西范坪矿区土壤铜元素的高光谱响应与反演模型研究

为解决传统的土壤地球化学测量方法成本高、效率低等问题,研究了利用可见-近红外光谱技术检测土壤重金属含量的简易方法。研究对西范坪矿区土壤反射光谱进行微分、连续统去除等六种变换,利用逐步回归法和皮尔逊相关系数选出对土壤铜含量敏感的特征波段,组成综合特征变量集,再应用不同的特征变量选取方法和参数建立估算模型并检验。结果表明：不同的光谱变换方法对土壤铜含量信息提取能力不同,每种光谱变换都对应特定的敏感波谱区间;基于综合光谱变换信息建立的土壤铜含量反演模型精度优于基于单种光谱变换信息建立的模型;利用综合光谱变换信息建立土壤铜含量反演模型,后向剔除法优于前向引入法和逐步回归法,当Removal取0.20时得到最优回归模型,其模型决定系数R2和预测模型决定系数R2_pre分别达到了0.851和0.830,建模均方根误差RMSEC和预测均方根误差RMSEP分别为0.349和0.468 mg·kg-1,能较好地检测土壤铜含量,同时为其他土壤重金属元素的光谱检测提供了思路。     

平面常宽凸集的Firey-Sallee定理

常宽凸集是一类广泛应用在机械设计、医学等领域的特殊几何图形.本文探讨平面中的常宽凸集,简化证明著名的Firey-Sallee定理,即宽度相等的正Reuleaux多边形中Reuleaux三角形的面积最小.     

电荷耦合器件的γ辐照剂量率效应研究

对电荷耦合器件进行了不同剂量率的γ辐照实验,通过多种参数的测试探讨了剂量率与电荷耦合器件性能退化的关系,并对损伤的物理机理进行分析。辐照和退火结果表明:暗信号和暗信号非均匀性是γ辐照的敏感参数,电荷转移效率和饱和输出电压随剂量累积有缓慢下降的趋势;暗场像素灰度值整体抬升,像元之间的差异显著增加;电荷耦合器件的暗信号增量与剂量率呈负相关性,器件存在潜在的低剂量率损伤增强效应。分析认为,剂量率效应是由界面态和氧化物陷阱电荷竞争导致的。通过电子-空穴对复合模型、质子输运模型和界面态形成对机理进行了解释。     

飞秒脉冲通过多层体光栅后的衍射光强分布

基于单层体光栅的耦合波理论和矩阵光学知识,推导出飞秒脉冲通过多层体光栅衍射后的耦合波方程,并由此得到衍射光强谱和瞬时衍射效率表达式.研究结果发现衍射光强谱和瞬时衍射光强分布是与多层体光栅的结构参量,如中间填充层厚度、光栅层厚度、光栅常量及体光栅材料的折射率调制度有关系的.通过调节这些参量,获得了不同形状和不同宽度的衍射脉冲.该结果可用于设计基于多层体光栅的光通信和光脉冲整形器件.     

Primary result of ^236U measurement with accelerator mass spectrometry at CIAE

The rare isotope ^236U has a half-life of 2.342（3）×107 years, and is produced principally by thermal neutron capture on ^235U. The isotopic atom ratio of ^236U/^238U depends on the integral thermal neutron flux received by the material of interest. ^236U is potentially useful as a ＂fingerprint＂ for indicating the presence of neutron-irradiated uranium usually originating from nuclear activity. By extracting negative molecular ion UO^- from the uranium oxide target, simulating the ^236U^16O^- beam transport with ^238U^16O^- and ^208Pb2^16O^- pilot molecular ion beam, transporting the ^236U-containing ion beam with a high resolution injection magnet analyzer and electrostatic analyzer system, and finally identifying and detecting ^236U with a time-of-flight detector （TOF）, a method for AMS （Accelerator Mass Spectrometry） measurement of ^236U was established on the HI-13 Accelerator AMS system at China Institute of Atomic Energy.     

Tuning for the first 9-cell TESLA cavity of PKU

A method based on circuit model is used to tune the first home-made 9-cell TESLA type superconducting niobium cavity at Peking University.After tuning,a flat field profile with a final π-mode frequency within 3 kHz of target frequency is achieved.The field flatness is measured by a bead-pull method,and the relative electric field is calculated from the frequency shift perturbed by the bead stepping along the axis of the cavity.     

Linear Analysis of Folded Double-Ridged Waveguide Slow-Wave Structure for Millimeter Wave Traveling Wave Tube

A novel slow-wave structure （SWS）, the folded double-ridged waveguide structure, is presented and its linear gain properties are investigated. The perturbed dispersion equation is derived and the small signal growth rate is caleulated for dimensions of the ridge-loaded region and the parameters of the electron beam. The novel structure has potential applications in the production of high power and broad band radiation. For a cold beam, the linear theory predicts a gain of 1.1-1.27dB/period and a 3-dB small-signal gain bandwidth of 30% in W-band. A comparison between the folded double-ridged waveguide SWS and folded waveguide SWS （FWSWS） shows that with the same physical parameters, the novel SWS has an advantage over the FWSWS on the bandwidth and electron efficiency.     

基于Bragg光纤光栅传感器的监测系统设计

设计了基于Bragg光纤光栅传感技术的监测系统,当光纤光栅所处环境的温度、应力、应变或其他物理量发生变化时,光栅的周期或纤芯折射率将发生变化,从而使反射光的波长发生变化,通过测量反射光中心波长的变化,就可以获得待测物理量的变化情况.该系统利用薄膜滤波器完成对外界的分布式测量,采用光强度检测,能够实现无缝监测.     

照度对测量三代微光像增强器MTF的影响分析

通过探讨像增强器MTF测试仪用光源的出射光照度对测量结果的影响,对透过测量狭缝的光强分布进行了分析.调节狭缝面的入射光照度,对不同照度作用下的调制传递函数进行了对比测量,经与微光像增强器的饱和输出亮度比较,得出:三代微光像增强器的MTF测试值随入射光照度分布呈抛物线分布,其最大值与微光像增强器的自动亮度控制特性有关.适当选择入射光照度,可确保被测像增强器既有足够的输出信噪比,而不进入饱和区域.     

稀疏阈值的超分辨率图像重建

为了解决基于字典学习的超分辨重构算法耗时过长的问题,提出了基于稀疏阈值模型的图像超分辨率重建方法。首先,将联合字典理论与图像块稀疏阈值方法相结合,训练得到高、低分辨率过完备图像字典对。接着,通过稀疏阈值OMP算法对图像特征块进行稀疏表示。然后,通过高分辨率字典重构出初始的超分辨图像。最后,通过改进迭代反投影算法对初始的超分辨图像进行全局优化,从而进一步提高图像重构质量。实验结果表明,超分辨图像重构平均峰值信噪比(PSNR)为30.1 d B,平均结构自相似度(SSIM)为0.937 9,平均计算时间为10.2 s。有效提高了超分辨重构的速度,改善了重构高分辨图像的质量。