CARS difference spectroscopy

Summary In this paper we present a vibrational-spectroscopy technique which combines the advantages of coherent anti-Stokes Raman spectroscopy (CARS) and linear Raman difference spectroscopy. The method which we call CARS difference spectroscopy can be applied for the study of small frequency shifts and/or bandwidth changes in the CARS spectra of liquid mixtures and solutions. First we develop the theory necessary for the interpretation of experimental data obtained from CARS difference measurements of mixtures of two Raman-active liquids and present some model calculations for benzene-toluene mixtures of different concentrations. Then the experimental arrangement used for CARS difference measurements as well as some examples of recorded spectra are described. We show that it is possible to observe the effects of dilution on CARS spectra with high accuracy by applying the discussed technique. Paper presented at the “XI European CARS Workshop”, Florence, Italy, 23–25 March, 1992.     

窗口竞争性自适应重加权采样策略的近红外特征变量选择方法

通过消除光谱中的冗余信息变量,挑选出代表样品性质的特征变量代替全谱建立定量模型,可以提高近红外分析结果的准确性。基于进化论中适者生存原理的竞争性自适应重加权采样（CARS）算法因具有计算速度快、筛选得到的特征波长少等优点,在近红外特征变量筛选方面得到了广泛的应用。然而该方法在计算过程中容易出现校正集和验证集结果不一致情况。这是因为算法过于强调校正集交叉验证结果,且并未考虑相邻变量之间的协同作用。为了建立更加稳健的变量筛选方法,通过结合“窗口”以及CARS算法的优势,提出了一种基于窗口竞争性自适应重加权采样（WCARS）策略的近红外特征变量筛选方法,并将其应用于复杂植物样品近红外光谱与其化学成分含量之间的建模分析。采用WCARS方法可以实现准确定量分析,且通过与竞争性自适应重加权采样（CARS）方法结果相比较,WCARS方法得到的校正集和预测集结果一致,在一定程度上减少了过拟合问题的出现。该策略能有效增强特征变量选择的稳健性,提高了定量模型的可信度,具有一定的应用价值。     

Theoretical modeling of single-laser-shot, chirped-probe-pulse femtosecond coherent anti-Stokes Raman scattering thermometry

Chirped-probe-pulse (CPP) femtosecond (fs) coherent anti-Stokes Raman scattering (CARS) spectroscopy for single-laser-shot temperature measurements in flames is discussed. In CPP fs CARS, a giant Raman coherence is created in the medium by impulsive pump-Stokes excitation, and the temperature-dependent temporal decay of this initial coherence is mapped into the frequency of the CARS signal using a CPP. The theory of the CPP fs CARS technique is presented. A computer code has been developed to calculate theoretical CPP fs CARS spectra. The input parameters for the calculation of the theoretical spectra include the temperature, probe time delay, ratio of the resonant and nonresonant susceptibilities, and parameters for characterizing the pump, Stokes and probe pulses. The parameters for characterizing the pump, Stokes and probe pulses are determined from the best fit of theoretical spectra to experimental spectra acquired from calibration flames at a known temperature. For spectra acquired in subsequent measurements, these laser parameters are fixed and temperature is determined as one of the fit parameters from the best fit of theoretical spectra to experimental spectra. For single-laser-shot CPP fs CARS temperature measurements performed in steady, near-adiabatic flames, the best-fit temperature distribution width is typically less than 1.5% of the mean temperature. The mean temperature is accurate to within approximately 3% with respect to the adiabatic flame temperature. The most significant limitation on temperature measurement accuracy is associated with the evaluation of the theoretical laser parameters. Significant improvements in the temperature measurement accuracy are expected once monitoring equipment capable of characterizing the spectrum and phase of each laser pulse is incorporated in the experiments.     

