论文

针对宝钢硅钢常化退火过程中产生的退火炉辊印缺陷问题,通过实际生产的大数据与产品质量问题相结合,将数据挖掘、数据分析方法应用到实际,一定程度上解决了现场实际生产中的痛点,为现场生产提供决策支撑,避免了以前通过人工识别判定存在疏漏和无法定量判断的问题,形成了一套具有鲁棒性和可操作性的钢铁生产过程数据分析方法。通过智慧决策系统平台获取实际生产和表检仪数据,基于Pearson相关系数算法进行变量挑选和特征工程,并应用随机森林算法对数据建立分类预测模型,实现了质量问题的溯源和监控,通过数据量化预测了炉辊印缺陷是否可通过轧制消除的质量问题,识别准确率达到96.43%。 In views of the normalizing annealing furnace roll marks problem occurred in the process of normalizing annealing of silicon steel in Baosteel,by combining big data from actual production with product quality problems,data mining and data analysis methods were applied to actual production to solve the pain points and provide decision support,a robust and practical data analysis method for the steel production process has been developed,which avoided the previous problems of omission and non-quan... 

采用电解法和扫描电镜研究了300 t转炉-RH精炼钙处理对无取向硅钢板（%:≤0.005C、1.2～2.2Si、0.2～0.6Mn、≤0.20P、≤0.005S、0.2～0.6Al、0～0.01Ca）中夹杂物的影响。结果表明,钢中Al含量为0.25%和0.35%时,钢中溶解氧均小于1×10^-4%,钙处理后都会产生CaS夹杂物,尤其是含0.35%Al的钢水;钙处理可以有效减少钢中的夹杂物数量,尤其是0.5μm以下的微细夹杂物数量;钙处理后夹杂物的种类以AlN、CaS为主,同时还含有少量的氧化物夹杂物以及AlN-CaS复合夹杂物,尺寸主要为1.5～5.0μm。 The effect of 300 t converter-RH refining calcium treatment on inclusions in non-oriented silicon steel sheet （%:≤0.005C,1.2～2.2Si,0.2～0.6Mn,≤0.20P,≤0.005S,0.2～0.6A1,0～0.01Ca） has been studied by electrolysis and scanning electron microscope.Results show that with 0.25%and 0.35%Al content in steel,all the dissolved oxygen in liquid is less than 1×10^-4%,and the CaS inclusions are produced after calcium treatment,especial for the liquid containing 0.35%Al;the amount of inclusions in ste... 

选用国内外无取向硅钢标杆企业A、B、C的产品,采用非水溶液电解提取+SEM(EDS)方法,分析了其典型牌号对应成品试样中的夹杂物。结果表明,从磁性能控制角度而言,受钢的化学成分及生产工艺影响,A、B、C企业成品试样的夹杂物尺寸、种类、数量存在差异,这些差异对其磁性能产生显著影响;从夹杂物控制角度而言,A、B、C企业对应成品试样的夹杂物,以MnS、CuxS复合或者AlN、MnS、CuxS复合为主,其中1.0μm以下的夹杂物数量分别为2.34×107个/mm3、2.98×107个/mm3和11.98×107个/mm3,1.0μm以上夹杂物数量均很少,夹杂物的平均尺寸从大到小依次为A企业、B企业、C企业。 According to the production of three benchmarking manufacture enterprises A,B and C of non-oriented silicon steel at home and abroad,the inclusions in finished steel samples with typical grades were investigated by electrolytic extraction from non-aqueous solution and SEM(EDS).Results show that,from the viewpoint of improving the magnetic properties,affected by chemical composition and production technique,the size,type and amount of inclusions in finished steel samples are various,which will af... 

