作为钢铁冶金制造的核心工序,高炉炼铁是典型的高能耗过程,其运行能耗约占钢铁总能耗的50%以上,其中,80%的能耗是焦炭和煤粉等燃料消耗.因此,对表征高炉燃料消耗的燃料比参数进行监测,并尽可能早地识别影响燃料比异常波动的关键因素,对...作为钢铁冶金制造的核心工序,高炉炼铁是典型的高能耗过程,其运行能耗约占钢铁总能耗的50%以上,其中,80%的能耗是焦炭和煤粉等燃料消耗.因此,对表征高炉燃料消耗的燃料比参数进行监测,并尽可能早地识别影响燃料比异常波动的关键因素,对于高炉炼铁过程的节能降耗具有重要意义.本文针对先验故障知识少的高炉燃料比监测与异常识别难题,提出一种基于核偏最小二乘(Kernel partial least squares,KPLS)鲁棒重构误差的故障识别方法.该方法首先建立过程变量与监测变量的KPLS监测模型,然后根据非线性映射空间的协方差矩阵和核空间Gram矩阵之间的关系,反向估计原始空间变量的正常估值.为了增强算法的鲁棒性,采用迭代去噪算法减少异常数据对原始空间正常估值的影响.通过利用原始空间正常估值和真实值来构造故障识别指标,并给出故障识别指标的控制限.基于实际工业数据的高炉数据实验表明所提方法不仅可以监测出正常工况下影响燃料比异常变化的潜在因素,还可识别出异常工况下影响燃料比异常变化的关键因素,具有很好的工程应用前景.展开更多
The Karakoram Highway(KKH),a part of the China–Pakistan Economic Corridor(CPEC),is a major highway connecting northern Pakistan to China.The inventorying and analysis of landslides along KKH are challenging because o...The Karakoram Highway(KKH),a part of the China–Pakistan Economic Corridor(CPEC),is a major highway connecting northern Pakistan to China.The inventorying and analysis of landslides along KKH are challenging because of poor accessibility,vast study area,limited availability of ground-based datasets,and the complexity of landslide processes in the region.In order to preserve life,property,and infrastructure,and to enable the uninterrupted and efficient operation of the KKH,it is essential to strengthen measures for the prevention and control of geological disasters.In the present study,SBASInSAR(Small Baseline Subsets-Interferometric Synthetic Aperture Radar)was used to process 150 scenes of Sentinel 1-A images in the year 2017 along the Karakoram Highway.A total of 762 landslides,including 57 complex landslides,126 rock falls,167 debris slides,and 412 unstable slopes,ranging in size between 0.0017 and 10.63 km2 were identified.Moreover,this study also gains an inventory of 40 active glacier movements in this region.Landslide categorization,displacements characteristics,spatial distribution,and their relationship with various contributing factors have been successfully investigated along the entire KKH using image interpretation and frequency-area statistics.The criteria adopted for landslides categorization is presented in the study.The results showed that the 2-D ground deformation derived in Hunza valley echoes well with the general regional landslides characteristics.The spatial distribution analysis revealed that there are clumped distributions of landslides in the Gaizi,Tashkurgan,and Khunjerab in China,as well as in Hunza valley,and north of Chilas city in Pakistan.Statistical results indicated that these landslides mainly occur on south-facing slopes with a slope angle of 20°–45°and elevation relief of 550–2,100 m.Landslide development is also related to low vegetation cover and weathering effects in mountain gullies.Overall,our study provides scientific data support and theoretical references for p展开更多
文摘作为钢铁冶金制造的核心工序,高炉炼铁是典型的高能耗过程,其运行能耗约占钢铁总能耗的50%以上,其中,80%的能耗是焦炭和煤粉等燃料消耗.因此,对表征高炉燃料消耗的燃料比参数进行监测,并尽可能早地识别影响燃料比异常波动的关键因素,对于高炉炼铁过程的节能降耗具有重要意义.本文针对先验故障知识少的高炉燃料比监测与异常识别难题,提出一种基于核偏最小二乘(Kernel partial least squares,KPLS)鲁棒重构误差的故障识别方法.该方法首先建立过程变量与监测变量的KPLS监测模型,然后根据非线性映射空间的协方差矩阵和核空间Gram矩阵之间的关系,反向估计原始空间变量的正常估值.为了增强算法的鲁棒性,采用迭代去噪算法减少异常数据对原始空间正常估值的影响.通过利用原始空间正常估值和真实值来构造故障识别指标,并给出故障识别指标的控制限.基于实际工业数据的高炉数据实验表明所提方法不仅可以监测出正常工况下影响燃料比异常变化的潜在因素,还可识别出异常工况下影响燃料比异常变化的关键因素,具有很好的工程应用前景.
基金supported by National Key Research and Development Program of China(Grant No.2017YFC1501005)National Natural Science Foundation of China(Grant Nos.41661144046,42007232)+3 种基金the Science and Technology Major Project of Gansu Province(Grant No.19ZD2FA002)the Science and Technology Planning Project of Gansu Province(Grant No.18YF1WA114)the Fundamental Research Funds for the Central Universities(Grant Nos.lzujbky-2021-ey05)Second Tibetan Plateau Scientific Expedition and Research Program(Grant No.2019QZKK0902)。
文摘The Karakoram Highway(KKH),a part of the China–Pakistan Economic Corridor(CPEC),is a major highway connecting northern Pakistan to China.The inventorying and analysis of landslides along KKH are challenging because of poor accessibility,vast study area,limited availability of ground-based datasets,and the complexity of landslide processes in the region.In order to preserve life,property,and infrastructure,and to enable the uninterrupted and efficient operation of the KKH,it is essential to strengthen measures for the prevention and control of geological disasters.In the present study,SBASInSAR(Small Baseline Subsets-Interferometric Synthetic Aperture Radar)was used to process 150 scenes of Sentinel 1-A images in the year 2017 along the Karakoram Highway.A total of 762 landslides,including 57 complex landslides,126 rock falls,167 debris slides,and 412 unstable slopes,ranging in size between 0.0017 and 10.63 km2 were identified.Moreover,this study also gains an inventory of 40 active glacier movements in this region.Landslide categorization,displacements characteristics,spatial distribution,and their relationship with various contributing factors have been successfully investigated along the entire KKH using image interpretation and frequency-area statistics.The criteria adopted for landslides categorization is presented in the study.The results showed that the 2-D ground deformation derived in Hunza valley echoes well with the general regional landslides characteristics.The spatial distribution analysis revealed that there are clumped distributions of landslides in the Gaizi,Tashkurgan,and Khunjerab in China,as well as in Hunza valley,and north of Chilas city in Pakistan.Statistical results indicated that these landslides mainly occur on south-facing slopes with a slope angle of 20°–45°and elevation relief of 550–2,100 m.Landslide development is also related to low vegetation cover and weathering effects in mountain gullies.Overall,our study provides scientific data support and theoretical references for p
