High-chromium vanadium-titanium magnetite(HVTM)is a crucial polymetallic-associated resource to be developed.The allpellet operation is a blast furnace trend that aims to reduce carbon dioxide emissions in the future....High-chromium vanadium-titanium magnetite(HVTM)is a crucial polymetallic-associated resource to be developed.The allpellet operation is a blast furnace trend that aims to reduce carbon dioxide emissions in the future.By referencing the production data of vanadium-titanium magnetite blast furnaces,this study explored the softening-melting behavior of high-chromium vanadium-titanium magnetite and obtained the optimal integrated burden based on flux pellets.The results show that the burden with a composition of 70wt%flux pellets and 30wt%acid pellets exhibits the best softening-melting properties.In comparison to that of the single burden,the softening-melting characteristic temperature of this burden composition was higher.The melting interval first increased from 307 to 362℃and then decreased to 282℃.The maximum pressure drop(ΔPmax)decreased from 26.76 to 19.01 kPa.The permeability index(S)dropped from 4643.5 to 2446.8 kPa·℃.The softening-melting properties of the integrated burden were apparently improved.The acid pellets played a role in withstanding load during the softening process.The flux pellets in the integrated burden exhibited a higher slag melting point,which increased the melting temperature during the melting process.The slag homogeneity and the TiC produced by over-reduction led to the gas permeability deterioration of the single burden.The segregation of the flux and acid pellets in the HVTM proportion and basicity mainly led to the better softening-melting properties of the integrated burden.展开更多
The formation mechanism of calcium vanadate and manganese vanadate and the difference between calcium and manganese in the reaction with vanadium are basic issues in the calcification roasting and manganese roasting p...The formation mechanism of calcium vanadate and manganese vanadate and the difference between calcium and manganese in the reaction with vanadium are basic issues in the calcification roasting and manganese roasting process with vanadium slag.In this work,CaO–V_(2)O_(5) and MnO_(2)–V_(2)O_(5) diffusion couples were prepared and roasted for different time periods to illustrate and compare the diffusion reaction mechanisms.Then,the changes in the diffusion product and diffusion coefficient were investigated and calculated based on scanning electron microscopy (SEM) with energy dispersive X-ray spectroscopy (EDS) analysis.Results show that with the extension of the roasting time,the diffusion reaction gradually proceeds among the CaO–V_(2)O_(5) and MnO_(2)–V_(2)O_(5) diffusion couples.The regional boundaries of calcium and vanadium are easily identifiable for the CaO–V_(2)O_(5) diffusion couple.Meanwhile,for the MnO_(2)–V_(2)O_(5) diffusion couple,MnO_(2) gradually decomposes to form Mn_(2)O_(3),and vanadium diffuses into the interior of Mn_(2)O_(3).Only a part of vanadium combines with manganese to form the diffusion production layer.CaV_(2)O_(6) and MnV_(2)O_(6) are the interfacial reaction products of the CaO–V_(2)O_(5) and MnO_(2)–V_(2)O_(5) diffusion couples,respectively,whose thicknesses are 39.85 and 32.13μm when roasted for 16 h.After 16 h,both diffusion couples reach the reaction equilibrium due to the limitation of diffusion.The diffusion coefficient of the CaO–V_(2)O_(5) diffusion couple is higher than that of the MnO_(2)–V_(2)O_(5) diffusion couple for the same roasting time,and the diffusion reaction between vanadium and calcium is easier than that between vanadium and manganese.展开更多
As a data-driven approach, Deep Learning(DL)-based fault diagnosis methods need to collect the relatively comprehensive data on machine fault types to achieve satisfactory performance. A mechanical system may include ...As a data-driven approach, Deep Learning(DL)-based fault diagnosis methods need to collect the relatively comprehensive data on machine fault types to achieve satisfactory performance. A mechanical system may include multiple submachines in the real-world. During condition monitoring of a mechanical system, fault data are distributed in a continuous flow of constantly generated information and new faults will inevitably occur in unconsidered submachines, which are also called machine increments. Therefore, adequately collecting fault data in advance is difficult. Limited by the characteristics of DL, training existing models directly with new fault data of new submachines leads to catastrophic forgetting of old tasks, while the cost of collecting all known