Hot Rolling Production Research Articles

Retraction Note: Research on hot-rolling steel products quality control based on BP neural network inverse model

Retraction Note: Research on hot-rolling steel products quality control based on BP neural network inverse model

A novel cost-sensitive quality determination framework in hot rolling steel industry

In the hot rolling industry, a high-precision quality determination framework is the guarantee for intelligent decision-making regarding products. Conventional methods ignore the significance of faulty products, which are assigned the same attention as the qualified products. To overcome these limitations, this paper proposes a novel compensated cost-sensitive multi-layer perceptron with minority attention (CCS-MLP-MIAT) to construct a high-precision quality determination framework. We first transform an imbalanced multi-class ordinal industrial problem into multiple imbalanced binary ordinal classifications using one-versus-one (OVO) decomposition. Then, we implement a minority attention mechanism and a cost-sensitive loss for MLP. Subsequently, a novel cost compensation coefficient is introduced and determined by Artificial Hummingbird Algorithm (AHA) to further improve the identification rate of faulty products. The advantage lies in the model to allocate more attention to the faulty products, thus promoting intelligent decision-making of hot rolling production. The proposed method is validated on eight UCI benchmark datasets and two real-cases of hot-rolled strip crown by cross validation. Experimental results demonstrate that the excellent generalization and robustness of CCS-MLP-MIAT in addressing the quality determination of hot-rolled steel.

Stacking ensemble learning fused with whale optimisation algorithm based method for crown prediction of hot-rolled strip

Precision of strip shape is one of the important indexes to measure the quality of strip steel in hot rolling. Accurate prediction of strip shape can realise the timely adjustment of the production system, which is the important foundation for ensuring the quality and stable production of hot-rolled strip steel products. Given the poor accuracy of the traditional mechanism model, this paper proposes a stacking ensemble learning for crown prediction that consist of the random forest (RF), extreme gradient boosting (XGBoost), light gradient boosting machine (LightGBM), categorical boosting (CatBoost) and linear regression, whose hyperparameters are optimised by whale optimisation algorithm (WOA) Firstly, the key features are selected by analysing the feature importance and the Pearson correlation coefficient between rolling parameters. Then, the hyperparameters of the stacking ensemble learning are optimised by four meta-heuristic algorithms during modeling process, among which WOA has the best optimisation effect. Compared to RF, XGBoost, LightGBM, CatBoost and the stacking, proposed method has the highest prediction accuracy that the hit rate reaches 98.58% in the range of 5.0 μm of crown deviation. In addition, the effect of strip width, bending force, rolling force and roll shifting on crown are analysed based on Shapley additive explanations (SHAP).

Hot rolled prognostic approach based on hybrid Bayesian progressive layered extraction multi-task learning

Hot-rolled strip products have diverse applications, and enhancing the detection, diagnostics, and prognostics of product quality during hot rolling is essential. Nevertheless, the multivariable, strong coupling, nonlinear, and time-varying nature of the production process poses a rigorous challenge for accurate hot-rolled prognostics. This paper implements a progressive layered extraction (PLE) multi-task learning (MTL) framework to simultaneously estimate multiple quality indicators, such as strip crown, center line deviation, exit temperature, wedge, width, and symmetry flatness. Additionally, the paper proposes the implements of Hybrid Bayesian Neural Network (HBNN) experts and a gating network with attention mechanism to integrate private and shared task features. It also puts forth an auxiliary task involving a Variational Autoencoder with Generative Adversarial Networks (VAE-GAN) to extract latent states from the original sequence. Moreover, an adaptive joint loss optimization is employed to update the weight of individual task losses for MTL training problems, and three sets of field hot-rolled datasets are used for model evaluation. In the experimental validation, considering the noisy field data and limited conditions in the real hot rolled production, comparative experiments are conducted to demonstrate the improved generalization and robustness of the proposed MTL approach. These experiments involve different percentages of the total data, ranging from 5% to 20%, and various prediction horizons ranging from 1 to 50 steps for model establishment. In addition, the paper discusses the interpretation of the model and strategies for further enhancing model performance.

Modeling and Simulation of Shape Control Based on Digital Twin Technology in Hot Strip Rolling.

Focusing on the problem of strip shape quality control in the finishing process of hot rolling, a shape model based on metal flow and stress release with the application of varying contact rolling parameters is introduced. Combined with digital twin technology, the digital twin framework of the shape model is proposed, which realizes the deep integration between physical time-space and virtual time-space. With the utilization of the historical data, the parameters are optimized iteratively to complete the digital twin of the shape model. According to the schedule, the raw material information is taken as the input to obtain the simulation of the strip shape, which shows a variety of export shape conditions. The prediction absolute error of the crown and flatness are less than 5 μm and 5 I-unit, respectively. The results prove that the proposed shape simulation model with strong prediction performance can be effectively applied to hot rolling production. In addition, the proposed model provides operators with a reference for the parameter settings for actual production and promotes the intelligent application of a shape control strategy.

