柴油机进气道喷射甲醇的排放与经济性研究

在一台增压中冷的电控单体泵柴油机上采用柴油/甲醇组合燃烧（DMCC）方式进行排放特性与经济性试验研究。结果表明:采用DMCC方式能够大量减少NO_x排放,同时CO₂排放也有所降低。可以通过在排气管上加装氧化催化转换器（DOC）很好地清除掉DMCC方式增加的HC和CO排放,其催化效率在90%以上,比原机排放低很多,DMCC方式的甲醛排放比原机高,DOC对其催化率达98%,接近超低排放。DMCC模式外特性的热效率提高了7.607%,且存在一个最佳燃料匹配区间。     

Analysis of differential plaque depositions in the brains of Tg2576 and Tg-APPswe/PS1dE9 transgenic mouse models of Alzheimer disease

Adequate assessment of plaque deposition levels in the brain of mouse models of Alzheimer disease (AD) is required in many core issues of studies on AD, including studies on the mechanisms underlying plaque pathogenesis, identification of cellular factors modifying plaque pathology, and developments of anti-AD drugs. The present study was undertaken to quantitatively evaluate plaque deposition patterns in the brains of the two popular AD models, Tg2576 and Tg-APPswe/ PS1dE9 mice. Coronally-cut brain sections of Tg2576 and Tg-APPswe/PS1dE9 mice were prepared and plaque depositions were visualized by staining with anti- amyloid β peptides antibody. Microscopic images of plaque depositions in the prefrontal cortex, parietal cortex, piriform cortex and hippocampus were obtained and the number of plaques in each region was determined by a computer-aided image analysis method. A series of optical images representing a gradual increase of plaque deposition levels were selected in the four different brain regions and were assigned in each with a numerical grade of 1-6, where +1 was lowest and +6, highest, so that plaques per unit in mm(2) increased "sigmoidally" over the grading scales. Analyzing plaque depositions using the photographic plaque reference panels and a computer-aid image analysis method, it was demonstrated that the brains of Tg2576 mice started to accumulate predominantly small plaques, while the brains of Tg-APPswe/PS1dE9 mice deposited relatively large plaques.     

基于可能性信息质量合成的双模态红外图像融合算法选取

针对双模态红外图像在融合时异类差异特征两两合成出现信息冗余导致所选择的融合算法相互冲突,造成融合效果差甚至失效的问题,提出了一种基于可能性信息质量合成的双模态红外图像融合算法选取方法。首先计算双模态红外图像多融合算法下不同差异特征的融合有效度,利用可能性框架得到对应的可能性分布向量子集;其次计算向量子集的信息量和可信度,并对多个向量子集进行加权合成;然后构建基于信息质量的排序函数,得到每种融合算法下的非支配子集;最后构建多融合算法得分函数的联合分布对多种融合算法优化选择。实验结果表明,将基于质量来整合多个差异特征的方法运用于双模态红外图像融合算法选取中,所选出的融合算法在加权综合指标上高于其他算法均值55%以上,证明了本文方法的有效性和合理性;由多组实验算得本文方法平均耗时10.083 s,在时间效率上也符合实时图像融合应用的工程需求。     

基于可能性分布联合落影的红外图像融合算法选取

针对双模态红外图像实际融合需求中常涉及到异类差异特征协同优化融合,且现有差异特征属性无法针对性地调整融合算法进行有效驱动,导致融合效果差等问题,提出了基于可能性分布联合落影的双模态红外图像融合算法选取的方法。首先计算双模态红外图像差异特征幅值的融合有效度,利用K最近邻法得到差异特征幅值的概率密度分布,得到差异特征频次属性的分布;然后通过差异特征幅值属性与频次属性构造差异特征权重函数,建立异类差异特征权重函数与多融合算法间的可能性分布合成,得到异类差异特征权重函数多融合算法融合有效度的联合落影;最后构建融合性能指标动态选取最优融合算法。实验结果表明,本文方法所选出的最优融合算法在等级得分指标上明显优于其他算法,验证了本文将可能性分布联合落影运用于双模态红外图像最优融合算法选取中具有可行性。     

3D Printing of Poly(3-hydroxybutyrate) Porous Structures Using Selective Laser Sintering

Summary: Poly(3-hydroxybutyrate) (PHB) 3D porous cubes were successfully built with Selective Laser Sintering (SLS), one of the many existing 3D printing technologies. The resulting cubes presented shape and dimensions very close to the corresponding virtual model. Moreover, they were resistant to handling without presenting any visible damage. The PHB powder did not present variation in thermal properties and chemical composition after 32.15 hours of SLS process as observed by proton nuclear resonance (¹H NMR) and differential scanning calorimetry (DSC) analysis, indicating that it can be re-utilized to print additional structures without affecting the reproducibility of the process.     

有1分配格的若干种等价定义

对有单位元1的分配格,给出若干种等价定义,并根据分配格所满足的幂等律、交换律、结合律、吸收律和分配律证明了最新定义与原始定义的等价性。     

Bessel and annular beams for materials processing

Non‐Gaussian beam profiles such as Bessel or annular beams enable novel approaches to modifying materials through laser‐based processing. In this review paper, properties, generation methods and emerging applications for non‐conventional beam shapes are discussed, including Bessel, annular, and vortex beams. These intensity profiles have important implications in a number of technologically relevant areas including deep‐hole drilling, photopolymerization and nanopatterning, and introduce a new dimension for materials optimization and fundamental studies of laser‐matter interactions.     

改进的Zernike矩工业CT图像边缘检测

为提高工业计算机断层扫描(CT)图像亚像素边缘检测的精度和速度,研究了一种改进的Zernike矩边缘检测方法。该方法采用Sobel边缘算子快速检测出图像所有可能的边缘,通过Zernike矩算子对所有可能的边缘进行重新检测,最后,检测出图像的亚像素边缘并计算其精确位置。由于采用Sobel算子检测出可能的边缘使后续Zernike矩算子检测范围缩小,从而减小了运算量,提高了运算速度。对实际CT图像进行的实验结果表明:改进的Zernike矩工业CT图像边缘检测精度绝对误差<0.24 pixel,改进算法的运算速度提高了约70%。     

基于联邦数据服务的企业级信息处理服务

针对当前企业级信息处理业务中普遍存在的问题,结合数据网格技术,提出一种新型的联邦数据服务FDSS。FDSS建立了联邦数据服务模型FDSM,提供对广域分布环境下异构企业信息的高效处理和透明管理,并通过本域数据服务和多域联邦数据服务实现信息资源的统一访问和跨域共享。FDSS适用