data to retrain the models is excessively high. DL-based fault diagnosis methods cannot learn continually and adaptively in dynamic environments. A new Continual Learning Fault Diagnosis method(CLFD) is proposed in this paper to solve a series of fault diagnosis tasks with machine increments. The stability–plasticity dilemma is an intrinsic issue in continual learning. The core of CLFD is the proposed Dual-branch Adaptive Aggregation Residual Network(DAARN).Two types of residual blocks are created in each block layer of DAARN: steady and dynamic blocks. The stability–plasticity dilemma is solved by assigning them with adaptive aggregation weights to balance stability and plasticity, and a bi-level optimization program is used to optimize adaptive aggregation weights and model parameters. In addition, a feature-level knowledge distillation loss function is proposed to further overcome catastrophic forgetting. CLFD is then applied to the fault diagnosis case with machine increments. Results demonstrate that CLFD outperforms other continual learning methods and has satisfactory robustness.展开更多
The antioxidant defense system is essential for animals to cope with homeostasis disruption and overcome oxidative stress caused by adverse environmental conditions such as hypoxia.However,our understanding of how thi...The antioxidant defense system is essential for animals to cope with homeostasis disruption and overcome oxidative stress caused by adverse environmental conditions such as hypoxia.However,our understanding of how this system works in subterranean rodents remains limited.In this study,Brandt’s vole Lasiopodomys brandtii was exposed to normoxia(21%O_(2))or hypoxia(mild or severe hypoxia:10%or 5%O_(2))for 6 h.Changes in key enzymes of the classic enzymatic antioxidant system at both mRNA and enzyme activity levels,and tissue antioxidant levels of the low-molecular-weight antioxidant system were determined in brain,liver,and kidney.Transcript levels of the upstream regulator NF-E2-related factor 2(Nrf2)were also measured.We found that the mRNA expression of Nrf2 and its downstream antioxidant enzyme genes in L.brandtii were relatively conserved in response to hypoxia in most tissues and genes tested,except in the liver.Hepatic Nrf2,Cu/Zn SOD,GPx1,and GPx3 levels were significantly upregulated in response to mild hypoxia,whereas Mn SOD level decreased significantly in severe hypoxia.Unmatched with changes at the RNA level,constitutively high and relatively stable antioxidant enzyme activities were maintained throughout.For the low-molecular-weight antioxidant system,an abrupt increase of cerebral ascorbic acid(AA)levels in hypoxia indicated a tissue-specific antioxidant response.Although hypoxia did not cause significant oxidative damage in most tissues tested,the significant decrease in antioxidant enzyme activities(GPX and GR)and increase in lipid peroxidation in the kidney suggest that prolonged hypoxia may pose a critical threat to this species.展开更多
基金supported by the National Natural Science Foundation of China (Nos.52174277 and 52204309)the China Postdoctoral Science Foundation (No.2022M720683).
文摘High-chromium vanadium-titanium magnetite(HVTM)is a crucial polymetallic-associated resource to be developed.The allpellet operation is a blast furnace trend that aims to reduce carbon dioxide emissions in the future.By referencing the production data of vanadium-titanium magnetite blast furnaces,this study explored the softening-melting behavior of high-chromium vanadium-titanium magnetite and obtained the optimal integrated burden based on flux pellets.The results show that the burden with a composition of 70wt%flux pellets and 30wt%acid pellets exhibits the best softening-melting properties.In comparison to that of the single burden,the softening-melting characteristic temperature of this burden composition was higher.The melting interval first increased from 307 to 362℃and then decreased to 282℃.The maximum pressure drop(ΔPmax)decreased from 26.76 to 19.01 kPa.The permeability index(S)dropped from 4643.5 to 2446.8 kPa·℃.The softening-melting properties of the integrated burden were apparently improved.The acid pellets played a role in withstanding load during the softening process.The flux pellets in the integrated burden exhibited a higher slag melting point,which increased the melting temperature during the melting process.The slag homogeneity and the TiC produced by over-reduction led to the gas permeability deterioration of the single burden.The segregation of the flux and acid pellets in the HVTM proportion and basicity mainly led to the better softening-melting properties of the integrated burden.