A hybrid artificial intelligence algorithm for fault diagnosis of hot rolled strip crown imbalance

In the production process of hot continuous rolling, due to the imbalance between the number of normal cases and fault cases, the traditional supervised learning methods are often unable to deal with them efficiently. In order to address this problem, a new Hybrid Artificial Intelligence Algorithm is proposed, which is based on Light Gradient Boosting Machine (LightGBM) and combines resampling technique with Support Vector Machine (SVM) to deal with multi-class imbalance problems. Specifically, Support Vector Machines-Synthetic Minority Oversampling Technique (SVM-SMOTE) is used to increase the training data, so as to change the distribution of training data and improve the accuracy of the model, and the Particle Swarm Optimization (PSO) algorithm is used to find the hyper parameters of the model to determine the optimal hyper parameters combination. At the same time, the method of SHapley Additive exPlanation (SHAP) based on game theory is used to reveal the factors affecting the convexity of strip steel plate in hot rolling process and its interpretability. In check to see the validity of the model, experiments were carried out on UCI data set and hot rolling production data set, and multi-model comparison was carried out. The values of Geometric mean (G-mean), macroscopic F1 fraction (F1-Macro) and Matthews correlation coefficient (MCC) tested on the hot rolling data set are 0.992, 0.992 and 0.987 respectively. The model has a good effect. The results show that the model is significantly better than traditional models in solving the problem of strip convexity diagnosis during hot rolling process.

Analysis of flatness and critical crown of hot-rolled strip based on thermal–mechanical coupled residual stress analytical model

Residual stress is introduced during hot-rolled strip production due to various factors, such as temperature distribution, crown change, and metal transverse flow. Excessive residual stress can cause flatness defects in the strip, while insufficient residual stress may result in deformation during subsequent cutting. To explore the mechanism of residual stress formation in hot-rolled strip and how to control it, an analytical model for the coupling of temperature, crown and residual stress is established. Firstly, based on the symplectic space Hamiltonian system, an analytical solution for the strip cross section temperature field in each inter-stand zone is obtained. Then the residual stress analytical model is derived based on the fiber rod model combined with the law of volume invariance, lateral metal flow function and high temperature stress relaxation phenomenon. Meanwhile, to verify the accuracy of the analytical model, a multi field coupled finite element model based on the measured data validation is established. The calculation results of the analytical model are in good agreement with the finite element model. Furthermore, the buckling model of strip in symplectic space Hamiltonian system is derived, and the critical buckling stress of different strip flatness defects is directly calculated. Finally, the residual stress analytical model is combined with critical buckling stress to calculate the critical exit crown ratio to keep the strip flat.

A reinforcement learning integral sliding mode control scheme against lumped disturbances in hot strip rolling

In the hot strip rolling production process, due to the influence of external disturbances and unmodeled errors which are defined as lumped disturbances (LDs), the looper operating angle and strip tension may not satisfy the requirements of normal production under the set values, which may lead to the product not satisfying the requirements. In this paper, an advanced reinforcement learning (RL) algorithm combined with integral sliding mode control (ISMC) is utilized to deal with LDs, and then the robustness of the system is improved. Firstly, the hot strip rolling is a complex nonlinear system, the looper model is linearized to be approximated around the operating point for simplicity. Secondly, the ISMC is addressed for LDs in hot strip rolling; meanwhile, due to the chattering phenomenon in conventional sliding mode control, a novel reaching law is designed; whereafter an off-policy RL algorithm is adopted to obtain the optimal controller. Finally, the effectiveness and superiority of the addressed method are elaborated by an example.

Deep learning-based prediction framework of temperature control time for wide-thick slab hot rolling production

Deep learning-based prediction framework of temperature control time for wide-thick slab hot rolling production

Development of a computer program for calculating the geometric parameters of the deformation zone and mill tuning parameters in the process of producing hot-rolled seamless tubes

Determining the optimal tuning parameters for a piercing mill is a labor-intensive process. There are various calculation methods available for determining the tuning parameters of a piercing mill but implementing these methods can be time-consuming. Therefore, the developed computer program for calculating the geometric parameters of the deformation zone during the piercing of round billets and the tuning parameters of the piercing mill represents an innovative tool for calculating the tuning parameters of the piercing mill. The program allows for the establishment of optimal parameter values required to achieve the desired quality and accuracy of the steel piercing process. The main functions of the program include the analysis and processing of input data, such as the billet diameter, steel grade, roll calibration, and piercing mill tools. Based on this data, the program utilizes a developed mathematical model to calcu-late the optimal tuning parameters of the piercing mill, considering the required process characteristics and the desired level of product quality. One of the key features of the program is its high accuracy and universality. It takes into con-sideration multiple factors that affect the piercing process and enables reliable production forecasting. The program has a user-friendly interface, making it convenient for both specialists and shop floor personnel in the field of hot-rolled seamless tube production. The developed computer program is a valuable tool for manufacturers and research centers involved in steel piercing. It helps improve production efficiency, enhance product quality, and reduce time and costs associated with piercing mill setup. The program represents a significant contribution to the development and improve-ment of piercing technology in industrial settings

