近红外定性分析模型的稳健性与适应性分析

研究了近红外定性分析中模型的稳健性与适应性问题。讨论了单独建模延长建模周期对模型性能的影响,结果显示这种方法可以显著提高模型的稳健性,同时对模型的台间适应性也有所提升,但提升作用有限;研究了不同仪器联合建模的方法对模型性能的影响,该方法不仅可以显著提高模型的适应性,而且可以有效提高模型的稳健性。对比单独建模,可以缩短建模时间,减少模型建立的工作量;延长模型的适用期限,提高建模效率。结果表明,对模型适应性的测试,单独建模其正确识别率较低,不能满足应用的要求,而多台仪器联合建模可以达到90%以上,提升效果明显;对模型稳健性的测试,联合建模也能得到相比单独建模更好的模型识别效果,具有很好的应用价值。     

Cyclic multiaxial and shear finite deformation responses of OFHC Cu. Part II: An extension to the KHL model and simulations

In this part, the Khan–Huang–Liang (KHL) constitutive model was extended to account for kinematic hardening characteristic behavior of materials. The extended model is then generalized and used to simulate experimental response of oxygen free high conductivity (OFHC) copper under cyclic shear straining and biaxial tension–torsion (multiaxial ratchetting) experiments presented in Part I (Khan et al., 2007). In addition, a new modification for the non-linear kinematic hardening rule of Karim–Ohno (Abdel-Karim and Ohno, 2000) is proposed to simulate multiaxial ratchetting behaviors. Although, the kinematic hardening contributes the most to the response, it is shown that, the loading rate effect, and a coupled isotropic and kinematic hardening effect should also be considered while simulating the multiaxial ratchetting behavior of OFHC copper. Furthermore, the newly modified kinematic hardening rules is able to fairly well simulate the multiaxial ratchetting experiments under different loading conditions, irrespective of the value of applied axial tensile stress, shear strain amplitude, pre-cyclic hardening and/or loading sequence.     

Efficient Thickness of Solid Oxide Fuel Cell Composite Electrode

The efficient thickness of a composite electrode for solid oxide fuel cells was directly calcu-lated by developing a physical model taking into account of the charge transfer process, the oxygen ion and electron transportation, and the microstructure characteristics of the elec-trode. The efficient thickness, which is defined as the electrode thickness corresponding to the minimum electrode polarization resistance, is formulated as a function of charge trans-fer resistivity, effective resistivity to ion and electron transport, and three-phase boundary length per unit volume. The model prediction is compared with the experimental reports to check the validity. Simulation is performed to show the effect of microstructure, intrinsic material properties, and electrode reaction mechanism on the efficient thickness. The results suggest that when an electrode is fabricated, its thickness should be controlled regarding its composition, particle size of its components, the intrinsic ionic and electronic conductivities,and its reaction mechanisms as well as the expected operation temperatures. The sensitivity of electrode polarization resistance to its thickness is also discussed.     

Modeling adiabatic shear failure from energy considerations

The numerical simulation of dynamic structural failure by localized shear is quite complex in terms of constitutive models and choice of adequate failure criteria, along with a pronounced mesh-sensitivity. As a result, the existing numerical procedures are usually quite sophisticated, so that their application for design purposes is still limited. This study is based on the implementation of a simple energy-based criterion, which was developed on experimental considerations (Rittel et al., 2006), and uses a minimal number of adjustable parameters. According to this criterion, a material point starts to fail when the total strain energy density reaches a critical value. Thereafter, the strength of the element decreases gradually to zero to mimic the actual structural behavior. The criterion was embedded into commercial finite element software and tested by simulating numerically four typical high-rate experiments. The first is the dynamic torsion test of a tubular specimen. The second concerns the failure mode transition in mode II fracture of an edge crack in plain strain. The last two involve dynamic shear localization under high rate compression of a cylindrical and a shear compression specimen. A very good adequation was found both qualitatively and quantitatively. Qualitatively, in terms of failure path selection, and quantitatively, in terms of local strains, temperatures and critical impact velocity. The proposed approach is enticing from an engineering perspective aimed at predicting the onset and propagation of dynamic shear localization in actual structures.     

An improved transmission line laser model for multimode laser diodes incorporating thermal effects

An improved transmission line model to study the thermal effects in semiconductor laser diodes is reported in this paper. The temperature effects in the laser characteristics are obtained by incorporating temperature dependent gain and carrier density equations for the laser cavity. These primary factors are introduced in the regular transmission line laser model to estimate the static and dynamic characteristics of an 1.3μm InGaAsP double heterostructure laser diode. The results show good agreement with the experimental observation and solution of rate equations referred in the literature. The key feature of this model is that it provides the laser spectra at various temperatures. Based on the model, time dependent evolution of optical spectrum, temperature dependent optical output and frequency chirp are evaluated. Further the distribution of photon and electron density within the cavity is also determined.     

