Search

In the hot rolling process, strip shape is a great factor in assessing the quality of non-oriented silicon steel products. As the downstream industry has increasingly quality requirements for silicon steel products, the traditional predictive model is unable to meet the demand for silicon steel control. In order to improve the accuracy of the pre-set crown model, a stochastic configuration network (SCN) is used to build a data-driven crown prediction model. Additionally, we conduct correlation analysis of all features using Spearman correlation analysis to mitigate the influence of redundant variables. To improve the network's performance, we investigated the effect of weights and biases on SCN performance in four different distributions. We utilized a zero-symmetric distribution with adaptive parameters to decrease the model's training time and enhance its generalization performance. The results indicate that the model achieved the highest prediction accuracy under the normal distribution of adaptive parameters, with the correlation coefficient of 0.9483. The accuracy of crown deviation within ±5 μm is 96.3 %, the root mean square error and the mean absolute error are 2.525 and 1.925 respectively. Furthermore, the model prediction results were used to analyze the impact of four key rolling parameters on the crown of silicon steel. Compared to the original SCN, this method can effectively enhance the prediction accuracy of the silicon steel crown.

Flatness defects are complicated issues that occur during hot-rolled strip production. With more research on flatness control during the rolling process, flatness control technology during the rolling stage has improved. However, there is no online method for measuring strip flatness during the cooling stage after rolling. Therefore, this study focused on the evolution of stress and flatness of steel strips during run-out table cooling. Providing guidance for strip flatness controlled during run-out table cooling. A fully coupled finite element model based on temperature, phase transitions, and strain was developed. The calculated results of the model were in good agreement with the measured data. Furthermore, the patterns of evolution of stresses and flatness of the steel strip during run-out table cooling were obtained. The results showed that large compressive stresses formed along the edge region, as well as small tensile stresses in the middle region, causing edge wave defects to occur. With an increase in the initial temperature drop distance of the strip edge area, the edge wave degree also increased. Moreover, the cooling efficiencies of the upper and lower steel strip surfaces were not consistent, resulting in uneven distributions of stresses along the thickness of the material. Finally, the longitudinal and transverse warpage phenomena during the cooling process were discussed in detail.

The closed-loop edge drop control system of cold rolled silicon steel strip is a typical time-delay control system, and the measurement time-delay has a great influence on the stability of the control system. In order to improve the stability and control accuracy of the closed-loop edge drop control system, an edge drop control model based on model predictive control for cold rolled silicon steel strip is proposed in this paper. In order to improve the accuracy of control system model parameters, the priori efficiency coefficients of work roll shifting (WRS) for edge drop control are determined by rolling experiments, and the online optimization of priori efficiency coefficients is realized by process data. On this basis, the time delay analysis of control system is completed by equal flow method, and a closed-loop control strategy for edge drop control is developed by introducing predictive aging value into model predictive control. Finally, trajectory smoothing is used to complete the smooth output of predicted control, and then improve the dynamic characteristics of the closed-loop edge drop control system. Theoretical analysis, simulation experiments and application show that the model prediction-based edge drop control method has higher accuracy and faster response characteristics compared with conventional control method, and can realize high-precision edge drop control process of cold rolled silicon steel strip.

In plate hot rolling, the prediction of head deformation of steel plates is crucial to improve the yield. In this paper, a novel prediction model based on neural networks for head deformation of medium-thick plates is proposed. The dataset of parameters of head deformation of medium-thick plates was established based on machine vision, and the artificial neural networks optimized by improved sparrow search algorithm (ISSA-ANN) was proposed to predict the irregular length and area of the head of finished medium-thick plates. During model parameters setting, mean error (MAE), root mean square error (RMSE), mean absolute percentage error (MAPE), and correlation coefficient (R) are used as the evaluation indexes. Results show that ISSA-ANN performs better than other models in this paper in predicting the head deformation of medium-thick plates. Meanwhile, the effect of key variables on head deformation is investigated and suggestions for improving the head shape of medium-thick plates are given according to the influence of key variables of the model.