Dual-pump coherent anti-Stokes Raman scattering thermometry in a sooting turbulent pool fire

We present a dual-pump coherent anti-Stokes Raman scattering (CARS) instrument, which has been constructed for the probing of temperature fluctuations in turbulent pool fires of meter-scale. The measurements were performed at the Fire Laboratory for Accreditation of Models and Experiments (FLAME) facility at Sandia National Laboratories, which provides a canonical fire plume in quiescent wind conditions, with well-characterized boundary conditions and access for modern laser-diagnostic probes. The details of the dual-pump CARS experimental facility for the fire-science application are presented, and single-laser-shot CARS spectra containing information from in-fire N₂, O₂, H₂, and CO₂ are provided. Single-shot temperatures are obtained from spectral fitting of the Raman Q-branch signature of N₂, from which histograms that estimate the pdf of the enthalpy-averaged temperature fluctuations at the center of the fire plume are presented. Results from two different sooting fire experiments reveal excellent test-to-test repeatability of the fire plume provided by FLAME, as well as the CARS-measured temperatures. The accuracy and precision of the CARS temperatures is assessed from measurements in furnace-heated air, where the temperature can be accurately determined by a thermocouple. At temperatures in excess of 500 K, the furnace results show that the CARS measurements are accurate to within 2-3% and precise to within ±3-5% of the measured absolute temperature.     

Oxygen concentration and temperature measurements in N2–O2 mixtures using rotational coherent anti-Stokes Raman spectroscopy

The accuracy and precision of oxygen concentration and temperature measured by dual-broadband rotational Coherent Anti-Stokes Raman Spectroscopy (CARS) were investigated in nitrogen-oxygen mixtures at atmospheric pressure and temperatures between 290 and 1410 K. The relative standard deviation of temperatures evaluated from pure oxygen rotational CARS spectra was found to be around 5%, and the mean temperature was the same as for nitrogen CARS spectra, except for temperatures above 1000 K, where the temperature was 120 K below the correct value. The in situ calibrated oxygen concentrations were within 10% of the correct value, with a standard deviation of around 1.2% for the mixtures of 12 and 20% oxygen in nitrogen. For the lowest oxygen concentrations considered in this study (2 and 4%), the systematic errors in the evaluated concentrations were very large, and the standard deviation of repeated single-shot measurements was above 2%. However, employing weighting in the spectral fitting routine reduced the errors in the concentration and the single-shot standard deviation was lowered to 0.5%. Finally, it was shown that spectral interference (from oxygen) in a rotational CARS spectrum of nitrogen generally had little impact on the temperature evaluated from fitting the spectra to theoretical nitrogen spectra.     

NO气体的空间交叉CARS

采用具有高空间分辨率的空间交叉BOX CARS相位匹配方法,完成了对NO气休室温下、Q支(v=0→v=1)振转CARS谱的测量,研究了它与NO气体压强及激光线宽的关系.从CARS基本理论出发,考虑到激光线宽及不同线型,计算出理论曲线,与实验谱线进行了比较,洛伦兹线型的理论模拟和实验结果符合较好.     

Coherent anti‐Stokes Raman scattering in Raman lasers and Raman wavelength converters

In this paper recent research progress on the use of Coherent Anti‐Stokes Raman Scattering (CARS) in Raman lasers and Raman wavelength converters is reviewed. The latest insights in the physical nature and behavior of CARS are addressed, and the recent performance breakthroughs in CARS‐based Raman wavelength conversion are discussed. Based on the new findings regarding the behavior of CARS, a physical explanation for apparent inconsistencies in various experimental observations of Raman wavelength conversion is provided. To conclude it is shown that these recent insights also pave the way to the development of a novel CARS‐based mechanism for reducing the heat dissipation in Raman lasers.     

