主成分分析在温室气体时序红外光谱处理中的应用研究

傅里叶变换红外光谱技术可同时测量多种温室气体组分的浓度。仪器噪声和谱线重叠效应会对光谱数据的质量产生影响,从而影响各组分反演浓度的结果。针对上述问题,使用不同数量的主成分对时序测量光谱矩阵进行重建,并将重建光谱矩阵与原始光谱矩阵之间的欧式距离和余弦距离作为动态选择主成分数量的判据,重建时序测量光谱,从而提高时序光谱数据的质量。采用该方法分别对数值仿真光谱、标准气体测量光谱和外场实验测量光谱进行了处理。结果表明,叠加0.001 RMS噪声的数值仿真光谱经过光谱重建后,光谱的结构特征未明显改变,重建光谱与原始光谱之间的残差标准差为4.191×10^-4,有效降低了测量光谱中噪声的影响。采用该方法对标准气体的平均测量光谱进行了重建,并比较了重建光谱与平均光谱的反演浓度精度。1 min平均测量光谱反演各组分浓度的精度为CO₂:0.24μmol·mol^-1、 CH₄:5.24 nmol·mol^-1、 N₂O:2.92 nmol·mol^-1和CO:...     

多区间主成分交互验证法用于气体光谱定性分析

提出了一种多区间主成分交互验证法.该方法根据目标气体的光谱特征吸收峰,在全光谱范围内划分出多个特征波数区间,同时,对单组分气体构成的参考光谱矩阵在划分出的特征波数区间内做主成分分析,依据特征波数区间内累计贡献率大于99%的主分数和单组分气体在特征区内光谱最大值的排序,识别出目标气体可能含有的组分,再提取这些可能组分的特...     

近红外漫反射光谱的主成分分析

本文研究了主成分分析的应用，说明了在主成分分析过程中所产生的新变量如何提供新的光谱信息，该信息能改善对原光谱的解释。     

红外光谱结合主成分分析鉴别不同产地黄柏

利用傅里叶变换红外光谱仪收集6个不同产地黄柏的红外光谱,原始光谱经过多点基线校正、五点平滑预处理后,统计各个产地黄柏的平均光谱,结果显示,6个产地黄柏的平均红外光谱整体上相似;在指纹特征区域1 800～500cm-1,选取经过归一化的红外光谱,建立主成分分析鉴别模型.建模结果表明,三个主成分能代表98%的指纹区变量信息,样本在主成分空间中聚集成为不同的类别,基本实现不同产地黄柏的鉴别,此外,在一定程度上,样本分布疏密度反映样本的亲缘关系.提取模型的载荷因子,分析显示,6个产地黄柏的差异主要在蛋白质、糖类、脂类、生物碱类、黄柏甾醇类、黄柏内酯、黄柏酮、黄柏酮酸八种物质成分上体现.因此,红外光谱结合主成分分析法可以快速、无损地鉴别不同产地黄柏,并且能反映不同产地黄柏物质成分含量的差异.     

基于主成分分析和聚类分析的不同产地绒柄牛肝菌红外光谱鉴别研究

采用傅里叶变换红外光谱结合主成分分析和聚类分析建立快速鉴别不同产地绒柄牛肝菌的方法。采集15个产地绒柄牛肝菌样品的红外光谱信息,用多元散射校正(multiplicative signal correction,MSC)、二阶求导(Second derivative,SD)、Norris平滑的组合方法对原始光谱进行优化处理,MSC+SD+ND(15,5)预处理后的光谱数据进行主成分分析和聚类分析,并通过主成分载荷图分析不同产地绒柄牛肝菌样品差异的原因。结果显示,该方法的重现性,精密度及稳定性的RSD值分别为0.17%,0.08%,0.27%,表明方法稳定、可靠。主成分分析的前3个主成分累积贡献率达到87.24%,能表达红外光谱的主要信息,主成分得分散点图中同一产地样品成簇聚集,不同产地样品分布于相对独立的空间,能有效区分不同产地样品。主成分载荷图显示,随主成分贡献率降低,主成分所捕获的样品信息减少,其中PC1在3 571,2 958,1 625,1 456,1 405,1 340,1 191,1 143,1 084,935,840,727cm~(-1)波数捕获大量样品信息,归属为糖类、蛋白质、氨基酸、脂肪、纤维素等化学物质的吸收峰,表明这些化学物质含量的差异是区分不同产地绒柄牛肝菌样品的主要依据。基于离差平方和法(Ward method)及欧氏距离(Euclidean distance)进行聚类分析,能直观显示不同产地样品的分类情况及样品之间的相关性,15个产地样品基本能够按照产地来源正确聚类,正确率为93.33%。傅里叶变换红外光谱结合主成分分析和聚类分析,可以有效鉴别绒柄牛肝菌产地来源,并且能够分析不同产地样品具有差异的原因,为野生食用菌的鉴别分类和应用研究提供可靠依据。     

