复杂场景分类是遥感图像解译的一项重要内容。本文通过优化ResNet18深度残差网络和随机森林,实现了遥感图像复杂场景的高精度分类。首先通过数据扩充将数据库扩充以缓解因训练样本少带来的过拟合问题,然后采用ResNet18深度残差网络自动...复杂场景分类是遥感图像解译的一项重要内容。本文通过优化ResNet18深度残差网络和随机森林,实现了遥感图像复杂场景的高精度分类。首先通过数据扩充将数据库扩充以缓解因训练样本少带来的过拟合问题,然后采用ResNet18深度残差网络自动提取遥感图像场景特征,最后使用随机森林分类器实现复杂场景分类任务并分别在NWPU-RESISC45和UC Merced Land Use数据库上进行了实验。结果表明,本文模型场景分类准确率分别为98.86%和99.17%,与单独使用ResNet18深度残差网络相比,本文模型分类准确率分别提高3.36%和1.71%,相比于其他场景分类方法,本文模型分类准确率分别提高5.23%和1.55%。展开更多
通过水热法在锂硫电池正极材料硫碳复合物表面包覆纳米金属氢氧化物抑制多硫化物的穿梭,很好地改善了电池的循环性能。利用扫描电镜(SEM)、恒流充放电和交流阻抗等方法比较了不同包覆层氢氧化铝、氢氧化钴、氢氧化铈对锂硫电池性能的影...通过水热法在锂硫电池正极材料硫碳复合物表面包覆纳米金属氢氧化物抑制多硫化物的穿梭,很好地改善了电池的循环性能。利用扫描电镜(SEM)、恒流充放电和交流阻抗等方法比较了不同包覆层氢氧化铝、氢氧化钴、氢氧化铈对锂硫电池性能的影响。其中,用氢氧化铝包覆的硫碳复合材料显示了较好的电化学性能,在100 m A/g充放电条件下,首次充放电比容量为1 192 m Ah/g,80次循环后放电比容量为797 m Ah/g,容量保持率达67%。0.5 C条件下,放电比容量达754 m Ah/g。展开更多
Lithium-sulfur(Li-S)batteries have attracted enormous interest due to their super-high theoretical energy density(2600 W·h/kg)in recent years.However,issues such as lithium dendrites and the shuttle effect severe...Lithium-sulfur(Li-S)batteries have attracted enormous interest due to their super-high theoretical energy density(2600 W·h/kg)in recent years.However,issues such as lithium dendrites and the shuttle effect severely hampered the large-scale application of Li-S batteries.Herein,a novel bifunctional gel polymer electrolyte,poly(N,Ndiallyl-N,N-dimethylammonium bis(trifluoromethylsulfonylimide))-P(VDF-HFP)(PDDA-TFSI-P(VDF-HFP),PTP),was prepared by anion exchange reaction to tackle the above problems.Benefited from the interaction between TFSI-and quaternary ammonium ion in PTP,a higher lithium-ion transference number was obtained,which could availably protect Li metal anodes.Meanwhile,due to the adsorption interactions between PDDA-TFSI and polysulfides(LiPSs),the shuttle effect of Li-S batteries could be alleviated effectively.Consequently,the Li symmetric batteries assembled with PTP cycled more than 1000 h and lithium metal anodes were protected effectively.Li-S batteries assembled with this polymer electrolyte show a discharge specific capacity of 813 mA·h/g after 200 cycles and 467 mA·h/g at 3 C,exhibiting excellent cycling stability and C-rates performance.展开更多
文摘复杂场景分类是遥感图像解译的一项重要内容。本文通过优化ResNet18深度残差网络和随机森林,实现了遥感图像复杂场景的高精度分类。首先通过数据扩充将数据库扩充以缓解因训练样本少带来的过拟合问题,然后采用ResNet18深度残差网络自动提取遥感图像场景特征,最后使用随机森林分类器实现复杂场景分类任务并分别在NWPU-RESISC45和UC Merced Land Use数据库上进行了实验。结果表明,本文模型场景分类准确率分别为98.86%和99.17%,与单独使用ResNet18深度残差网络相比,本文模型分类准确率分别提高3.36%和1.71%,相比于其他场景分类方法,本文模型分类准确率分别提高5.23%和1.55%。
文摘通过水热法在锂硫电池正极材料硫碳复合物表面包覆纳米金属氢氧化物抑制多硫化物的穿梭,很好地改善了电池的循环性能。利用扫描电镜(SEM)、恒流充放电和交流阻抗等方法比较了不同包覆层氢氧化铝、氢氧化钴、氢氧化铈对锂硫电池性能的影响。其中,用氢氧化铝包覆的硫碳复合材料显示了较好的电化学性能,在100 m A/g充放电条件下,首次充放电比容量为1 192 m Ah/g,80次循环后放电比容量为797 m Ah/g,容量保持率达67%。0.5 C条件下,放电比容量达754 m Ah/g。
基金Project(21935006)supported by the National Natural Science Foundation of China。
文摘Lithium-sulfur(Li-S)batteries have attracted enormous interest due to their super-high theoretical energy density(2600 W·h/kg)in recent years.However,issues such as lithium dendrites and the shuttle effect severely hampered the large-scale application of Li-S batteries.Herein,a novel bifunctional gel polymer electrolyte,poly(N,Ndiallyl-N,N-dimethylammonium bis(trifluoromethylsulfonylimide))-P(VDF-HFP)(PDDA-TFSI-P(VDF-HFP),PTP),was prepared by anion exchange reaction to tackle the above problems.Benefited from the interaction between TFSI-and quaternary ammonium ion in PTP,a higher lithium-ion transference number was obtained,which could availably protect Li metal anodes.Meanwhile,due to the adsorption interactions between PDDA-TFSI and polysulfides(LiPSs),the shuttle effect of Li-S batteries could be alleviated effectively.Consequently,the Li symmetric batteries assembled with PTP cycled more than 1000 h and lithium metal anodes were protected effectively.Li-S batteries assembled with this polymer electrolyte show a discharge specific capacity of 813 mA·h/g after 200 cycles and 467 mA·h/g at 3 C,exhibiting excellent cycling stability and C-rates performance.
