stricts the high-precision and stable rolling of the hot-rolled Wide and thick plate production line. To solVe this industry problem , data-driVen monitoring and analysis technology for the status of hot-rolled Wide and thick plate rolling mills Was introduced. Based on deep perception of Various data in the roll- ing process , big data algorithms are applied for data correlation analysis , and a rolling process dynam- ic model and rolling mill status online monitoring system platform With multi-source data-driVen fusion mechanism are constructed. After being applied in releVant production lines , the accuracy of the sickle bending offset model has exceeded 95% , and the Vibration speed of the precision rolling mill has been reduced by 64 . 8% ~72 . 8% . It can effectiVely identify 90% of the rolling mill state data , effectiVely improVing the real-time accurate perception of the rolling mill state and the ability to analyze and pre- dict abnormal states.
Under the background of green and intelligent construction in the steel industry , traditional single-process quality control systems become insufficient to meet the demands of smart steel manufac - turing. They need to eVolVe into full-process integrated process control systems. These systems need to transition from mere model calculations to collaboratiVe Whole-process process model manufacturing , production data mining , and intelligent system decision-making , self-adaptation , and self-learning. with this objectiVe in mind , the system architecture , research and deVelopment content , and anticipa- ted Vision for the deVelopment and application of intelligent process control systems and process models across four dimensions : intelligent egaipment , horizontal process floW, Vertical information transfer , and time Were explored.
In the sintering process , accurate prediction of the sintering burn-through point is essential to control the operation of the sintering machine , as it determines the quality of the sintered material and the efficiency of energy use. HoWeVer , the current common sintering burn-through point prediction models mainly focus on single-step prediction , and they tend to ignore the spatio-temporal characteris- tics of the sintering process data. In VieWof the multi-step prediction task requirements of the sintering burn-through point and the complex spatio-temporal characteristics of process data , a multi-step pre- diction model of the sintering burn-through point based on encoder-decoder architecture Was proposed. The spatio-temporal encoder-task decoder architecture effectiVely extracts the spatio-temporal features of sintering process data by stacking temporal conVolutional netWorks and spatial attention mechanism , and adopts a task-specific guidance decoder to correspond to each step of prediction in the form of in- dependent units. The effectiVeness of the proposed method Was Verified by an actual sintering industrial process case. This study can assist operators to obtain information about the future state of the sintering process in adVance , so as to adjust the operating parameters more accurately , reduce the quality fluc - tuation and energy Waste caused by the lag , and improVe the quality stability of the sinter.
The end-point carbon content and temperature of the basic oxygen furnace (BOF) are the crucial factors for ensuring the seamless operation of steelmaking production. consequently , a noVel end-point prediction model based on the L6Vy flights Whale optimization algorithm ( LWOA) and ε - tWin support Vector regression ( εTSVR) has been established. Firstly , the box plot Was employed to filter the data. Subsequently , the key factors influencing the end-point carbon content and temperature Were identified through metallurgical mechanism and spearman correlation analysis. Finally , the pa- rameters in the εTSVRalgorithm Were automatically optimized by utilizing LWOA, Which possesses the characteristics of simple adjustment parameters and rapid conVergence speed. The simulation results indicate that the proposed LWOA-εTSVRprediction model haVe hit rates of 89% and 93% at the end- point carbon content and temperature satisfying the error tolerance of ±0 . 005% and ± 10 ℃ , respec- tiVely . MeanWhile , the double hit rate reaches 83% . compared With the other three prediction mod- els , the proposed LWOA-εTSVRmodel demonstrates superior adVantages. Furthermore , by setting dif- ferent error interVals , the reliability of the performance of the proposed prediction model Was also Veri- fied. MoreoVer , the proposed prediction model has higher prediction accuracy than the actual steel plant process , proViding robust technical support for steel enterprises.
The prediction of the endpoint carbon content and molten steel temperature in conVerter steelmaking is of great significance for the precise control of molten steel composition and temperature and the improVement of product quality . Due to the high dimension , high noise and strong nonlinearity of the conVerter steelmaking process data , it is difficult to obtain high-quality data. Direct modeling not only has a loWhit rate but also is prone to oVerfitting. To address this problem , a random forest prediction model based on adaptiVe SMOTEdata enhancement technology Was proposed. Firstly , the feature selection of the endpoint carbon content and molten steel temperature Was performed by recur- siVe elimination method. Secondly , the adaptiVe SMOTEalgorithm Was used to enhance the original data. Finally , random forest Was used to predict the endpoint carbon content and molten steel tempera- ture respectiVely . The actual industrial data shoW that the prediction hit rate of the end-point carbon content Within the target error Value ±0 . 02% is 88 . 9% , and the prediction hit rate of the end-point molten steel temperature Within the target error Value ± 20 ℃ is 92 . 3% , Which is significantly im- proVed , and proVides a reference for end point prediction and control of conVerter steelmaking.
