Comprehensive prediction of surface quality of hot-rolled strip with multiple and single defects
SUN Jian-liang1, SUN Meng-qian1, GUO He-song1, JI Jiang2, XU Li-pu2, PENG Yan1
1. National Engineering Research Center for Equipment and Technology of Cold Rolled Strip,Yanshan University, Qinhuangdao 066004, Hebei, China; 2. China Nation Heavy Machinery Research Institute Co., Ltd., Xi′an 710032, Shaanxi, China
Abstract：In view of the problem that the surface quality defects of the hot-rolled strip are serious,have many types and different mechanisms,and cannot be diagnosed online. A method for the comprehensive diagnosis of surface quality problems of the hot-rolled strip with multiple defects and single defects was presented by combining the deep belief network model of each dimension reduction layer. Considering the traditional deep belief network model in the shortfalls,layer number and the number of nodes were put forward and set up the integration of each dimension reduction layer belief network model,nodes were reduced due to network algorithm and the layer number of forecasting error caused by uncertainty,and used the correlation between classic deep belief network forecasting results,and the forecasting results of dimension reduction layer integration was the only anticipation value. In view of the problem of surface quality multiple defects in the hot-rolled strip production,considering the different input parameters of different surface quality defects,a comprehensive diagnosis strategy for surface quality multiple defects and single defects of the hot-rolled strip was established. In a typical surface become warped in the process of hot rolling production line production leather,metal surface quality defects such as seal and burr problem,application integration of each dimension reduction deep belief network model and multiple flaws and single comprehensive diagnosis strategy,the results show that the surface quality defects of the model prediction accuracy were above 80%,for the hot-rolled strip surface defects with good diagnosis effect.
孙建亮, 孙孟乾, 郭贺松, 计江, 徐利璞, 彭艳. 热轧带材多缺陷和单缺陷表面质量综合预报[J]. 钢铁, 2021, 56(1): 51-58.
SUN Jian-liang, SUN Meng-qian, GUO He-song, JI Jiang, XU Li-pu, PENG Yan. Comprehensive prediction of surface quality of hot-rolled strip with multiple and single defects[J]. Iron and Steel, 2021, 56(1): 51-58.
[1] 王国栋. 我国钢铁工业主要技术发展方向[J]. 金属世界导报,2017-12-12(B08).(WANG Guo-dong. China′s iron and steel industry technology development direction[J]. World Metals,2017-12-12(B08).) [2] 汪水泽,刘洋,孙宜强,等. CSP流程薄规格热轧带钢氧化铁皮控制技术[J]. 钢铁,2018,53(1):54.(WANG Shui-ze,LIU Yang,SUN Yi-qiang,et al. Control technology of oxide scale for thin gauge hot rolled strip in CSP process[J]. Iron and Steel,2018,53(1):54.) [3] 杨哲,龙卓,杨晗,等.不锈钢中厚板质量提升工艺研究[J].轧钢,2019,36(5):84.(YANG Zhe,LONG Zhuo,YANG Han,et al. Research on improving the quality of stainless steel heavy plate[J]. Steel Rolling,2019,36(5):84.) [4] 王畅,于洋,李研,等.高强镀锌板表面发麻缺陷产生机理[J].中国冶金,2019,29(6):21.(WANG Chang,YU Yang,LI Yan,et al. Formation mechanism of pitting defect on surface of high strength galvanized sheet[J]. China Metallurgy,2019,29(6):21.) [5] 张宏亮,冯光宏,崔怀周. 大规格Q420B热轧角钢轧制中间道次开裂缺陷分析[J]. 钢铁,2019,54(9):73.(ZHANG Hong-liang,FENG Guang-hong,CUI Huai-zhou. Analysis of crack defects in Q420B angle steel during hot rolling process[J]. Iron and Steel,2019,54(9):73.) [6] 徐海卫,于洋,李飞,等.IF钢热轧薄板翘皮缺陷的产生原因及机制[J].钢铁,2012,47(9):53.