基于数字孪生的自组网多模态数据快速融合

随着自组网性能的提升与大规模应用,其内部数据呈现出多模态特征,数据量更是达到了海量级别,为数据融合处理工作带来了极大的挑战。为此,提出基于数字孪生的自组网多模态数据快速融合方法。实时采集自组网数据后,利用NLM算法与卡尔曼滤波算法去除数据中的噪声与冗余信息。然后,构建数字孪生自组网（包括自组网、孪生网络与服务系统）,从服务系统加载的数据中提取多模态数据特征,搭建双线性融合模型,从而实现对多模态数据的快速融合处理。实验表明：应用该方法后,多模态数据融合过程的时延始终保持在3 s以下,融合后多模态数据质量系数可达到0.9,证明该方法具有更优的数据融合性能。     

热输入对铝合金光纤激光穿透焊缝成形的影响

在2.5mm厚6061铝合金光纤激光穿透焊接实验的基础上,着重研究了焊接热输入对焊缝成形的影响。结果表明,光纤激光穿透焊的焊缝熔宽随焊接热输入的增加而增加,且背面熔宽的增幅更快。激光功率越高,获得稳定全熔透焊缝的热输入调节范围越大。焊接接头熔合区附近为柱状晶组织,焊缝中心为柱状晶和等轴晶的混合组织。随焊接热输入的逐渐降低,焊缝区显微组织逐渐细化,且混合组织中等轴晶所占比例逐渐减少。另外,随热输入的逐渐降低,焊缝区的平均显微硬度缓慢增长,焊缝区上部和下部的显微硬度数据分散性逐渐减小,当焊接热输入约为90J/mm时,焊缝上部和下部的显微硬度对称性最好。     

煅烧温度对湿化学法制备AlN粉体的影响

以六水合氯化铝和尿素为原料,甲醇为溶剂,在NH₃气氛下通过湿化学法制备AlN粉体,并采用X射线衍射仪、扫描电子显微镜和激光粒度仪对产物进行表征。结果表明:煅烧温度在900 ℃以上可获得六方AlN粉体;煅烧温度为1 000 ℃时获得的AlN粉体具有球形特征,粉体的平均晶粒尺寸为17.0 nm,平均粒径为159.5 nm。使用透射电子显微镜进一步表征了AlN粉体的微观结构。拉曼光谱结合能谱分析表明,存在于AlN粉体中的非晶杂质不是残留的含碳副产物,而是粉体表面水解产生的氢氧化铝。     

一维交替型脉冲微分系统

讨论一维空间中超前型与滞后型交替的脉冲微分系统.首先考虑具常系数的脉冲微分系统平凡解稳定的充分条件;其次研究了具变系数的脉冲微分系统的振动性,并给出了其解的表示式.     

Effect of 6H-SiC （1120） substrate on epitaxial graphene revealed by Raman scattering

A nonpolar SiC(1120) substrate has been used to fabricate epitaxial graphene (EG). Two EGs with layer numbers of 8-10 (referred to as MLG) and 2-3 (referred to as FLG) were used as representative to study the substrate effect on EG through temperature dependent Raman scattering. It is found that Raman lineshifts of G and 2D peaks of the MLG with temperature are consistent with that of a free graphene predicted by theory calculation, indicating that the substrate influence on the MLG is undetectable. While Raman G peak lineshifts of the FLG to that of the free graphene are obvious, however, its lineshift rate (-0.016 cm^-1/K) is almost one third of that (-0.043 cm^-1/K) of a EG on 6H-SiC (0001) in the temperature range from 300 K to 400 K, indicating a weak substrate effect from SiC (1120) on the FLG. This renders the FLG a high mobility around 1812 cm²- ·V^-1-·s^-1 at room temperature even with a very high carrier concentration about 2.95× 10¹³ cm^-2 (p-type). These suggest SiC (1120) is more suitable for fabricating EG with high performance.     

Influence of defects in SiC （0001） on epitaxial graphene

Defects in silicon carbide(SiC) substrate are crucial to the properties of the epitaxial graphene(EG) grown on it. Here we report the effect of defects in SiC on the crystalline quality of EGs through comparative studies of the characteristics of the EGs grown on SiC(0001) substrates with different defect densities. It is found that EGs on high quality SiC possess regular steps on the surface of the SiC and there is no discernible D peak in its Raman spectrum. Conversely, the EG on the SiC with a high density of defects has a strong D peak, irregular stepped morphology and poor uniformity in graphene layer numbers. It is the defects in the SiC that are responsible for the irregular stepped morphology and lead to the small domain size in the EG.