On the use of inline phase transformation sensors in a hot strip mill: case studies

Dynamic cooling control process on the run-out table (ROT) in a hot strip mill (HSM) is an important step to achieve the hot-rolled steel product quality required by customers. The measurements of strip temperature and the amount of transformed phase on the ROT are complementary to each other for the tracking of the material state of the strip during the cooling process. Infrared pyrometers have been widely used for the measurement of strip temperatures, while commercial systems for inline measurement of steel’s phase transformation only became available in recent years. Since 2015, a Transformation Monitor sensor system using EMSpec Technology has been deployed on the ROT of the HSM#2 at Tata Steel in the Netherlands for inline measurement of phase transformation. This paper presents use case studies of the Transformation Monitor. It will demonstrate the advantages of in-situ real time measurement of phase transformation, not only to reveal the internal microstructure of steel strips but also to measure the evolution of transformation–time trajectory on the run-out cooling table, as a complimentary measurement to pyrometer-based surface temperature measurement.

A knowledge-based NSGA-II algorithm for multi-objective hot rolling production scheduling under flexible time-of-use electricity pricing

With the increasing demand and shortage of energy, the iron and steel industry has focused on achieving low-carbon and smart manufacturing. This study investigates an energy-oriented scheduling problem arising from hot rolling production under flexible time-of-use (TOU) electricity tariffs, where the electric power comes from both the enterprise-owned power plant and the state-owned power grid. First, we model the studied scheduling problem as a multi-objective prize-collecting vehicle routing problem with special constraints. This model selects slabs from the order pool and sequences them to form rolling units, aiming to simultaneously minimize the transition costs in specifications, penalty costs, and electricity costs. Next, we develop a knowledge-based NSGA-II algorithm (KB-NSGA-II) by introducing a self-adaptive recombination procedure and a shortest path search (SPS)-based local intensification procedure. Furthermore, the comparative experimental results indicate that the KB-NSGA-II outperforms other competing algorithms in terms of the following indicators: (1) the hypervolume indicator is improved by approximately 57%; (2) the Spacing indicator is decreased by approximately 40%; and (3) the success ratio indicator of KB-NSGA-II is at least 60% higher than that of other competing algorithms. Finally, we apply a multi-criteria decision-making method named VIKOR to a case study under flexible TOU pricing strategies. These comparison results and the case study confirm that the KB-NSGA-II algorithm obtained better computational performance than the other competing algorithms and provided more flexible solutions under TOU pricing.

DA-CBGRU-Seq2Seq Based Soft Sensor for Mechanical Properties of Hot Rolling Process

High-precision prediction of material properties is an advanced direction of hot rolled product technology in manufacturing processes. However, complex disturbance usually produces wrong or missing data in the acquisition and entry stages. The long process and the diverse operating conditions bring great challenges to constructing soft sensors for the hot rolling process. To address this problem, a new time-series soft sensor model framework is proposed for mechanical properties prediction. Firstly, in order to process the raw dataset, the isolation forest algorithm is used to remove the outliers, and the random forest algorithm is used to fill in missing values. Secondly, the auxiliary variables are extracted by the combination of the XGBoost algorithm and prior mechanistic knowledge. Then, a DA-CBGRU-Seq2Seq time-series model based on a two-stage attention mechanism is designed for soft sensor modeling. In the encoder stage, external factors are introduced by an entity embedding layer. A combination structure of bidirectional gated recurrent units and convolutional layers is designed for extracting and learning the input features adaptively with the input attention mechanism. In the decoder stage, the long-term time dependence of the time series is learned by the temporal attention mechanism. Finally, the reliability of the proposed sensor model is compared and verified by the hot rolling process. The results indicate that the proposed method has an accuracy of over 95% in predicting the mechanical properties of a real hot-rolling process dataset, and has better performance compared to other comparative soft sensor prediction methods.