利用DSC、XRD,结合热焓分析对硅钢级氧化镁和二氧化硅的固相反应行为进行了研究。比较了不同二氧化硅原料及混合方法对硅钢级氧化镁反应性的影响,找到了一种更为有效的硅钢级氧化镁反应性的评估方法,克服了传统柠檬酸活性法在比较不同厂家生产的氧化镁的反应性上的局限性。进一步研究了Na₂B₄O₇添加剂、TiO₂添加剂对氧化镁和二氧化硅反应行为的影响,结果表明,Na₂B₄O₇添加剂含量的增加对反应有利,而TiO₂添加剂对反应有阻碍作用。 The solid-state reaction behavior between silicon-steel grade MgO and SiO₂ was investigated by DSC and XRD combined with enthalpimetric analysis.The influence of different kinds of SiO₂ and different preparation methods were studied and a specifical measurement that is more reliable was proposed to evaluate reactivity of silicon steel grade MgO.It overcomes the limitation of the traditional citric acid activity（CAA） method,which can not be used to evaluate the reactivity of... 

【机构】 宝山钢铁股份有限公司规划与科技部； ...

无取向硅钢中夹杂物的存在会抑止晶粒生长,使基体的均匀连续性中断,其在钢中的形态、含量及分布情况都不同程度影响着硅钢的性能,尤其是对磁性能起关键的作用。因此,全尺度分布考察夹杂物对无取向硅钢夹杂物的研究极为重要。本实验确定了适用于不同牌号无取向硅钢夹杂物全尺度分布的分析方法:样品制备—小样电解—过滤喷金—根据不同牌号的要求选择合适的放大倍率扫描观测—夹杂物颗粒的分类统计。通过统计的结果,结合电解的失重量可以得到不同尺度的体积分布数据。实验分析了不同牌号和工艺无取向硅钢夹杂物的种类、形貌、大小和尺度分布,并初步考查了夹杂物与磁性能的关系,对无取向硅钢的工艺研究具有一定参考价值。 Inclusions in non-oriented silica steel could inhibit the growth of grain and cause discontinuity of micro-structure.The configuration,content and size distribution of inclusion have different effects on the performance of silica steel,especially significant on the magnetic property.Therefore,it is very useful to completely characterize inclusions with full size distribution in silica steel.In our research,full size analysis method for inclusion in silica steel had been established as follows: s... 

磁性能指标是硅钢产品最关键的质量指标之一，但是目前磁性能判定100%依赖于样品的离线实验室检测结果，生产线配置的在线检测仪的测量结果由于精度问题，不宜直接用于成品牌号判级。本文在现有硅钢产品质量管控体系基础上，利用大数据技术对生产数据进行分析与建模，构建不同磁性能指标在线检测模型，并在现有信息系统上完成模型库的集成部署，实现硅钢产品全长、多指标磁性能结果的拟合数据输出，支撑取样优化、精准分切、辅助综合判定等功能应用，进一步优化硅钢产品质量管控体系。 The magnetic performance index of silicon steel products is one of the most critical quality indexes.However,at present,100%determination of magnetic performance depends on the offline laboratory test results of samples,and the measurement results of the online detector configured in the production line cannot be applied in practice due to the accuracy problem.Based on the existing quality control system of silicon steel products,big data technology was used to analyze and model the production d... 

结合工业化生产的无取向硅钢,进行了RH精炼添加稀土合金实验。结果表明,1.15%(质量分数)Si钢的脱硫反应,主要发生在添加稀土合金之后的前5min。最佳的稀土合金添加量为0.6～0.9kg/t钢。钢液经过稀土处理后,加入的稀土总量越多,稀土氧硫化物夹杂物的尺寸就越大,但热轧带钢再结晶效果会逐渐变差,成品带钢晶粒尺寸先是快速长大,而后逐渐减小。最佳的钢中存留稀土含量与钢的化学成分有关,应严格控制在2.0×10-3%～6.0×10-3%(质量分数)。在此范围内,钢的铁损先是快速降低,而后缓慢升高,钢的磁感应强度则单调降低。 Based on the industrial production of non-oriented electrical steel,rare earth(RE) alloy treatment during the RH refining process was studied.The results showed that the effects of desulfurization and total concentration of RE remained in steel mainly depended on the chemical compositions of different steel grades.For 1.15wt% Si steel grade,the desulfurization reaction mainly focused on the initial 5min after RE alloy added during the RH refining process.The suitable RE alloy addition was 0.6-0.... 

