基于LSP与GLCM方法的碎米识别特性研究

随着农业机械化的不断推进,稻米加工过程中的碎米率逐步上升,碎米的快速准确识别对于大米加工与分级等具有重要意义,高效碎米识别方法的建立与应用是提高碎米的利用价值的有效途径。以米粒为研究对象,采用计算机视觉技术提取米粒特征,实现碎米自动分类识别。首先对米粒图像预处理,提取4维形状特征,利用局部相似模式(LSP)和灰度共生矩阵(GLCM)提取米粒4维纹理特征,建立支持向量机(SVM)、线性判别分类器(LDA)、K近邻(KNN),随机森林(RF),4种机器学习方法训练碎米识别分类器进行分类识别,利用验证集数据分析各类分类器的性能特征。结果表明,基于纹理特征及形状特征的支持向量机分类器的碎米识别精度最高,正确率达到97.56%。

风冷燃料电池结构优化及装配压力设计

风冷质子交换膜燃料电池简化了传统燃料电池的结构,极大地增强了其便携性而备受关注。风冷燃料电池的装配压力以及密封件的设计对于电池的性能具有重要影响。通过建立三维模型,运用有限元方法对风冷石墨流场板及密封件进行结构力学仿真分析。结果表明当直流道石墨流场板中密封圈的厚度为0.50 mm,双极板长度为225 mm,装配压力为2.5 MPa时,风冷燃料电池双极板应力分布均匀性最好,电池性能最优。

造孔剂（NH4）2CO3和尿素含量对TiAl多孔材料性能的影响

以 Ti、Al 元素粉末为原料,分别添加(NH4)2CO3和尿素,利用偏扩散反应,制备高孔隙度,孔特征可控的 TiAl 金属间化合物多孔材料。通过 XRD、SEM、金相显微技术等表征手段,研究造孔剂含量对 TiAl 金属间化合物多孔材料的物相结构、孔径、总孔隙率、透气度、膨胀率及力学性能的影响。结果表明:造孔剂对多孔材料物相组成没有影响;分别添加 7%碳酸铵和尿素后,平均孔径分别对应为 30.6 μm 和 28.8 μm,体积膨胀率增大到45.6%和 44.4%,总孔隙率为 49.9%和 48.3%,透气度为 307.3 m3/(m2 kPa1 h1)和 302.1 m3/(m2 kPa1 h1),极限抗拉强度分别为 27.85 MPa 和 32.49 MPa。

Water Distribution and Removal along the Flow Channel in Proton Exchange Membrane Fuel Cells

Distribution expressions of total gas pressure and partial water vapor pressure along the channel direction were established based on lumped model by analyzing pressure loss in the channel and gas diffusion in the layer. The mechanism of droplet formation in the flow channel was also analyzed. Effects of the relative humidity, working temperature and stoichiometry on liquid water formation were discussed in detail. Moreover, the force equilibrium equation of the droplet in the flow channel was deduced, and the critical flow velocity for the water droplet removal was also addressed. The experimental results show that the threshold position of the liquid droplet is far from the inlet with the increase of temperature, and it decreases with the increase of the inlet total pressure. The critical flow velocity decreases with the increase of the radius and the working pressure.