During the production of cold-rolled strips, the high-speed operation of rolling mills generates vibrations that significantly impact the production efficiency, the strip quality, and the rolling mill service life. Therefore, investigating the vibration characteristics of rolling mills is of paramount importance. Previous studies predominantly treated the rolls as point masses, overlooking variations in the vibrations along the lengths of the rolls. In this study, the working, intermediate, and backup rolls were regarded as continuous elastic bodies, and they were simplified as Timoshenko beams. The modal superposition method was used to solve the vertical vibration equations for the rolls of a six-high rolling mill; thus, the natural frequencies and modal shapes for the rolls were obtained for both free and forced vibrations. A finite-element model validation confirmed that the accuracy of the novel Timoshenko beam six-high rolling mill vibration model introduced in this paper is significantly higher than that of the traditional Euler beam model when the roller length-to-diameter ratio is relatively small. Further analysis based on this model indicates that the forced vibration of the rolls is primarily influenced by fluctuations in rolling force. The amplitude of forced vibration is affected by factors such as the friction coefficient in the deformation zone, rolling speed, and cumulative rolling distance of the work rolls. These findings provide a precise representation of the characteristics of forced vibrations in rolling mills and offer valuable theoretical insight to ensure stable rolling mill operation.

Background: Most doctors are used to locate the vascular stenosis position first then clinically estimate vascular stenosis by the CAG images instead of using mouse, keyboard and computer during the preoperative diagnosis. Therefore, it is necessary to have an effective, intuitive and natural method to replace the existing preoperative diagnosis mode for diagnosing the coronary heart disease. Methods: This work presents the gesture interactive diagnostic method for coronary heart disease based on wavelet transform and Semi-continuous Hidden Markov Model in Augmented Reality. Firstly, the dynamic gesture trajectory is analyzed by using K-means cluster algorithm to determine the gesture trajectory plane. Secondly, the integrity of the trajectory is preserved on the sensor plane by employing trajectory projection and space rotation transformation. Thirdly, the features include the azimuth of gesture trajectory and the changes of hand shape are extracted. Meanwhile, the azimuth of gesture trajectory is analyzed by utilizing wavelet transform and the singularity of the trajectory is detected to compute the state number of SCHMM for all the dynamic gestures. The gesture training is completed by employing Baum-Welch algorithm with multi-features and multi-observation sequences. Ultimately, gesture recognition is realized by using Viterbi algorithm based on self-learning mechanism. It is then applied into the gesture interactive diagnosis of coronary heart disease and the diagnosis of the vascular stenosis are intuitively and naturally realized with augmented reality and gesture interactive techniques. Results: The number of states of dynamic gesture is determined by using wavelet transform which precisely detect singular points of gesture trajectory. The recognition rate of dynamic gesture is average of 96.1% and the semi-dynamic gesture is average of 92.6%, the total recognition rate is average of 95.6%. The dynamic gesture recognition rate of the proposed method is 7% higher than the traditional method. The intuitive and natural gesture interactive experiments and the affirmed measuring results are approved by surgeons. Conclusions: The proposed method is a feasible, usable and reliable gesture interactive diagnostic method for coronary heart disease based on wavelet transform and Semi-continuous Hidden Markov Model in Augmented Reality. The results demonstrate that the proposed method could improve the gesture recognition rate, extend the gesture manipulation space and expand the diagnostic method for coronary heart diseases.

Edge rolling plays a vital role in width control of strip rolling, where plastic deformation concentrates in the edge of the slab. Then, the center of the slab remains its initial shape. The local deformation is different from overall one which obtained in horizontal rolling. In this paper, dog-bone shape is creatively described by a continuous symmetric parabola used to describe, and based on the complex mathematical function, a kinematically admissible continuous velocity field for edge rolling is established. Finally, the upper deformation power function is yielded. Based on minimum of the power, the results in edge rolling force and torque are obtained. Some vertical rolling forces for copper and aluminum experiments are calculated by the function, and then, also the results are verified by the traditional models. Anyway, the maximum error is less than 18 %. This is a crossover study on both mathematical curve and mechanical analysis which can be seen as a combination of the theory and engineering.