TOTAL OXYGEN CONTENT IN LIQUID STEEL AS A CRITERION FOR EVALUATING NONMETALLIC INCLUSIONS IN PIPE STEELS

A statistical analysis of the total oxygen content in samples of pipe steels at the stage of continuous casting at the continuous-casting machines of the casting and rolling complex of JSC “Vyksa Metallurgical Plant” was carried out. The microstructures of rolled samples were studied, and the chemical composition of non-metallic inclusions was determined by X-ray spectral microanalysis. A laboratory analysis of liquid steel samples and hot-rolled steel products of the casting and rolling complex was carried out using the method of reduction melting in an inert gas flow. Relationships have been constructed between the duration of metal presence in a steel-pouring ladle and the magnesium content in non-metallic inclusions, as well as between the average score of non-metallic inclusions in rolled metal and the total oxygen content in liquid steel at the stage of continuous steel casting.

An Improved NSGA-II for Energy-oriented Hot Rolling Production Scheduling Problems

With the sharply increasing energy demand and the shortage of energy, the scheduling goal of hot rolling production (HRP) is to select and sequence slabs to reduce cost and energy consumption. A multi-objective HRP optimal scheduling model is formulated and transferred as a multi-objective prize collecting vehicle routing problem (MO-PCVRP) with special constraints that simultaneously minimizes the jump cost of width, thickness, and hardness between adjacent slabs, the total length of unselected slabs, and the energy cost. Furthermore, this paper introduces a greedy strategy to intensify the local search of the standard NSGA-II algorithm, and names it local-intensification NSGA-II (LI-NSGA-II). The comparison experiments show that the LI-NSGA-II algorithm obtains better comprehensive performance than other compared algorithms in solving multi-objective scheduling problems of HRP.

Justification of the Production Technology Efficiency for Low-Alloy Steels with Improved Properties and Quality at Low-Cost. Part 1. Hot-Rolled Products

Justification of the Production Technology Efficiency for Low-Alloy Steels with Improved Properties and Quality at Low-Cost. Part 1. Hot-Rolled Products

Controlled Coil Chilling As Backbone Link in Hot-Rolled Sheet and Plate Steel Production

Controlled Coil Chilling As Backbone Link in Hot-Rolled Sheet and Plate Steel Production

Examination of the adhesion of the scale formed on hot-rolled steel with different silicon content using a tensile test

In the hot-rolling process, steel was subjected to high temperatures which results in the formation of an oxide layer called scale. The oxide scale can be affected to the surface characteristic of the hot-rolled steel product. The scale must be completely removed from the surface of the steel before further processing. This research aimed to examine the adhesion behavior of scale on hot-rolled steel with different silicon contents (0.01, 0.12, 0.18, and 0.29 wt%) using a tensile testing machine with an observation setup. The results showed that the scale thickness decreased with increasing silicon content in the range of 9 µm to 12 µm. The scales were composed of hematite and magnetite. The results of the tensile test showed that the strain initiating the first spallation and the mechanical adhesion energy tend to increase with increasing silicon content. However, it decreased by 0.29 wt% Si hot-rolled steel. This result indicates that scale was difficult to remove after the hot rolling process for higher Si-containing steel.

Steel Strip Defect Sample Generation Method Based on Fusible Feature GAN Model under Few Samples.

Due to the shortage of defect samples and the high cost of labelling during the process of hot-rolled strip production in the metallurgical industry, it is difficult to obtain a large quantity of defect data with diversity, which seriously affects the identification accuracy of different types of defects on the steel surface. To address the problem of insufficient defect sample data in the task of strip steel defect identification and classification, this paper proposes the Strip Steel Surface Defect-ConSinGAN (SDE-ConSinGAN) model for strip steel defect identification which is based on a single-image model trained by the generative adversarial network (GAN) and which builds a framework of image-feature cutting and splicing. The model aims to reduce training time by dynamically adjusting the number of iterations for different training stages. The detailed defect features of training samples are highlighted by introducing a new size-adjustment function and increasing the channel attention mechanism. In addition, real image features will be cut and synthesized to obtain new images with multiple defect features for training. The emergence of new images is able to richen generated samples. Eventually, the generated simulated samples can be directly used in deep-learning-based automatic classification of surface defects in cold-rolled thin strips. The experimental results show that, when SDE-ConSinGAN is used to enrich the image dataset, the generated defect images have higher quality and more diversity than the current methods do.

A joint model for hot-rolled strip quality in advance prediction

ABSTRACT Aiming at the problem that previous product quality prediction models cannot directly assess product quality before hot rolling, this paper presents a joint model for process parameters generation and quality prediction based on Conditional Generative Adversarial Nets and Artificial Neural Network. After generated actual process parameters by the generation module, this model would predict the quality of products ahead of schedule according to the generated parameters before hot rolling, which do not rely on inputting actual process parameter online anymore. Finally, the model has been trained and tested with the actual data from a certain hot rolling plant. The experimental results show that the generated process parameters agree well with actual production and quality prediction accuracy can meet the production requirements, which confirms the proposed model can be applied to simulate the actual rolling process and predict strip quality ahead of hot rolling production, providing a reference for the adjustment of planning and scheduling in the future.