Previous analytical results on flow splitting are generalized to consider multiple boiling channels systems. The analysis is consistent with the approximations usually adopted in the use of systems codes (like RELAP5 ...Previous analytical results on flow splitting are generalized to consider multiple boiling channels systems. The analysis is consistent with the approximations usually adopted in the use of systems codes (like RELAP5 and TRACE5, among others) commonly applied to perform safety analyses of nuclear power plants. The problem is related to multiple, identical, parallel boiling channels, connected through common plena. A theoretical model limited in scope explains this flow splitting without reversal. The unified analysis performed and the confirmatory computational results found are summarized in this paper. New maps showing the zones where this behavior is predicted are also shown considering again twin pipes. Multiple pipe systems have been found not easily amenable for analytical analysis when dealing with more than four parallel pipes. However, the particular splitting found (flow along N pipes dividing in one standalone pipe flow plus N -1 identical pipe flows) has been verified up to fourteen pipes, involving calculations in systems with even and odd number of pipes using the RELAP5 systems thermal-hydraulics code.展开更多
A novel laser processing technique was developed for making channels in the nano regime in this paper.A Nd:YAG laser was used to dry fabricate micro channels(25μm～100μm di- ameter)in a 1 cm^3 fused silica substrate...A novel laser processing technique was developed for making channels in the nano regime in this paper.A Nd:YAG laser was used to dry fabricate micro channels(25μm～100μm di- ameter)in a 1 cm^3 fused silica substrate by thermal-induced processing.By controlling the locations of these initiating micro channels on a silica cube,1D-controllable self-connecting nano fractures can be formed as rectangular channels.These nano channels are smooth and with extremely high aspect ratio(～10~4 depth to width ratio).A possible mechanism is proposed to explain the formation of the nano channels.This laser-based nano channel fabrication technique is fast and inexpensive,and with potential applications in capillary electrophoresis and electro-osmosis driven nano-filtration.展开更多
视觉惯性里程计(visual-inertial odometry,VIO)通过融合视觉和惯性数据来实现位姿估计.在复杂环境中,惯性数据受噪声干扰,长时间运动会导致累积误差,同时大多数VIO忽略了模态间局部信息交互,未充分利用不同模态的互补性,从而影响位姿...视觉惯性里程计(visual-inertial odometry,VIO)通过融合视觉和惯性数据来实现位姿估计.在复杂环境中,惯性数据受噪声干扰,长时间运动会导致累积误差,同时大多数VIO忽略了模态间局部信息交互,未充分利用不同模态的互补性,从而影响位姿估计精度.针对上述问题,本文提出了一种基于注意力机制与局部交互的视觉惯性里程计(attention and local interaction-based visual-inertial odometry,ALVIO)模型.首先,该模型分别提取到视觉特征和惯性特征.其次,保留惯性特征的历史时序信息,并通过基于离散余弦变换(discrete cosine transform,DCT)的通道注意力机制增强低频有效特征,抑制高频噪声.接着,设计了多模态局部交互与全局融合模块,利用改进的分散注意力机制与MLP-Mixer逐步实现模态间的局部交互与全局融合,根据不同模态的贡献调节局部特征权重,实现模态间互补,再在全局维度上整合特征,得到统一表征.最后,将融合的特征进行时间建模和位姿回归得到相对位姿.为了验证模型在复杂环境下的有效性,对公开数据集KITTI和EuRoC进行了低质量处理并实验,实验表明,ALVIO相较于直接特征拼接模型、多头注意力融合模型、软掩码融合模型,平移误差分别减少了49.92%、32.82%、37.74%,旋转误差分别减少了51.34%、25.96%、29.54%,且具有更高的效率和鲁棒性.展开更多
利用高精度和稳定的AIRS/Aqua(Atmospheric InfraRed Sounder on board Aqua)数据对SVISSR/FY-2C(Stretched Visible and Infrared SpinScan Radiometer on board Feng Yun 2C)的两个分裂窗通道IR1(Infra Red 1,10.9μm)和IR2(InfraRed ...利用高精度和稳定的AIRS/Aqua(Atmospheric InfraRed Sounder on board Aqua)数据对SVISSR/FY-2C(Stretched Visible and Infrared SpinScan Radiometer on board Feng Yun 2C)的两个分裂窗通道IR1(Infra Red 1,10.9μm)和IR2(InfraRed 2,11.9μm)进行交叉辐射定标的方法,并利用赤道附近2006年12月和2007年12月的AIRS和SVISSR数据完成了交叉辐射定标,结果表明,SVISSR数据与卷积得到的AIRS数据高度线性相关,SVISSR/FY-2C传感器的两个分裂窗通道不仅存在定标误差,而且定标误差随时间的变化呈现增大的趋势。相对于AIRS/Aqua测量值,当SVISSR的通道亮温从220K变化到340K时,2006年12月IR1通道的温度调整量从5.8K变化到-4.4K,而2007年12月IR1通道的温度调整量从6.9K变化到-5.1K;2006年12月IR2通道的温度调整量从2.2K变化到-1.5K,而2007年12月IR2通道的温度调整量从6.3K变化到-6.1K。展开更多
文摘Previous analytical results on flow splitting are generalized to consider multiple boiling channels systems. The analysis is consistent with the approximations usually adopted in the use of systems codes (like RELAP5 and TRACE5, among others) commonly applied to perform safety analyses of nuclear power plants. The problem is related to multiple, identical, parallel boiling channels, connected through common plena. A theoretical model limited in scope explains this flow splitting without reversal. The unified analysis performed and the confirmatory computational results found are summarized in this paper. New maps showing the zones where this behavior is predicted are also shown considering again twin pipes. Multiple pipe systems have been found not easily amenable for analytical analysis when dealing with more than four parallel pipes. However, the particular splitting found (flow along N pipes dividing in one standalone pipe flow plus N -1 identical pipe flows) has been verified up to fourteen pipes, involving calculations in systems with even and odd number of pipes using the RELAP5 systems thermal-hydraulics code.