Hydrogen CARS thermometry in H2–N2 mixtures at high pressure and medium temperatures: influence of linewidths models

In order to improve the accuracy of H₂ CARS thermometry, H₂ Q-branch CARS spectra have been recorded for various H₂-N₂ mixtures in a high-pressure cell at different pressures and temperatures (up to 40 bar and 875 K). Due to the low spectral resolution of broadband CARS experiments, the relevant spectral lineshape factor is the linewidth ratio &(Q(3))/&(Q(1)), since Q(1) and Q(3) are the most intense lines of the Q-branch spectrum in this temperature range. For the first time, the speed-inhomogeneous effects are accounted for in the simulation of the CARS profiles. The evaluated temperatures are in good agreement with reference values obtained by thermocouples. The specific role on the accuracy of H₂ CARS thermometry of the speed inhomogeneity is carefully analyzed, in connection with the influence of the nitrogen concentration.     

BOXCARS测量燃烧场的温度

 介绍了用BOXCARS技术测量CH₄/air火焰的温度。给出了甲烷气体不同流量下和火焰不同高度处的温度测量结果，分析了温度随甲烷气体流量变化和火焰高度变化的变化趋势，根据实际的实验参数计算了相应的空间分辨率。实验结果表明：利用BOXCARS技术可进行高精度、高空间分辨的温度测量。     

BOXCARS测量燃烧场的温度

介绍了用BOXCARS技术测量CH4/air火焰的温度。给出了甲烷气体不同流量下和火焰不同高度处的温度测量结果,分析了温度随甲烷气体流量变化和火焰高度变化的变化趋势,根据实际的实验参数计算了相应的空间分辨率。实验结果表明：利用BOXCARS技术可进行高精度、高空间分辨的温度测量。     

CARS thermometry revisited in lightof the intramolecular perturbation

The rigid rotor approximation (RRA) is commonly assumed in the Raman cross section used in thermometric analysis based on coherent anti‐Stokes Raman scattering (CARS). In this paper, we discuss instead the role of the coupling between molecular vibrations and rotations in view of the alterations found in the amplitude of CARS signals of basic molecules and, in the end, we demonstrate that the deviation of a few percent from the RRA results in corrections to the measured temperature that are comparable to the thermometric accuracy of very well‐known Q‐branch CARS measurements on nitrogen, which is unanimously regarded as the fundamental molecule in CARS thermometry. Copyright © 2010 John Wiley & Sons, Ltd.     

基于CARS技术的超燃冲压发动机点火过程温度测量

相干anti-Stokes Raman散射（coherent anti-Stokes Raman scattering,CARS）技术作为一种非接触测量手段,已广泛应用于多种发动机模型燃烧室温度测量及地面试验.然而,目前的工作主要集中在稳态燃烧场温度的测量,缺乏用高分辨率的单脉冲来测量瞬变的燃烧火焰温度及组分浓度的研究.基于CARS理论,结合多参数拟合算法,开发了基于MATLAB的CARS光谱计算和拟合程序CARSCF;利用McKenna平面火焰炉在不同工况下进行了温度测量,并与DLR测量结果进行对比,结果显示开发的CARSCF具有较高的测量重复性和准确性;最后将CARS技术应用于测量超燃冲压发动机点火过程中的温度测量,获取了点火过程中的温度.结果显示,在来流Mach数为3的条件下,H₂/air点火过程中温度呈现急剧上升然后缓慢下降,而CARS信号则呈现急剧上升然后急剧下降随后又缓慢上升的趋势,并且在点火过程中最高温度为1 511 K.      

Highly sensitive single-beam heterodyne coherent anti-Stokes Raman scattering

Single-beam coherent anti-Stokes Raman-scattering (CARS) microspectroscopy achieves a complete CARS scheme with a femtosecond laser. Here, we introduce heterodyne detection in a simple experimental extension: the optical fields driving the CARS process and the local oscillator used for heterodyning are derived from a single beam of ultrashort laser pulses by pulse shaping. The heterodyne signal is amplified by more than 3 orders of magnitude and is linearly dependent on the concentration of Raman scatterers. This dramatically increases the sensitivity of chemically selective detection at microscopic resolution while maintaining the simplicity of the single-beam setup.     