Currently , many imported temper mill flatness closed-loop feedback control systems employ a sequential control strategy , Which proVes to be ineffectiVe in controlling compound WaVes and fourth- order flatness deViations. MoreoVer , these systems struggle to handle the collaboratiVe regulation among different flatness control mechanisms. Focusing on a six-high CvCtemper mill at a steel plant , the existing sequential control strategy Was optimized and a flatness closed-loop feedback control strate- gy based on the RMsprop gradient descent algorithm Was proposed. The proposed strategy considers three regulation methods : Work roll bending , intermediate roll bending , and CvCshifting , aiming to quickly determine the combination of regulation quantities that minimizes flatness deViations. This en- sures the full utilization of each control mechanism s(/) potential. simulation and comparatiVe experiments demonstrate that the flatness control strategy based on the RMsprop algorithm significantly enhances the control of high-order and complex flatness defects , markedly improVing flatness quality compared to the original sequential control strategy . Furthermore , the RMsprop-based flatness control strategy Was implemented Via CFC programming , addressing program load rate issues on-site by integrating the strategy into the primary control system in parallel. Results confirm that the RMsprop-based strategy achieVes excellent application outcomes.
surface defects are common quality issues in strip steel production. To improVe production quality and assist engineers in deVeloping production optimization strategies , a method for identifying key factors of strip steel surface defects based on adaptiVe integrate gradient (adaptiVe integrate gradi- ent , AIG) Was proposed. The proposed AIG aims to extract defect-related information from practical industrial data , and identifies the key Variables that are most likely to cause the defect. In the process of defect-related information extraction , defect labels are transformed to defect ratio through a special - designed method , meanWhile , deep neural netWorks (deep neural netWork , DNN) Were used to build a defect ratio prediction model , Which effectiVely preVents defect-related modeling issues caused by practical data characteristics such as data imbalance and data oVerlap. Additionally , based on the trained DNN, an adaptiVe baseline sample selection strategy Was constructed to optimize the gradient integration calculation process , alloWing the final Variable importance indicators to better reflect the correlation With defects , achieVing an effectiVe transformation from abstract features to highly interpret- able identification results. Finally , the effectiVeness of the proposed method Was Validated on an inclu- sion defect dataset.
LFrefining is a metallurgical process inVolVing multi-process , multi-steel type and complex enVironment , a single-process control model cannot meet the demand for full-process automation and intelligence in smelting. Based on data and experience , by adopting self-learning algorithms , mecha- nistic models , expert system , image recognition , mathematical model to create models for alloy , slag formation , temperature control , argon bloWing , and Wire feeding , and using a Variety of detection de- Vices and multi-functional robots to achieVe the intelligent temperature sampling , and closely integra- ting multi-process model and L1 automated control system , an intelligent control system for the full- process of LFrefining Was realized. The system adopts a modularized and multi-threaded control meth- od , completing the smelting steps by a single-step incremental Way through time sequence analysis. It is applied in 150t double-station LF, and the aVerage operation time is shortened about 4 min , the hit rate of molten steel prediction temperature error Within ± 8 ℃ is more than 93% , and the success rate of temperature measurement and sampling is 97% .
The connection and scheduling of the casting-rolling interface is an important factor affecting the production efficiency and economic benefits of iron and steel enterprises. For the continuous cast- ing and hot rolling production site of a steel enterprise , a multi-agent simulation model of the casting- rolling interface Was established , and the complete production and transportation process on the pro- duction site Were modeled by designing the roles and tasks of Various agents and the cooperatiVe inter- action mechanism betWeen agents. A Variety of different production rhythms Was simulated and ana- lyzed by multi-agent model , and the precise matching betWeen different production rhythms of continu- ous casting and different production rhythms of hot rolling on the production line Was studied. The sim- ulation results shoWthat the model can accurately reflect the actual production process , Which Verifies the effectiVeness of the simulation model. According to the simulation model , as the interVal time be- tWeen the different streams of the casting machine increases , the conVeying time of the hot slab in the process of traVersing trolley and the total conVeying time of the hot slab shoWed a trend of first decrea- sing and then rising , and reached their loWest When the interVal time betWeen the different streams of the casting machine Was 2 min. when only producing hot billets , With the increase in the casting speed , the aVerage total conVeying time increases , the total production time decreases , and the aVer- age reheating time increases first and then decreases after the casting speed aboVe 1 . 9 m/min. As a result of the optimization , the aVerage reheating time Was reduced to 95 . 7% and the aVerage total con- Veying time Was reduced to 68 . 2% of the pre-optimization leVel.
For the continuous annealing unit of cold rolling , in order to ensure the stability of the speed of the process section , the front and back of the process section Were equipped With looper for buffe- ring. HoWeVer , there is no mature model for calculating the limiting speed of the process section for strips of different sizes and coil Weights under the complex and Variable production conditions in the entry section. By quantitatiVely calculating the production rhythm of the entry section and the impact of the production-limited entry looper Volume on the process section speed , the entry section-limited process section limit speed setting strategy Was established , Which solVed the stopping accidents caused by the mismatch betWeen the speed of the entry section and the process section With the loWVolume of the entry looper ; and in combination With the actual conditions of the production site , it analyzes the potential influencing factors of the entrance section of tWo continuous annealing units in a cold rolling mill that restrict the speed of the process section. By improVing strip uncoiling rhythm in the entry sec- tion , increasing the tailing speed , shortening the Welding time and increasing the entry looper filling speed , the aVerage Value of the limiting speed of the process section of No. 1 and No. 2 continuous an- nealing units Was increased from 266 and 258 m ● min — 1 to 369 and 349 m ● min — 1 When the Weight of incoming rolls Was 27 t and the minimum Volume of the entry looper control Was 30% .