地面长波红外高光谱成像气体探测研究

报道了地面长波红外遥测的新进展 ,具体阐述了窗扫时空调制傅里叶光谱成像技术的实现过程.演示装置基于角锥反射镜M ichelson干涉具 ,构成了空间调制干涉 ;采用了制冷型长波红外焦平面探测器组件 ,通过对数据立方体的采集、重组、基线校正、切趾、相位校正和傅里叶变换等处理 ,实现了长波红外波段高光谱成像.自研的CHIPED-1长波红外高光谱成像原理实验装置的探测灵敏度指标噪声等效辐射通量密度NESR在单次采样时达到了5.6 × 10-8 W · (cm-1 · sr · cm2 )-1 ,与商品化时间调制干涉高光谱成像仪相当 ;反映了技术的先进性 ,并留有较大的改进空间.通过测试聚丙烯薄膜的透过率曲线 ,CHIPED-1红外高光谱成像原理实验装置的光谱响应范围达到了11. 5 μm.文章还以室外高楼和乙醚气体的探测实验为例 ,研究了二维分布化学气体VOC的高光谱成像探测方法.在复杂背景和低试验浓度情况下 ,从同一波数的红外光谱切片上 ,观察不出乙醚蒸气的存在 ,但是进行了差谱处理后 ,可以清楚看到乙醚蒸气的空间分布.高光谱方法应用在有机蒸气VOC的红外探测领域 ,相对于宽波段热成像方法 ,具有灵敏度高、抗干扰能力强和识别种类多等诸多优势.     

基于主成分分析的复杂光谱定量分析方法的研究

针对超短脉冲激光与气体相互作用产生的非线性荧光光谱的特点,提出了光谱分析的预处理、特征提取以及定量分析的有效方法.采用小波算法对光谱数据进行压缩和降噪处理,使光谱数据由3979点压缩降噪至664点.对处理后的光谱特征峰团的强度进行主成分分析,结果表明2个主成分即可包含98%的光谱信息.采用第一主成分对不同浓度下的谱线强度进行拟合与误差计算,有效地提取了与气体浓度有关的特征参量.实现了气体的定量分析.同已有的定量分析方法相比,定量分析的精度有较大提高,特别在所关心的低浓度情况下,定量分析的精度改善更为明显.3种气体的拟合与计算显示同以往方法相比误差由0.2694降为0.02.     

黄芩采收季节的红外光谱三级鉴别与主成分分析

中药的采收季节与药物质量有着密切的关联,研究不同季节采收药材的差异并对其进行准确识别,对中药药理研究与质量控制意义重大。采用红外光谱三级鉴别法,对春季和秋季采收的黄芩进行研究。结果表明,春季与秋季采收的黄芩红外光谱吸收峰位置较相似,但是酯类化合物中羰基CO的1 740 cm-1特征吸收峰,黄酮类成分1 614 cm-1附近特征吸收峰及糖类成分1 071 cm-1附近吸收峰的相对峰强度差别显著。通过二者相对峰强度的不同发现春季采收的黄芩药材中黄酮类和酯类成分高于秋季采收的药材。比较二者在1 300～400 cm-1波数范围内的二阶导数红外光谱,得出其所含糖类物质有所不同。二者的差异在二维相关红外光谱上更加直观显著。对16个批次的春季与秋季采收黄芩药材红外光谱进行了主成分分析,结果表明二者可以得到准确区分。综上,红外光谱三级鉴别法与主成分分析相结合,可以准确鉴别春季和秋季采收的黄芩,并分析其化学成分的差异。     

