Efficient utilization of sinter return fine is an important measure to reduce cost,increase efficiency,save energy and reduce emission.A new path of green and efficient utilization of return fine was proposed to produ...Efficient utilization of sinter return fine is an important measure to reduce cost,increase efficiency,save energy and reduce emission.A new path of green and efficient utilization of return fine was proposed to produce composite pellets.The metallurgical properties of composite pellets under the condition of hydrogen-rich blast furnace were studied.The experimental results indicate that the coated concentrate was consolidated for the composite pellets through normal Fe_(2)O_(3) recrystallization.Near the surface of core return fine,the liquid phase formed due to its low-melting point,assimilated the adjacent concentrate,and then consolidated with the temperature decreasing.Compared with regular pellets,the com-pressive strength and reduction swelling index of composite pellets were decreased,but the reducibility index and softening-melting properties were improved.In addition,the reduction degradation index of composite pellets was sig-nificantly higher than that of sinter.Therefore,adding composite pellets was conducive to indirect reduction in blast furnace,reducing fuel ratio and improving production efficiency.According to the effect of the roasting system on the metallurgical properties,the roasting temperature and time were determined as 1250℃and 30 min,respectively.The composite pellets can be produced under the traditional pelletizing process.展开更多
随着全球气候变暖,近年来极端降水事件及其引发的洪涝灾害频发,极端降水事件的模拟与精细化研究显得尤为重要。随着区域气象站网的加密建设,为极端降水事件的精细化研究提供可能。为了将区域站短序列数据应用到日极端降水量的研究中,本...随着全球气候变暖,近年来极端降水事件及其引发的洪涝灾害频发,极端降水事件的模拟与精细化研究显得尤为重要。随着区域气象站网的加密建设,为极端降水事件的精细化研究提供可能。为了将区域站短序列数据应用到日极端降水量的研究中,本研究首先基于年最大值法(annual maximum,AM)和超阈值峰值法(peak over threshold,POT)抽样方法与44种概率分布模型,选择最优抽样方法与概率分布模型,并在此基础上提出对于短序列数据计算日极端降水量的订正方案,通过国家站分析论证,优选出最佳订正方案,将该订正方法应用到只有短序列实测数据的区域站中,优选插值参数并比较不同空间插值方法对插值精度的影响,选择最优的插值方法实现日极端降水量的精细化研究。结果表明,POT1抽样方法与广义帕累托模型是最适用于计算河北省日极端降水量的抽样方法与模型;本研究提出的区域站订正与计算日极端降水量方法可行,将区域站考虑进来后与国家站联合插值使得在空间上更加精细。展开更多
由于加工工艺的影响,应用零件表面划痕形态也会呈现出多种不同的表现形式,受到背景纹理样式的干扰,零件表面的细微划痕依然难以得到准确分辨与检测,为解决此问题,设计基于深度学习的零件表面细微划痕检测方法。在深度学习网络的彩色空...由于加工工艺的影响,应用零件表面划痕形态也会呈现出多种不同的表现形式,受到背景纹理样式的干扰,零件表面的细微划痕依然难以得到准确分辨与检测,为解决此问题,设计基于深度学习的零件表面细微划痕检测方法。在深度学习网络的彩色空间环境中,对滞后多阈值进行分割处理,再联合投资回报率(return on investment, ROI)提取权限条件,完成基于深度学习的划痕特征参量提取。在此基础上,建立边缘模板,通过处置细微划痕配准需求的方式,实现对划痕的准确定位,完成基于深度学习零件表面细微划痕检测方法的设计与应用。与机器视觉型检测方法相比,深度学习型检测方法能够根据划痕所表现出的具体形态,对其进行检测与分辨,可避免背景纹理样式对于零件加工工艺的影响。展开更多
基金financial support from the National Natural Science Foundation of China (U1960205)China Minmetals Science and Technology Special Plan Foundation (2020ZXA01).
文摘Efficient utilization of sinter return fine is an important measure to reduce cost,increase efficiency,save energy and reduce emission.A new path of green and efficient utilization of return fine was proposed to produce composite pellets.The metallurgical properties of composite pellets under the condition of hydrogen-rich blast furnace were studied.The experimental results indicate that the coated concentrate was consolidated for the composite pellets through normal Fe_(2)O_(3) recrystallization.Near the surface of core return fine,the liquid phase formed due to its low-melting point,assimilated the adjacent concentrate,and then consolidated with the temperature decreasing.Compared with regular pellets,the com-pressive strength and reduction swelling index of composite pellets were decreased,but the reducibility index and softening-melting properties were improved.In addition,the reduction degradation index of composite pellets was sig-nificantly higher than that of sinter.Therefore,adding composite pellets was conducive to indirect reduction in blast furnace,reducing fuel ratio and improving production efficiency.According to the effect of the roasting system on the metallurgical properties,the roasting temperature and time were determined as 1250℃and 30 min,respectively.The composite pellets can be produced under the traditional pelletizing process.
文摘随着全球气候变暖,近年来极端降水事件及其引发的洪涝灾害频发,极端降水事件的模拟与精细化研究显得尤为重要。随着区域气象站网的加密建设,为极端降水事件的精细化研究提供可能。为了将区域站短序列数据应用到日极端降水量的研究中,本研究首先基于年最大值法(annual maximum,AM)和超阈值峰值法(peak over threshold,POT)抽样方法与44种概率分布模型,选择最优抽样方法与概率分布模型,并在此基础上提出对于短序列数据计算日极端降水量的订正方案,通过国家站分析论证,优选出最佳订正方案,将该订正方法应用到只有短序列实测数据的区域站中,优选插值参数并比较不同空间插值方法对插值精度的影响,选择最优的插值方法实现日极端降水量的精细化研究。结果表明,POT1抽样方法与广义帕累托模型是最适用于计算河北省日极端降水量的抽样方法与模型;本研究提出的区域站订正与计算日极端降水量方法可行,将区域站考虑进来后与国家站联合插值使得在空间上更加精细。
文摘由于加工工艺的影响,应用零件表面划痕形态也会呈现出多种不同的表现形式,受到背景纹理样式的干扰,零件表面的细微划痕依然难以得到准确分辨与检测,为解决此问题,设计基于深度学习的零件表面细微划痕检测方法。在深度学习网络的彩色空间环境中,对滞后多阈值进行分割处理,再联合投资回报率(return on investment, ROI)提取权限条件,完成基于深度学习的划痕特征参量提取。在此基础上,建立边缘模板,通过处置细微划痕配准需求的方式,实现对划痕的准确定位,完成基于深度学习零件表面细微划痕检测方法的设计与应用。与机器视觉型检测方法相比,深度学习型检测方法能够根据划痕所表现出的具体形态,对其进行检测与分辨,可避免背景纹理样式对于零件加工工艺的影响。