The inaccuracy of roll thermal expansion and roll wear settings affects strip steel plate shape and thickness in hot-rolling process. Roll wear and thermal expansion are nonlinear time series that cannot be neglected in the roll engineering. In order to make up for the shortcomings of the existing models, a novel prediction model of roll thermal expansion and roll wear based on Informer is developed. Firstly, multiple experiments are conducted to determine hyperparameter such as training epoch, dropout, batch size and learning rate. Then, the predicted results of finishing mill group (F1–F7 stand) are compared with the long-short-term memory (LSTM) model, the artificial neural network (ANN) model and the recurrent neural network (RNN) model. The results show that the model based on Informer network has highly accurate in predicting roll thermal expansion and roll wear. Finally, the application results show that the hit rate is 98.412 % within the tolerance range ± 4 μm and the hit rate within the ±40 μm thickness tolerance deviation range reach 98.94 % which proves that the roll wear and thermal expansion model based on Informer has outstanding advantages in the hot rolling shape control system.

Variable gauge rolling is a new process to obtain a plate for which the thickness changes continuously by continuously and dynamically adjusting the roll gap upward and downward in the rolling process. This technology is an effective method for producing lightweight, low-cost, and economical plates. However, variable gauge rolling is an unsteady process, and the changes in the force and deformation parameters are complex. In this research, based on the minimum energy theory of the variational principle and considering the characteristics of the roll movement and workpiece deformation comprehensively, the internal plastic deformation, friction, shear and tension powers, and the minimum result of the total power functional in upward and downward rolling are obtained with the first integral and then with a variation of adopting the specific plastic power and strain rate vector inner product. The analytical results of the deformation and force parameters are also established using the variational method. Then the precision of this model is certified using the measured values in a medium plate hot rolling plant and the experimental data for Tailor Rolled Blank rolling. Good agreement is found. Additionally, the variation rule of bite angle, neutral angle, and location neutral points are shown, and the change mechanism of the friction parameter on the stress state effect coefficient is given in variable gauge rolling. This research proposes a new mathematical model for rolling process control that provides a scientific basis and technical support for obtaining an accurate section shape in variable gauge rolling production.

The path planning problem is an important problem in the research area of robot, games and group animation. This paper shows a 2.5-dimensional terrain grid which can reduce the amount of computation. By applying the fuzzy logic theory, the terrain trafficability of the rugged road can be evaluated based on different gradient, roughness, elevation difference; the trafficability factor can be achieved and applied to the heuristic function. The improved algorithm can solve the symmetry problem of path planning on uneven surfaces, reduce the search space.

In order to produce ultrafine grained structures, interstitial-free steel sheets have been processed using a novel severe plastic deformation technique semi continuous equal channel angular extrusion (SC-ECAE) which is based on equal channel angular extrusion (ECAE) in an incremental way. The deformation was carried out at room temperature and individual specimen was repeatedly processed to various passes. An overall grain size which is 0.55 um was achieved after 10 passes (or an equivalent total strain of 4.8). The present paper reports the evolution of microstructures during deformation, which were examined and characterized using high resolution EBSD in a field emission gun SEM. The mechanisms of grain refinement are discussed.

In the hot-rolling metal forming process, the consistency and accuracy of the thickness of the metal strip are the most important factors for the product quality control. The current method of utilizing a mechanism prediction model with pre-defined parameters does not perform well due to some limits on the model assumptions and environmental interference. Manually tuning these parameters of the mechanism model may even result in worse performance. To resolve this problem, an advanced randomized learner model, termed stochastic configuration network (SCN), is employed to build a data-driven prediction model which can be trained by using a dataset collected from a real-world hot-rolling production site. Based on the rolling theory and gray relational analysis (GRA), 36 features are selected as the inputs of the prediction model. Experimental results with comparisons show that our proposed method is feasible and outperforms other machine learning methods, such as deep learning models and the random vector functional link (RVFL) model. [ABSTRACT FROM AUTHOR]