基金supported by the National Natural Science Foundation of China(Nos.52174277 and 51874077)the Fundamental Funds for the Central Universities,China(No.N2225032)+1 种基金the China Postdoctoral Science Foundation(No.2022M720683)the Postdoctoral Fund of Northeastern University,China。
文摘The formation mechanism of calcium vanadate and manganese vanadate and the difference between calcium and manganese in the reaction with vanadium are basic issues in the calcification roasting and manganese roasting process with vanadium slag.In this work,CaO–V_(2)O_(5) and MnO_(2)–V_(2)O_(5) diffusion couples were prepared and roasted for different time periods to illustrate and compare the diffusion reaction mechanisms.Then,the changes in the diffusion product and diffusion coefficient were investigated and calculated based on scanning electron microscopy (SEM) with energy dispersive X-ray spectroscopy (EDS) analysis.Results show that with the extension of the roasting time,the diffusion reaction gradually proceeds among the CaO–V_(2)O_(5) and MnO_(2)–V_(2)O_(5) diffusion couples.The regional boundaries of calcium and vanadium are easily identifiable for the CaO–V_(2)O_(5) diffusion couple.Meanwhile,for the MnO_(2)–V_(2)O_(5) diffusion couple,MnO_(2) gradually decomposes to form Mn_(2)O_(3),and vanadium diffuses into the interior of Mn_(2)O_(3).Only a part of vanadium combines with manganese to form the diffusion production layer.CaV_(2)O_(6) and MnV_(2)O_(6) are the interfacial reaction products of the CaO–V_(2)O_(5) and MnO_(2)–V_(2)O_(5) diffusion couples,respectively,whose thicknesses are 39.85 and 32.13μm when roasted for 16 h.After 16 h,both diffusion couples reach the reaction equilibrium due to the limitation of diffusion.The diffusion coefficient of the CaO–V_(2)O_(5) diffusion couple is higher than that of the MnO_(2)–V_(2)O_(5) diffusion couple for the same roasting time,and the diffusion reaction between vanadium and calcium is easier than that between vanadium and manganese.
基金supported by the National Natural Science Foundation of China(Nos.52272440,51875375)the China Postdoctoral Science Foundation Funded Project(No.2021M701503).
文摘As a data-driven approach, Deep Learning(DL)-based fault diagnosis methods need to collect the relatively comprehensive data on machine fault types to achieve satisfactory performance. A mechanical system may include multiple submachines in the real-world. During condition monitoring of a mechanical system, fault data are distributed in a continuous flow of constantly generated information and new faults will inevitably occur in unconsidered submachines, which are also called machine increments. Therefore, adequately collecting fault data in advance is difficult. Limited by the characteristics of DL, training existing models directly with new fault data of new submachines leads to catastrophic forgetting of old tasks, while the cost of collecting all known data to retrain the models is excessively high. DL-based fault diagnosis methods cannot learn continually and adaptively in dynamic environments. A new Continual Learning Fault Diagnosis method(CLFD) is proposed in this paper to solve a series of fault diagnosis tasks with machine increments. The stability–plasticity dilemma is an intrinsic issue in continual learning. The core of CLFD is the proposed Dual-branch Adaptive Aggregation Residual Network(DAARN).Two types of residual blocks are created in each block layer of DAARN: steady and dynamic blocks. The stability–plasticity dilemma is solved by assigning them with adaptive aggregation weights to balance stability and plasticity, and a bi-level optimization program is used to optimize adaptive aggregation weights and model parameters. In addition, a feature-level knowledge distillation loss function is proposed to further overcome catastrophic forgetting. CLFD is then applied to the fault diagnosis case with machine increments. Results demonstrate that CLFD outperforms other continual learning methods and has satisfactory robustness.
基金This work was supported by the National Natural Science Foundation of China(grant no.U2004152)the China Postdoctoral Science Foundation(grant no.2020M672264).
文摘The antioxidant defense system is essential for animals to cope with homeostasis disruption and overcome oxidative stress caused by adverse environmental conditions such as hypoxia.However,our understanding of how this system works in subterranean rodents remains limited.In this study,Brandt’s vole Lasiopodomys brandtii was exposed to normoxia(21%O_(2))or hypoxia(mild or severe hypoxia:10%or 5%O_(2))for 6 h.Changes in key enzymes of the classic enzymatic antioxidant system at both mRNA and enzyme activity levels,and tissue antioxidant levels of the low-molecular-weight antioxidant system were determined in brain,liver,and kidney.Transcript levels of the upstream regulator NF-E2-related factor 2(Nrf2)were also measured.We found that the mRNA expression of Nrf2 and its downstream antioxidant enzyme genes in L.brandtii were relatively conserved in response to hypoxia in most tissues and genes tested,except in the liver.Hepatic Nrf2,Cu/Zn SOD,GPx1,and GPx3 levels were significantly upregulated in response to mild hypoxia,whereas Mn SOD level decreased significantly in severe hypoxia.Unmatched with changes at the RNA level,constitutively high and relatively stable antioxidant enzyme activities were maintained throughout.For the low-molecular-weight antioxidant system,an abrupt increase of cerebral ascorbic acid(AA)levels in hypoxia indicated a tissue-specific antioxidant response.Although hypoxia did not cause significant oxidative damage in most tissues tested,the significant decrease in antioxidant enzyme activities(GPX and GR)and increase in lipid peroxidation in the kidney suggest that prolonged hypoxia may pose a critical threat to this species.