Nano‐Graphene Oxide Based Multichannel Sensor Arrays towards Sensing of Protein Mixtures

Optical array‐based sensors are attractive candidates for the detection of various bio‐analytes due to their convenient fabrication and measurements. For array‐based sensors, multichannel arrays are more advantageous and used frequently in many electronic sensors. But most reported optically array based sensors are constructed on a single channel array. This difficulty is mainly instigated from the overlap in optical responses. In this report we have used nano‐graphene oxide (nGO) and suitable fluorophores as sensor elements to construct a multichannel sensor array for the detection of protein analytes. By using the optimized multichannel array we are able to detect different proteins and mixtures of proteins with 100 % classification accuracy at sub‐nanomolar concentration. This modified method expedites the sensing analysis as well as minimizes the use of both analyte and sensor elements in array‐based protein sensing. We have also used this system for the single channel array‐based sensing to compare the sensitivity and the efficacy of these two systems for other applications. This work demonstrated an intrinsic trade‐off associated with these two methods which may be necessary to balance for array‐based analyte detections.     

利用近红外及中红外融合技术对小麦产地和烘干程度的同时鉴别

小麦是制作馒头的主要原料之一,小麦中水、蛋白质、淀粉会因产地以及烘干程度的差异而不同,进而影响到加工成馒头的品质。所以实现对小麦产地和烘干程度的快速鉴别就显得尤为重要。感官评定是鉴别小麦产地和烘干程度常用的方法,对比感官评定,光谱分析可以识别样品中的分子结构等信息。基于此,尝试利用近红外和中红外光谱融合技术实现对不同产地和不同烘干程度的小麦同时鉴别。首先选取了两个不同产地的小麦,再利用微波干燥法对两个不同产地的小麦做烘干预处理,使烘干的小麦水含量为12%±0.5%,原麦水含量为18%±0.5%。分别标记为原麦A,烘干A,原麦B,烘干B,再将小麦研磨成粉末,过100目筛网筛选后,置于自封袋中备用。随后分别采集四种小麦样品的近红外和中红外光谱信息,在Matlab 7.10的环境下使用标准正态变量变换(standard normal variable transformation, SNVT)对采集到的原始光谱数据进行预处理,利用主成分分析对预处理后的数据进行降维处理,再结合线性判别分析（linear discriminant analysis,LDA）和支持向量机（support vector machine, SVM）分别建立小麦近红外、中红外光谱数据识别模型。另外利用联合区间偏最小二乘法（synergy interval partial least square, SiPLS）筛选出利用标准正态变量变换(SNVT)预处理后的小麦近红外和中红外光谱数据特征光谱区间,将筛选出的近红外和中红外光谱数据特征光谱区间融合后再结合线性判别分析（LDA）和支持向量机（SVM）建立小麦融合光谱信息的识别模型。然后比较同种光谱数据下利用线性判别分析（LDA）和支持向量机（SVM）建立的小麦识别模型识别率、比较同种建模方法下近红外和中红外光谱数据建立小麦识别模型识别率、比较同种建模方法下光谱数据融合和单一光谱数据建立小麦识别模型识别率。结果表明,同种光谱分析方法,利用SVM建立的四种小麦识别模型识别率高于利用LDA建立的小麦识别模型识别率。同种建模方法,近红外光谱数据建立的小麦识别模型识别率优于中红外光谱数据建立的小麦识别模型识别率。而在同种建模方法下,利用SiPLS筛选出近红外和中红外光谱数据的特征光谱区间数据融合后建立小麦识别模型识别率最高,光谱数据融合后结合LDA建立的小麦识别模型校正集识别率为98.75%,预测集识别率为97.50%;而将此选择的变量结合SVM建立的小麦识别模型的校正集和预测集识别率都达到100.0%。对比利用单一光谱数据建立的小麦识别模型识别率,光谱数据融合之后建立的小麦识别模型识别率得到显著提高,该研究从纵向和横向上全面地比较了光谱数据建立的小麦模型识别率,结果可为更准确地运用光谱融合技术建立小麦产地以及烘干程度识别模型提供参考。     

Multivariate analysis approach to the plasma protein profile of patients with advanced colorectal cancer

The aim of the present study was to identify the pattern of plasma protein species of interest as markers of colorectal cancer (CRC). Using matrix-assisted laser desorption/ionization-mass spectrometry (MALDI-MS), the plasma protein profile was determined in nine stage IV CRC patients (study group) and nine clean-colon healthy subjects (control group). Multivariate analysis methods were employed to identify distinctive disease patterns at protein spectrum. In the study and control groups, cluster analysis (CA) on the complete MALDI-MS spectra plasma protein profile showed a distinction between CRC patients and healthy subjects, thus allowing the identification of the most discriminating ionic species. Principal component analysis (PCA) and linear discriminant analysis (LDA) yielded similar grouping results. LDA with leave-one-out cross validation achieved a correct classification rate of 89% in both the patients and the healthy subjects.     