基于主成分分析和神经网络的近红外光谱苹果品种鉴别方法研究

提出了一种用近红外光谱技术快速鉴别苹果品种的新方法,首先用主成分分析法对苹果品种进行聚类分析并获取苹果的近红外指纹图谱,再结合人工神经网络技术进行品种鉴别。主成分分析表明,主成分1和主成分2的累积可信度已达98％,以主成分1和2对所有建模样本的得分值做出的得分图,对不同种类苹果具有很好的聚类作用。利用主成分分析得到的载荷图可以得到对于苹果品种敏感的特征波段,用特征波段图谱作为神经网络的输入建立三层BP人工神经网络模型。每个品种各25个苹果共75个用来建立神经网络模型,余下的共15个用于预测。对未知的15个样本进行预测,品种识别准确率达到100%。说明文章提出的方法具有很好的分类和鉴别作用,为苹果的品种鉴别提供了一种新方法。     

基于多尺度加权主成分分析的SF6红外光谱分析

利用红外光谱法分析SF6气体及其衍生物是判断气体绝缘组合电器(GIS)运行状态和故障的一种重要手段。传统的诊断方法过程繁琐、效率低下,而且受主观因素的影响较大。本文指出可以采用机器学习的方法实现GIS设备的故障诊断,并提出了多尺度加权主成分分析的特征提取方法。多尺度加权主成分分析结合了主成分分析和多尺度分解的特点,保证了尺度特征信息的最大化,并且修正了特征向量在数据分类时的权重。通过对广西电力研究院提供的SF6及其衍生物的红外光谱进行分析,证明了多尺度加权主成分分析算法对训练样本的分类效果要比标准的主成分分析算法好3～4倍。     

Principal component analysis algorithm in video compressed sensing

An algorithm of principal component analysis in video compressed sensing is proposed in the paper. Aiming at the compressed sensing problems of video sequences, the inter-frame correlation among the images is analyzed and the transform coefficients with lower value are removed according to the energy concentration characteristics of principal component analysis. Therefore, the sparse realization of video signals in the form of principal component analysis is accomplished and the possibility of the transformation being used in compressed sensing algorithm is verified. Finally, simulation results show that, with the comparison of the traditional algorithm based on wavelet transform, the proposed algorithm can not only improve the reconstructed quality and the visual effects of the video sequence, but also save the sampling resources. Moreover, it is more suitable for stream transmission of multimedia.     

On the use of principal component analysis to speed up radiative transfer calculations

Radiative transfer is computationally expensive. However, it is essential to many applications, in particular remote sensing retrievals. Principal component analysis of the optical depth and single scattering albedo profiles has been proposed as a possible method to help ease the computational burden. Here we show how the technique could be applied to a practical problem of CO₂ retrievals from high spectral resolution measurements of reflected sunlight in three near infrared bands. We obtain a speed improvement of more than 50 fold (compared to monochromatic computations), while reproducing the radiances to better than 0.1% accuracy.     

Improvement of sensitivity using principal component analysis by dual focused ion beam time-of-flight secondary ion mass spectrometry

It is important to develop and to practically use the method to analyze a micro-nanometer order area. Especially, three-dimensional microanalysis for minute structure that consists of the organic compounds and the polymer is difficult. We developed a novel three-dimensional microanalysis method by means of focused ion beam (FIB) for section processing and ToF-SIMS for mapping method. For the purpose of realization of three-dimensional microanalysis and a chemical and structural analysis of the organic matter, the sensitivity improvement of ToF-SIMS in the three-dimensional analysis device and the method of the spectral analysis are examined. To improve the sensitivity of ToF-SIMS, the sample stage was modified to arrange perpendicularly with the ToF optical axis, and the distortion of electric field was corrected. And, by analyzing the fragment ions by using the principal component analysis (PCA) to raise the efficiency of the spectrum analysis, spatial resolution has improved. As a result, the resolution of the device improved to sub micrometer order, and advanced to the achievement of the three-dimensional microanalysis.     

Measurement of transparence topography based on principal component analysis

An algorithm based on the principal component analysis (PCA) is proposed to measure transparent surface topography. First, many frames of random phase shifted fringe patterns are captured. Second, the first and second principal components are extracted from the phase shifted fringe patterns by PCA. Then, the phase is calculated from principal components. Finally, the correct global phase sign is determined by the clustering method considering local fringe phase monotonicity. The experimental results show that the proposed method can directly extract the three-dimensional (3D) shape, which does not need to determine the amount of phase shift, without precise phase shifter. It reduces the experimental hardware requirements and improves its adaptability in profilometry.     