(XU Hai-wei,YU Yang,LI Fei,et al. Origination mechanism of the edge shell defect on the hot-rolled IF steel strip[J]. Iron and Steel,2012,47(9):53.) [7] Pesin A,Pustovoytov D. Research of edge defect formation in plate rolling by finite element method[J]. Journal of Materials Processing Technology,2015,220:96. [8] GUAN Ming-fei,YU Hao. Fatigue crack growth behaviors in hot-rolled low carbon steels:A comparison between ferrite-pearlite and ferrite-bainite microstructures[J]. Materials Science and Engineering A,2012,559(2013):875. [9] David Bombac,Milan Terceij. Amelioration of surface cracking during hot rolling of AISI D2 tool steel[J]. Materials Science and Technology,2018,34(14):1723. [10] 高建红,杨军. 热轧IF钢溢出类毛刺产生机理及控制研究[J]. 轧钢,2018,35(4):17.(GAO Jian-hong,YANG Jun. Study on the generation mechanism of IF steel strip side over brim burr defect during hot rolling and its countermeasures[J]. Steel Rolling,2018,35(4):17.) [11] Hiroshi Utsunomiya,Kenichiro Hara,Ryo Matsumoto,et al. Formation mechanism of surface scale defect in hot rolling process[J]. CIRP Annals-Manufacturing Technology,2014,63(1):261. [12] 艾永好,徐科. 钢板表面在线监测技术的发展和展望[J]. 金属世界,2010(5):37.(AI Yong-hao,XU Ke. Development and prospect of on-line surface inspection of steel plates[J]. Metal World,2010(5):37.) [13] 徐科,杨朝霖,周鹏,等. 热轧带钢表面缺陷在线检测的方法与工业应用[J]. 机械工程学报,2010,46(8):1.(XU Ke,YANG Chao-lin,ZHOU Peng,et al. Technology of on-line surface inspection for hot-rolled steel strips and its industrial application[J]. Journal of Mechanical Engineering,2010,46(8):1.) [14] 白振华,王俊飞. 冷连轧机成品板面粗糙度控制技术的研究[J]. 钢铁,2006,41(11):46.(BAI Zhen-hua,WANG Jun-fei. Control technique for surface roughness of strip in cold tandem rolling[J]. Iron and Steel,2006,41(11):46.) [15] LI R,ZHANG Q D,ZHANG X F,et al. Control method for steel strip roughness in two-stand temper mill rolling[J]. Chinese Journal of Mechanical Engineering,2015,28(3):573.) [16] 张清东,张勃洋,李瑞,等. 钢板微观表面质量控制理论与技术研究进展[J]. 机械工程学报,2016,52(10):32.(ZHANG Qing-dong,ZHANG Bo-yang,LI Rui,et al. Advances in theory and technology for microscopic surface quality control of steel strip[J]. Journal of Mechanical Engineering,2016,52(10):32.) [17] WU Chu-han,ZHANG Liang-cha,QU Pei-lei,et al. A new method for prediction the three-dimensional surface texture transfer in the skin pass rolling of metal strip[J]. Wear,2019(426/427):1246. [18] 吴长春. 轧辊与钢板表面形貌及其控制的研究[D]. 北京:北京科技大学,2006.(WU Chang-chun. Study on Surface Morphology and Control of Roll and Steel Plate[D]. Beijing:University of Science and Technology Beijing,2006.) [19] 李志锋,麻永林,冯岩,等. 神经网络在冷轧301S不锈钢极薄带材表面硬度预测中的应用[J]. 钢铁,2014,49(5):63.(LI Zhi-feng,MA Yong-lin,FENG Yan,et al. Applications of neutral network in surface hardness prediction for cold rolling 301S stainless steel thin strip[J]. Iron and Steel,2014,49(5):63.) [20] 李洪波,夏春雨,张杰,等. 基于元胞自动机的冷轧工作辊表面形貌演变过程仿真与实验研究[J]. 机械工程学报,2020,56(3):207.(LI Hong-bo,XIA Chun-yu,ZHANG Jie,et al. Simulated and experimental study on attenuation process of the surface topography of cold rolling work rolls with cellular automaton[J]. Journal of Mechanical Engineering,2020,56(3):207.) [21] 尤媛,李洪波,夏春雨,等. 冷轧毛化工作辊表面粗糙度衰减过程的试验与数学模型研究[J]. 机械工程学报,2018,54(12):173.(YOU Yuan,LI Hong-bo,XIA Chun-yu,et al. Experimental and mathematical model study of attenuation process of the surface roughness of textured work rolls during cold rolling[J]. Journal of Mechanical Engineering,2018,54(12):173.) [22] 王少,陈斌,司小明,等. 热轧板带表面质量智能化自动判定系统的开发应用[J]. 中国冶金,2019,29(7):70.(WANG Shao,CHEN Bin,SI Xiao-ming,et al. Development and application of intelligent automatic judgment system for hot rolled strip surface quality[J]. China Metallurgy,2019,29(7):70.) [23] YANG Hu,YONG Yu. Learning restricted Boltzmann machines using mode-hopping MCMC[C]//Proceedings of the 2012 4th International Conference on Machine Learning and Computing. Xi′an: IACSIT—International Association of Computer Science and Information Technology,2012.
2021, Vol. 56 