高光谱漫反射无损检测鸡蛋受精状态

种鸡蛋孵化期间受精状态的检测需要消耗大量人力、物力,并且孵化期间的种鸡蛋不能保证均为健康蛋,需要能够在孵化早期将无精蛋和死精蛋快速准确挑选出来达到降低生产成本的目的。以白来航鸡蛋为研究对象,采用高光谱分选仪批量采集受精、未受精、死精三类鸡蛋共119枚在382～1 026 nm范围内的高光谱数据,其中受精蛋采集孵化3, 5, 7, 9, 11, 13和15 d的数据,并通过黑白校正方法对原始光谱图做校正处理,得到其漫反射率,经过实验对比以及根据实际生产需要,受精蛋选用孵化3和5 d的光谱数据作为建模数据。同时提出了一种将光谱数据转换为图像数据的方法,在最大化保证光谱原始数据的前提下达到了光谱向量数据可视化的效果,可以有效与深度学习图像识别算法相结合。采用连续投影算法(SPA)、竞争性自适应重加权算法(CARS)对光谱波段进行筛选,建立基于全波段、 CARS筛选的特征波长、 SPA筛选的特征波长与SVM、 RandomForest算法与AlexNet、 MobileNet网络的判别模型,其中AlexNet-5dFull Wave Bands准确率最高为93.22%。与通过不同特征波长算法筛...     

可调空间分辨USEDCARS技术研究

相干反斯托克斯拉曼散射（CARS）技术是一种非常重要的燃烧诊断技术,该技术具有非常强的抗干扰能力和非常高的测量精度。但空间分辨力不足会使CARS技术产生很强的空间平均效应,引起成CARS光谱畸变,进而造成CARS光谱分析困难,无法通过CARS光谱反演燃烧场参数。针对非稳腔空间增强探测CARS（USEDCARS）技术存在的空间分辨不足以及空间分辨力不易改变的特点,分析了影响USEDCARS技术测量空间分辨力的各种因素,采用一组轴棱锥对USEDCARS系统中的泵浦激光进行环状光束整形,并通过调节轴棱锥之间的距离获得了不同直径的环状光束,在此基础上,建立了空间分辨可调USEDCARS诊断系统。开展了空间分辨力分析实验,获得了CARS信号强度随空间位置的分布数据,以CARS信号总强度95%包含的空间区域代表CARS的纵向空间分辨力,以此计算得到了CARS系统空间分辨力为1.7～6.5 mm连续可调。其中,高分辨力情况,达到了现有BOXCARS技术的空间分辨力。利用所建立的空间分辨可调USEDCARS诊断系统测量了酒精/空气预混火焰温度参数,获得了不同空间分辨条件下的CARS光谱。空间分辨力为1.7 mm时,获得了高质量CARS光谱,通过光谱拟合给出了所测火焰的温度信息。分辨力分别为4.9和6.5 mm时获得了较强的CARS信号,但存在光谱畸变。结果显示,空间分辨力对CARS信号的强度和空间平均效应有很大地影响,提高测量的空间分辨力可以有效消除空间平均效应,获得准确的CARS光谱,增强光谱拟合精度,同时空间分辨可调的特性使该系统能够更好地适应不同实验条件下的诊断工作。     

Combined coherent anti-Stokes Raman spectroscopy and linear Raman spectroscopy for simultaneous temperature and multiple species measurements

The simultaneous application of pure rotational coherent anti-Stokes Raman spectroscopy (CARS) and vibrational linear Raman spectroscopy (LRS) for the measurement of temperature and species concentrations in combustion systems is demonstrated. In addition to the standard rotational CARS experimental setup, only one detection system (spectrometer and intensified CCD camera) for the collection of the LRS signals was applied. The emission of the broadband dye laser used for CARS was shifted to the deep red to avoid interferences with the LRS signals located in the visible region. First experimental results from a vaporizing propane spray using an engine injection system are shown.     