基金The project supported by the National Natural Science Foundation of China (50375031)the Hong Kong Research Grants Council (CUHK 4416/99E)
文摘A novel laser processing technique was developed for making channels in the nano regime in this paper.A Nd:YAG laser was used to dry fabricate micro channels(25μm～100μm di- ameter)in a 1 cm^3 fused silica substrate by thermal-induced processing.By controlling the locations of these initiating micro channels on a silica cube,1D-controllable self-connecting nano fractures can be formed as rectangular channels.These nano channels are smooth and with extremely high aspect ratio(～10~4 depth to width ratio).A possible mechanism is proposed to explain the formation of the nano channels.This laser-based nano channel fabrication technique is fast and inexpensive,and with potential applications in capillary electrophoresis and electro-osmosis driven nano-filtration.
文摘视觉惯性里程计(visual-inertial odometry,VIO)通过融合视觉和惯性数据来实现位姿估计.在复杂环境中,惯性数据受噪声干扰,长时间运动会导致累积误差,同时大多数VIO忽略了模态间局部信息交互,未充分利用不同模态的互补性,从而影响位姿估计精度.针对上述问题,本文提出了一种基于注意力机制与局部交互的视觉惯性里程计(attention and local interaction-based visual-inertial odometry,ALVIO)模型.首先,该模型分别提取到视觉特征和惯性特征.其次,保留惯性特征的历史时序信息,并通过基于离散余弦变换(discrete cosine transform,DCT)的通道注意力机制增强低频有效特征,抑制高频噪声.接着,设计了多模态局部交互与全局融合模块,利用改进的分散注意力机制与MLP-Mixer逐步实现模态间的局部交互与全局融合,根据不同模态的贡献调节局部特征权重,实现模态间互补,再在全局维度上整合特征,得到统一表征.最后,将融合的特征进行时间建模和位姿回归得到相对位姿.为了验证模型在复杂环境下的有效性,对公开数据集KITTI和EuRoC进行了低质量处理并实验,实验表明,ALVIO相较于直接特征拼接模型、多头注意力融合模型、软掩码融合模型,平移误差分别减少了49.92%、32.82%、37.74%,旋转误差分别减少了51.34%、25.96%、29.54%,且具有更高的效率和鲁棒性.
文摘利用高精度和稳定的AIRS/Aqua(Atmospheric InfraRed Sounder on board Aqua)数据对SVISSR/FY-2C(Stretched Visible and Infrared SpinScan Radiometer on board Feng Yun 2C)的两个分裂窗通道IR1(Infra Red 1,10.9μm)和IR2(InfraRed 2,11.9μm)进行交叉辐射定标的方法,并利用赤道附近2006年12月和2007年12月的AIRS和SVISSR数据完成了交叉辐射定标,结果表明,SVISSR数据与卷积得到的AIRS数据高度线性相关,SVISSR/FY-2C传感器的两个分裂窗通道不仅存在定标误差,而且定标误差随时间的变化呈现增大的趋势。相对于AIRS/Aqua测量值,当SVISSR的通道亮温从220K变化到340K时,2006年12月IR1通道的温度调整量从5.8K变化到-4.4K,而2007年12月IR1通道的温度调整量从6.9K变化到-5.1K;2006年12月IR2通道的温度调整量从2.2K变化到-1.5K,而2007年12月IR2通道的温度调整量从6.3K变化到-6.1K。