简要介绍了宝钢无取向电工钢的发展历程、产品大纲、电磁性能控制情况,以及高附加值新产品的开发、应用情况。概述了近年来宝钢无取向电工钢生产技术进步,主要包括夹杂物控制技术、连轧机高牌号硅钢制造技术、同板差控制技术、环保涂层开发技术等。 This paper briefly summarizes the development history, product catalogue, and magnetic properties of non-oriented electrical steel sheets at Baosteel, as well as the development and application of high-value-added steel grades. Recent advances in manufacturing electrical steel sheets are also introduced, including technologies for controlling inclusion, producing high-grade steel strips by a tandem rolling mill, and controlling the transverse thickness difference of steel sheets, as well as the ... 

提出以“云边一体化架构”构建硅钢智慧决策系统，来解决原硅钢制造L1～L5系统架构模式下的数字信息孤岛、业务功能割裂等问题。在此基础上，开发了云边协同的自学习型控制模型及业务决策模型，构建起硅钢“智慧大脑”，形成了以研发、制造、服务等核心业务数字化融合的智能化决策支持新模式，探索出一条钢铁制造业数字化、智能化转型之路。 SIDS(Silicon-steel Intelligent Decision-making System)based on \"cloud-edge integration architecture\" was proposed to solve the problems of data silos and business function fragmentation in the original L1～L5 system architecture.On this basis,the self-learning control model and decision-making model of cloud-edge collaboration were developed,the \"smart brain\" of silicon steel department was constructed,and a new intelligent decision-making support model of digital integration of core businesses s... 

采用RH精炼添加钙合金方式对硅钢进行钙处理。结果表明,钙合金添加量为0.67、1.00、1.67kg/t钢时,钢中钙含量分别为0、2×10-6、4×10-6;随着钙合金添加量增大,钢中夹杂物粒度逐渐由0～2μm向2～4、4～6μm偏移;不同钙处理条件下,钢中均存在粒径小于1μm和粒径为1～5μm的MnS、CuxS夹杂物,后者或单独存在,或同AlN、CaS夹杂复合;粒径为5～10μm区间,钢中的夹杂物基本以钙的氧、硫化物为主。与钙处理前相比,钙合金添加量为0.67、1.00、1.67kg/t钢时,粒径小于1.0μm的微细夹杂物减少幅度分别为68.06%、87.50%、94.94%。钙合金添加量为1.67kg/t钢时,可以去除钢中绝大部分的微细夹杂物。 Ca alloy was added into the liquid steel during RH refining,and the results show that Ca concentration in final Si steel sheets is insignificant,about 0,2×10-6 and 4×10-6 when the added amount of Ca is 0.67,1.00 and 1.67 kg/t steel,respectively.With the increase in the added Ca alloy amount,the inclusions in the steel gradually change from those of 0～2 μm to those of 2～4 and 4～6 μm.Under different Ca treatments,there exist MnS and CuxS inclusions whose size is below 1 μm as well as MnS and CuxS ... 

结合工业化生产的50A1300牌号无取向硅钢,分析了化学成分、RH精炼脱氧方式、板坯装炉温度以及热轧平整工艺等对磁性能的影响,探讨成品钢磁性能的改善。结果表明,采取改善后,50A1300牌号无取向硅钢的磁性能得到明显改善。2012年,该钢种平均铁损、磁感应强度分别达到5.26 W/kg、1.762 T,能够较好地满足用户市场需求及同行对标需要。 Based on the industrial manufacture of 50A1300 grade non-oriented silicon steel sheets,effects of chemical composition,deoxidization method in RH vacuum refining,charging temperature of steel slabs,hot rolling and flattening progress on magnetic properties are analyzed.The optimization methods of magnetic properties of steels are discussed. Results show that the magnetic properties are obviously improved by the adopted measures above. In 2012,the average core loss and the magnetic induction of 5...