Papers by Author: Fei He

Paper TitlePage

Authors: Fei He, Quan Yang, Bao Jian Wang
Abstract: With more and more process data acquired from manufacturing process, extracting useful information to build empirical models of past successful operations is urgently required to get higher product quality. Clustering is the important data mining methods, where feature extraction is a significant factor to ensure the accurate rate of clustering and classification. As a common non-linear feature extraction method, kernel principal component analysis (KPCA) uses the variance as the information metric, but the variance is not always effective in some cases. Since information entropy is nonlinear and can effectively represent the dependencies of features, the Renyi entropy is used as the information metric to extract the feature in this paper. Simulation data, Tennessee Eastman and hot rolling process data are used for model validation. As a result the proposed method has better performance on feature extraction, compared with traditional KPCA.
Authors: Fei He, Xiao Chen Wang, Zhi Guo Liang, Cui Ma
Abstract: In the hot strip rolling production process, mechanical properties are detected offline. The process data is used for clustering analysis to acquire the state in advance. During the mechanical properties detection, the outliers in the same steel grade are regarded as the focal points. Spectral clustering is one of the advanced methods, where the Euclidean distance as the common similarity measure, can only extract the features of the spherically distribution data and can not express the complex distribution data. In this paper geodesic distance is introduced to spectral clustering, which is used to do the production state clustering. The Tennessee Eastman and hot strip rolling process data are used for model validation, as a result the proposed method has better performance on clustering, compared with the Euclidean distance.
Authors: An Min Yin, Quan Yang, Xiao Chen Wang, Fei He, You Zhao Sun
Abstract: This paper described the application of a diffraction system based on X-ray area detector on pole figure measurement as well as corresponding computation of orientation distribution functions and the principle of rapid measurement texture. The impact of calculates the orientation distribution function on the conditions of the two-dimensional X-ray diffraction was analyzed; this was illustrated by an example of deep drawing steel sheets texture measurement. ̙̈́˰͇̱̓˰̶̴̿̾ͅ˰̸̱̈́̈́˰̷̱̹͂̿̈́̈́̾˰̸̵̈́˰͇̈́̿˽̴̵̹̹̱̼̽̾̓̿̾˰̴̵̵̳̈́̈́̿͂˰̈́̿˰̸̵̈́˰̵̱̹̱̀̀͂̿̀͂̈́˰̂θ position then fix it, reduce the sample rotation; the texture determination time can be significantly reduced. Reduce the Measuring range of angle χ˰̴̱̾ φ˰̴̵̿̓˰̾̿̈́˰̶̶̵̱̳̈́˰̸̵̈́ calculation of orientation distribution function, it also can significantly reduce the measurement of diffraction data. Several technical problems appeared on the on-line determination of texture based on an X-ray two-dimensional detector system and the possibility to improve the measurement speed and accuracy in the industrial production applications were then discussed.
Authors: Fei He, Zhi Guo Liang, Min Li, Jin Wu Xu
Abstract: In order to predict product quality and optimize production process, the product quality model needs to be built. According to the fact that the common methods always cost long training time and can not realize real-time update, an online product quality model based on the online support vector regression is here proposed. The real field data of zinc coating weights from strip hot-dip galvanizing are used for validation. The results show that the models based on the online support vector regression have a higher prediction precision and shorter training time than traditional support vector regression, which is convenient to complete the real-time update. The zinc coating weights forecasting model based on the online support vector regression for multi-group data has an average of the relative prediction error of 4.35%, thus for the model will be used as an analysis tools for the quality control.
Authors: Xiao Chen Wang, Quan Yang, Zhi Guo Liang, Fei He
Abstract: Towards the insufficiency of the research on asymmetric shape defects control, and the equipment of new asymmetric shape control device—inverse symmetric roll bending for UCM cold mill, in this paper, by ANSYS finite element simulation, high-precision rolls-strip coupling models were established, and the difference of control character among all the symmetric and asymmetric shape control methods were analyzed, then generalized overall shape setting control models with asymmetric shape control function was proposed. It was proved by field application that the control models can effectively control asymmetric and high-order shape defects with the coordination of all the symmetric and asymmetric shape control methods. Consequently, expansive research and apply prospect should be expected.
Authors: You Zhao Sun, Quan Yang, An Rui He, Xiao Chen Wang, Fei He
Abstract: The paper researched a seven-roll leveler in the cross cut shearing line of a factory. An analytical model of the roller intermesh was established based on the elastic-plastic bend theory. It was known the maximal deformation of strip happened on the second leveler roll in the leveling process by the analysis of FEM simulation for the calculation results of analytical model. However, the restriction applied to the strip at the entrance side was smaller than expected, which caused the deformation was smaller than the calculated value of analytical model. Aiming at this program, combined with the analysis of FEM simulation, it was improved by adding the adjustment of the intermesh of first roll.
Authors: Zhi Guo Liang, Quan Yang, Ya Dong Wan, Fei He, Xiao Chen Wang, Min Wang
Abstract: Nowadays, IOT (Internet of Things) technology for the future intelligent manufacturing is still in its initial stage. In steel production, especially in steel strip production process, the research on how to construct IOT architecture is lack of research. This paper focused on the construction of IOT in steel Strip production process and analyzed the development of the industrial wireless sensor network standards. Based on the requirements of industrial networking monitoring in steel strip production process, this study proposed a feasible IOT network architecture for steel strip production process, which provided the basis for promoting further application of IOT technology in steel flat production process.
Authors: Zhi Guo Liang, Xiao Chen Wang, Fei He
Abstract: As the detecting method of structured light, linear laser has been widely recognized in theory, but there still exist some problems when it comes to its application in industrial production. Based on the needs and features of steel bar and wire rod production, the present paper proposes the online detection of steel bar and wire rod’s cross-sectional shape and size by using linear laser as the structured light. Meanwhile, the present paper applies such a detecting method to the online detection of continuous casting billet’s size and the online automatic counting of steel bar and wire rod. Moreover, the validity and accuracy of the proposed detection are verified by experimental device and project cases.
Authors: Zhi Guo Liang, Quan Yang, Ke Xu, Fei He, Xiao Chen Wang, Min Wang
Abstract: Structured light 3D measurement technology with its simple structure, non-contact measurement, fast measurement speed and other advantages, has been widely used. Steel plate surface quality detection is not confined to the two-dimensional feature of gray detection, and local topography measurement for surface quality of steel plate detection becomes increasingly important. In this paper, steel plate surface 3D detection method based on structured light and the factors affecting the measurement accuracy are analyzed. Several effective methods of improving 3D detection accuracy are put forward. Compared with the traditional structured light 3D detection methods, the detection accuracy of new methods is remarkably improved, thus possessing better application values.
Authors: Min Wang, Yan Ping Bao, Quan Yang, Fei He, Xiao Chen Wang, Cui Ma, Zhi Guo Liang
Abstract: This paper studied the distribution of defects by ultrasonic testing and the source of the defect was verified by SEM/EDS. The results showed that the defects mainly distributed in inner and outer sides closed to the surface of the slabs. The number of the defects in inner side was 1.5~2 times of that in outer side. A window 40mm×180mm inclusion band in 1/8 of the inner side existed. The defects in hot-rolled plate mainly formed due to the bad plasticity and deformability inclusions, such as Al2O3, SiO2 and MgO·xCaO. So decreasing the inclusions in slab was important to decrease the defects in hot-rolled plate.
Showing 1 to 10 of 10 Paper Titles