高光谱成像的猕猴桃糖度无损检测方法

猕猴桃糖度是重要的猕猴桃内部品质衡量指标。传统的糖度检测耗时且有损样品,有效无损检测猕猴桃糖度含量对于其品质分级、储藏销售具有重大意义。基于高光谱成像技术的常见果蔬品质无损检测方法多数是采用竞争性自适应重加权算法(CARS)、连续投影算法(SPA)、主成分分析(PCA)、迭代保留信息变量法(IRIV)等算法中的某个单一算法提取特征光谱变量,而这些算法单独使用易导致预测结果的稳定性不足。对此,开展了基于高光谱成像技术的猕猴桃糖度的无损检测方法研究。以四川省雅安市“红阳”猕猴桃为研究对象,依次对猕猴桃样本编号并采集其在400~1 000 nm波长范围内的高光谱图像,计算感兴趣区域的平均光谱作为样本的有效光谱信息;分别采用多元散射校正（MSC）、标准正态变量变换（SNV）、直接正交信号校正（DOSC）等3种光谱数据预处理方法分析对预测模型精度的影响,对比结果显示DOSC的预处理效果最好;对预处理后的光谱分别采用一次降维(CARS,SPA,IRIV)、一次组合降维(CARS+SPA,CARS+IRIV)算法和二次组合降维算法((CARS+SPA)-SPA,(CARS+IRIV)-SPA))等7种算法提取特征光谱变量,并分别构建了预测猕猴桃糖度的3种模型,即支持向量回归机（SVR）、最小二乘支持向量机（LSSVM）和极限学习机（ELM）模型;最后对比了基于不同特征提取方法的3种模型的预测精度。研究结果表明：ELM模型具有最好的预测性能,而SVR模型的预测性能最差;(CARS+IRIV)-SPA所选特征光谱变量输入LSSVM、ELM模型,其获得的预测结果均优于其他算法所选特征光谱变量输入对应模型所得的预测结果,证明了(CARS+IRIV)-SPA算法在提高猕猴桃糖度含量检测精度方面的有效性。对比不同方法的预测结果可知,(CARS+IRIV)-SPA-ELM对猕猴桃糖度的预测性能最优,其相关系数R_c=0.945 1,R_p=0.839 0,均方根误差RMSEC=0.450 3,RMSEP=0.598 3,预测相对分析误差RPD=2.535 1,该方法为猕猴桃糖度的检测无损化、精准化、智能化发展提供了可靠的理论依据和技术支撑。     

Hydrogen CARS thermometry in a high-pressure H2–air flame. Test of H2 temperature accuracy and influence of line width by comparison with N2 CARS as reference

This work describes a further step towards the determination of the temperature accuracy of H₂ Q-branch CARS (Coherent Anti-Stokes Raman Scattering) at high pressure with regard to the influence of the H₂ line widths. In laminar steady H₂/air flames in the pressure range 1–15 bar and at fuel-rich conditions with stoichiometries between two and four, quasi-simultaneous temperature measurements were performed with H₂ and N₂ CARS. The temperature values deduced from H₂ CARS are in good agreement with the reference temperature from N₂ CARS. The influence of different line-width contributions on the accuracy of H₂ Q-branch thermometry was investigated in detail. Received: 10 April 2001 / Revised version: 22 May 2001 / Published online: 18 July 2001     

CARS difference spectroscopy. II—Determination of small frequency shifts and band width changes in binary mixtures

The experimental verification of CARS difference spectroscopy (CARS-DS) is presented and its application to binary mixtures is shown. The experimental set-up allows the simultaneous measurement of CARS spectra of two different samples under identical conditions. It is shown how the method can be used for the accurate determination of small frequency shifts and band width changes in liquid mixtures. The experimental spectra obtained were analysed by their numerical simulation using a least-squares fitting routine. As examples, experimental results for chloroform–deuterochloroform isotopic mixtures and binary mixtures of benzene and toluene of different concentrations are presented. In both cases frequency shifts and changes in the band width were observed which exhibit an almost linear dependence on the mole fraction of the probe species.