Ultrasound- and microwave-assisted convective drying of carrots – Process kinetics and product’s quality analysis

The aim of these studies was to investigate the influence of airborne ultrasound-assisted convective drying and microwave-assisted convective drying, as well as their combination, on process kinetics, total color change, water activity, content of carotenoids, polyphenols and antioxidant activity of carrots (Daucus carota L.). The global model of drying kinetics based on coupled ordinary differential equations was used to describe the moisture and material temperature profiles during drying. Application of ultrasound and microwave in convective drying reduced drying time in the range of 9–81%, but the shortest drying time was observed for simultaneous action of convection, ultrasound and microwave. The results of qualitative analysis showed a product improvement due to ultrasound as compared to convective drying and microwave-convective drying. The proposed mathematical model of drying kinetics successfully simulated real drying processes. The proposed mathematical model of drying kinetics successfully simulated real hybrid drying processes.     

高光谱技术结合CARS算法的库尔勒香梨可溶性固形物定量测定

由于高光谱数据量大、维数高,光谱噪声明显、散射严重等特征导致光谱建模时关键变量提取较为困难,同时,高光谱图像的获取会受非单色光、杂散光、温度等多种因素的影响,从而使高光谱数据与待测性质之间有一定非线性关系。为此,提出采用正自适应加权算法(CARS)对可见-近红外高光谱高维数据进行关键变量筛选,并与全光谱和经典变量提取方法SPA,MC-UVE,GA和GA-SPA方法进行比较。以200个库尔勒香梨为研究对象,采用SPXY方法将样本划分为校正集和预测集,校正集和预测集分别包含150个和50个样本。基于不同方法筛选的变量,分别建立线性PLS模型及非线性LS-SVM模型,r2,RMSEP和RPD用于模型性能的评估。综合比较发现,GA,GA-SPA和CARS变量筛选方法能够有效地筛选出原始高光谱数据中具有强信息且对外界影响因素不敏感的变量,适用于高光谱数据关键变量的提取,其中CARS变量筛选效果最佳,基于CARS获取的关键变量构建的非线性LS-SVM库尔勒香梨SSC含量预测模型获得了最优的预测结果,r2_pre,RMSEP和RPD分别为0.851 2,0.291 3和2.592 4。研究表明,CARS方法是一种有效的高光谱关键变量筛选方法,利用高光谱数据,非线性LS-SVM模型比线性PLS模型更适合于香梨品质的定量预测。     

油页岩含油率近红外光谱原位分析方法研究

为实现油页岩含油率的原位检测,采用便携式近红外光谱分析技术,针对吉林扶余油页岩基地2号钻井的66个岩芯样品开展了原位检测的分析建模方法研究。采用自制便携式近红外光谱仪器获得反射率、吸光度、K-M函数三种数据形式光谱数据,结合主成分-马氏距离(PCA-MD)剔除异常样品、无信息变量消除法(UVE)波长筛选及二者组合的四种建模数据优化方法,采用相同的数据预处理方法进行偏最小二乘(PLS)和反向传播神经网络(BPANN)两种方法的建模分析研究,确定最佳分析模型及方法。结果表明(1)不论是否采用四种不同的数据优化方法,两种建模方法所用建模数据库适合采用反射率或K-M函数的光谱数据形式;(2)两种建模方法,采用四种不同的数据优化方法,对相同数据库建模精度的影响不同：采用PLS建模方法、以PCA-MD和UVE+ PCA-MD两种方法进行数据优化、可以提高K-M函数光谱数据形式数据库的建模分析精度,采用BPANN建模方法、以UVE、PCA-MD 与UVE组合的 三种方法进行数据优化、对三种数据形式数据库的建模精度均有所提高;(3)除以反射率光谱数据并进行PCA-MD数据优化外,采用BPANN方法的建模精度好于PLS法。其中采用反射率光谱数据形式、只进行UVE数据优化外的BPANN建模精度最高,预测相关系数为0.92、标准偏差为0.69%。     

YSZ中温燃料电池的稳态模拟

依据同时考虑电化学及热平衡耦合的二维模拟软件 ,计算了薄膜钇稳氧化锆 (YSZ)中温燃料电池在不同工作条件下的稳态特性 .通过电流～电压关系参数进行自拟合实验 ,格点选取由平衡收敛性和计算效率而得 ,研究了不同连接体、气流流向设计等工作条件下的温度场 ,给出了不同工作温度下输出功率及电池效率与工作电压的关系 .对温度场的分析表明 :电池板内最高温度及最大温差以并流为最小 ,交叉流为最大 ,并流是最好的气流流向设计 .与以陶瓷材料作连接体相比 ,使用金属连接体能显著减小热应力和电池板内最高温度 ,受益最大的是交叉流 ,其最高温度及最大温差均小于陶瓷连接体的并流设计 .不同的气流流向对于输出功率及电池效率影响很小 ,对并流和金属连接体组合 ,给出了工程设计的燃料分布、电流密度、Nernst势及温度梯度在典型工作条件下的情形     

基于tri-PLS模型的油类混合物三维荧光光谱及其成分的研究

提出了一种光谱重叠的多种矿物油混合物组分含量测定的新方法。将偏最小二乘方法(PLS)推广至三维扩展(tri-PLS),不需要解决特征值问题。利用该方法对柴油、汽油和煤油混合物的三维荧光光谱进行研究,根据样本序列、激发波长、发射波长构造出三维数据矩阵,结合浓度矩阵应用tri-PLS法建立校正模型,对实验样本进行预测,实验结果表明tri-PLS方法的建模精度比常用的平行因子法优越。