First-principles study of SnS electronic properties using LDA,PBE and HSE06 functionals

Recently, tin sulphide (SnS) has emerged as a promising alternative to conventional CIGS and CZTC for use in solar cells, possessing such properties as non-toxicity, low cost and production stability. SnS has a high theoretically predicted efficiency above 20%, but the experimentally achieved efficiency so far is as low as 4.36%. The reason for the low achieved efficiency is unclear. One of the powerful tools to get deeper insights about the nature of the problem is first-principles calculation approaches. That is why SnS has become an attractive subject for first-principles calculations recently. Previously calculated data, however, show a widespread of such fundamental value as the bandgap varying from 0.26 to 1.26 eV. In order to understand a reason for that, in this work, we concentrate on a systematic study of calculation parameters effects on the resulting electronic structure, with the particular attention paid to the influence of the exchange-correlation functional chosen for calculations. Several exchange-correlation functionals (LDA, PBE and HSE06) were considered. The systematic analysis has shown that the bandgap variation can result from a tensile/compressive hydrostatic pressure introduced by non-equilibrium lattice parameters used for the calculations. The study of the applicability of three functionals has shown that HSE06 gives the best match to both experimentally obtained bandgap and the XPS valence band spectra. LDA underestimates the bandgap but qualitatively reproduces experimentally measured valence DOS similar to that of HSE06 in contrast to PBE. PBE underestimates the bandgap and does not match to the measured XPS spectra.     

基于卡尔曼滤波算法的城市交叉路口车辆检测及分类

为了向相关部门提供更多的过往车辆信息以满足道路交通的需求,设计了一个基于卡尔曼滤波算法的城市交叉路口车辆检测及分类系统,用于对过往的车辆进行检测、计数和分类.首先采用背景差分法和卡尔曼滤波算法对在检测区的车辆进行检测和跟踪;然后使用经过检测、处理的被测车辆图像触发距其最近的相机进行图形分割;最后,通过LDA分类器对分段车辆的几何形状及外观特征进行正确地分类.所提系统的有效性在摄取的3400帧视频序列上得到了验证,实验结果表明,系统的检测率可达97.44％,正确分类率可达88.0％,与先进的方法相比,取得了更好的检测性能.     

图像特征融合在人脸识别中的应用研究

主成分分析(PCA)和线性判别分析(LDA)是基于全局结构的特征提取方法,局部保持投影(LPP)和正交拉普拉斯脸(OLF)是基于局部结构的特征提取方法,全局结构特征的弱点是忽略了局部结构特征,局部结构特征的弱点是忽略了整体结构.基于此提出了一种全局与局部结构图像特征融合(GLSF)的提取方法,将PCA和LDA的提取结果融合到LPP中,既描述了全局结构,又考虑了局部结构.在ORL及Yale上的实验结果表明,GLSF方法比PCA,LDA,LPP,OLF等方法具有更高的识别率.     

四角切圆燃烧锅炉炉膛内流体流动的实验研究

本文采用一维ＬＤＡ技术通过旋转方法对四角切圆燃烧煤粉锅炉炉膛内的冷态空气动力场进行了实验研究,得出了该流动定量化的流动特征：中心较明显的向下流动;下部燃烧器区外旋气流沿壁面向下流动;在屏式过热器下面仍然存在较强的旋转流动。本文所提供的数据可用于数值计算结果的检验。     

Intermolecular interaction potentials of methane－argon complex calculated using LDA approaches

The intermolecular interaction potential for methane－argon complex is calculated by local density approximation (LDA) approaches. The calculated potential has a minimum when the intermolecular distance of methane－argon complex is 6.75 a.u.; the corresponding depth of the potential is 0.0163eV which has good agreement with experimental data. We also have made a nonlinear fitting of our results for the Lennard-Jones (12-6) potential function and obtain that V(R)=143794365.332/R^{12}-3032.093 / R^6 (R in a.u. and V(R) in eV).     

融合奇异值分解和线性鉴别分析的人脸识别算法

本文提出了奇异值分解（SVD）和线性鉴别分析（LDA）相结合的人脸识别算法。理论上，当两种数据或分类器具有一定的独立性或互补性时，数据融合或分类器融合才能改善识别率。SVD和LDA之间有着明显的互补之处，LDA在fisher准则下能最大限度地把不同的类别区分开来，但作为一种子空间方法，LDA敏感于位移、旋转等几何变换。而作为一种代数特征提取方法的SVD则具有位移、旋转不变性等优点。因此，将这两种方法相结合就有可能提高分类性能（好于单独的SVD方法和单独的LDA方法）。在ORL数据库上的实验表明，SVD和LDA相融合的识别方法的确提高了人脸识别率。