Multi-harmonic active structural acoustic control of a helicopter main transmission noise using the principal component analysis

This paper evaluates the feasibility and effectiveness, under controlled conditions, of active structural acoustic control of a helicopter main transmission in order to attenuate the vibration of the receiving roof panel and its sound radiation into the cabin. The vibroacoustic analysis of a typical helicopter drive train is conducted to extract the dominant tones generated by the various transmission stages. A finite element model of a Bell 407 transmission and simplified roof structure is developed in order to investigate various active control arrangements using piezoceramic actuators to control wave propagation in the gearbox supporting struts or receiving panel. The principal component analysis is used to extract the most significant control paths and reduce the control effort. A multiple frequency principal component least mean square (PC-LMS) algorithm is implemented on a laboratory setup and is used to successfully reduce up to seven gearbox tones simultaneously.     

一种突出构造信息的多光谱遥感变换方法

为突出多光谱遥感图像中的构造信息,提出了用主成分分析、最大噪声变换和原始数据联合进行信息增强的方法。首先对原始数据进行主成分分析和最大噪声变换,然后对变换结果与原始数据进行归一化处理的基础上进行波段选择,最后进行假彩色合成并输出结果。利用新疆哈图地区的ETM数据进行构造增强的效果表明,该方法不仅丰富了遥感图像的颜色、拉大了不同类地物间的色彩差异,为提取不同类地物信息提供了一条有效的途径,而且还有力增强了宏观构造、凸现了小构造和肠状构造,为展现研究区构造格局、建立遥感找矿模式提供了一套行之有效的方法。     

利用激光诱导荧光来探测远程宽带太赫兹波

远程太赫兹光谱分析在国土安全和环境监测方面的应用有着巨大的潜力.然而由于大气中的水气在太赫兹频率范围内的强吸收,远程宽带太赫兹探测一直以来都是一个挑战.文章介绍了一种通过相干控制非对称电离气体荧光来实现远程宽带探测太赫兹波的全光学技术.由于荧光在大气中的高透过率以及全方位辐射特性,该技术可用于太赫兹波的远程探测,具有最小水气吸收以及不受光信号收集方向限制两个优点.实验上已经获得了距离为10m量级的太赫兹波相干探测.     

矿区土壤Cu含量高光谱反演建模

为探究高光谱遥感手段反演土壤Cu含量方法的可行性,以湖南省某矿区为例,利用ASD地物光谱仪和实验室电感耦合等离子发射光谱法测定83个土壤样品350～2 500 nm光谱信号和Cu含量。在光谱重采样、一阶/二阶微分、标准正态变换预处理对比分析基础上,分别采用主成分分析与相关分析对潜在建模变量进行初步筛选,运用逐步回归方法确定最终模型变量,建立土壤Cu含量反演模型,基于最优模型识别Cu含量光谱指示特征波段。结果表明,相对于传统主成分分析方法,标准正态变换后的光谱全要素主成分分析逐步回归建模方法因保留土壤样品弱光谱信号能有效提升土壤Cu含量估算能力,R2达0.86,模型对于预测样本的估计效果较好,建模样本和预测样本的残差分别为0.76和1.29,且通过F检验;360～400,922～1 009,1 833～1 890与2 200～2 500 nm波段对研究区土壤Cu含量有较好指示性。研究结果将丰富南方矿区土壤Cu含量估算典型案例,同时为发展基于高光谱遥感的土壤环境监测手段提供理论支撑。     

Multi-objective optimization of Nd:YAG laser cutting of nickel-based superalloy sheet using orthogonal array with principal component analysis

Laser cutting of nickel-based superalloy sheets, having wide applications in aircraft and rocket industries, is important from the quality of cut point of view. Keeping this in view, a hybrid approach of Taguchi method (TM) and principal component analysis (PCA) has been applied for multi-objective optimization (MOO) of pulsed Nd:YAG laser beam cutting (LBC) of nickel-based superalloy (SUPERNI 718) sheet to achieve better cut qualities within existing resources. The three-quality characteristics kerf width, kerf deviation (along the length of cut), and kerf taper have been considered for simultaneous optimization. The input parameters considered are assist gas pressure, pulse width, pulse frequency, and cutting speed. Initially, single-objective optimization has been performed using TM and then the signal-to-noise (S/N) ratios obtained from TM have been further used in PCA for multi-objective optimization. The results of MOO include the prediction of optimum input parameter level and their relative significance on multiple quality characteristics (MQC). The responses at predicted optimum parameter level are in good agreement with the results of confirmation experiments conducted for verification tests.