Assessment of the environmental and physiological processes determining the accumulation of organochlorine compounds in European mountain lake fish through multivariate analysis (PCA and PLS)

The presence of organochlorine compounds (OCs)-namely hexachlorobenzene (HCB), hexachlorocyclohexanes (HCH), polychlorobiphenyls (PCBs #28, 52, 101, 118, 138, 153 and 180), dichlorodiphenyltrichloroethane (DDT) and dichlorodiphenyldichloroethane (DDE)-was examined in various fish tissues (muscle and liver) sampled in 23 mountain lakes in Europe. The dependence of these organochlorine compounds on geographical parameters (altitude, longitude, latitude and temperature) and physiological parameters (lipid content, age, weight and size) was assessed. Principal components analysis (PCA) and partial least squares (PLS) models were used for the analyses. PCA results showed that organochlorine compound concentrations in fish tissues increased with increasing altitude and decreasing temperatures. This trend appeared to be more marked for the less volatile compounds. Some differences were found between the muscle and liver in the effects of the percentage of lipids on the accumulation of organochlorine compounds and the behaviour of HCB. Moreover, PCBs tended to accumulate more in liver rather than in muscle. PCA scores clearly differentiated samples according to the lake of origin. PLS models confirmed that temperature and altitude were the main factors influencing the accumulation of most organochlorine compounds in the lipids of the fish tissue.     

基于吸光度的波长筛选方法用于近红外光谱定量模型的优化

建立一种基于吸光度的波长筛选方法,以近红外光谱测定中成药制剂的多糖含量为例,对模型优化效果进行验证.考虑模型稳定性,在计算机平台上搭建一种新的样本集划分框架,基于吸光度筛选出最优波段为400~1882&2072~2364 nm,建立偏最小二乘(PLS)模型得到的SEPAve、RP,Ave分别为27.13 mg L-1、0.856,与全扫描谱区(400~2498 nm)的PLS模型预测效果做比较.结果表明,基于吸光度的波长筛选方法,可以优选出高信噪比波长,从而提高了近红外光谱定量模型的性能.     

基于近红外高光谱成像快速无损检测注胶肉研究

采用近红外高光谱成像技术结合化学计量学方法建立注胶肉的快速无损检测模型。首先通过近红外高光谱成像系统获取含有不同浓度梯度卡拉胶的猪里脊肉高光谱图像,然后提取图像中的光谱数据,使用偏最小二乘法(Partial least square,PLS)探究光谱信息与不同掺假比例卡拉胶之间的定量关系。结果表明全波段光谱(900~1700 nm)所构建的PLS校正集模型均方根误差(Root mean square error,RMSE)为1.74%,预测模型RMSE为3.16%。表明基于全波段所建立的PLS模型具有较优的预测性能。利用连续投影算法(Successive projection algorithm,SPA)筛选获得11个特征波长,并优化全波长PLS模型,将预测集样品带入,以验证模型的预测效果,结果表明SPA算法结合PLS建模方法所建立的模型预测效果更优,预测集相关系数(R_P)为0.93,均方根误差(Root mean square error of prediction,RMSEP)为3.51%,预测偏差(Residual predictive deviation,RPD)为2.66。试验表明利用高光谱成像技术可实现对注胶猪肉的快速无损检测。     

葛根提取物中麦芽糊精的中红外快速定量分析

中药提取物因生产工艺、储存条件的要求,需在生产过程中添加一定量的麦芽糊精。目前国内外还没有标准方法测定中药提取物中麦芽糊精的含量。本研究旨在建立一种快速检测葛根提取物混合粉末中辅料麦芽糊精含量的中红外快速测定方法,为企业有效把握提取物质量提供技术支持。以不同麦芽糊精含量的葛根提取物混合粉末样品88份为实验材料,采集已知麦芽糊精含量样品的中红外光谱,对谱图进行基线校正和归一化预处理,通过偏最小二乘回归方法建立定量模型。结果表明:所建定量模型预测效果良好,预测模型稳定性强、预测精度高,预测模型的决定系数R2为0.9819,预测标准差为3.205。采用中红外光谱检测技术可以快速定量葛根提取物中麦芽糊精的含量,有效地监督提取物市场以次充好的混乱现象,为市场监督部门提供技术支持。     

SVR算法的生物混液左旋多巴含量紫外光谱建模

以酪氨酸和左旋多巴混合溶液中左旋多巴的紫外光谱数据为研究对象,首先用Kennard-Stone算法对样品集进行分割;然后使用ν-SVR和ε-SVR算法对核进行建模,以建立不同的核.紫外光谱定量分析模型中左旋多巴含量的功能;最后,采用粒子群算法对参数进行了优化,并与传统的PLS算法进行了比较.实验结果表明,由ν-SVR,...     

常压蒸馏产品质量软测量改进方法及应用

针对石油化工生产过程基于统计模型的产品质量软测量中普遍存在的软测量模型适用工作范围小、难以反映进料原料性质变化的问题,探讨和应用了机理分析和统计建模相结合的软测量方法。针对一个实际常减压蒸馏产品质量软测量和控制问题．通过机理分析、实际操作数据分析并结合操作人员经验,选择能够反映进料原料性质变化的过程变量作为统计模型的输入以克服进料原料性质变化的影响,将有些直接测量的输入变量按照机理关系进行计算得到的新的变量作为统计建模的输入变量,使其和产品质量之间具有更宽范围的近似线性关系,提高软测量模型的泛化能力：某厂常减压蒸馏装置的实际应用结果表明了该方法的有效性。     

工业流化床乙烯气相聚合中的在线测量与模型分析(II)

工业流化床乙烯气相聚合过程是一类典型的多变量复杂非线性系统。我们首先分析了若干主要过程变量(反应器温度等)对两个质量变量(树脂熔融指数和密度)的影响趋势，给出了两组变量间的非线性函数结构形式，然后运用多元统计投影原理和非线性函数的。Taylor逼近原理建立了流化床乙烯气相聚合过程变量和质量变量之间的主元多项式非线性偏最小二乘软测量(Nonlinear PLS)模型，并对理论结果进行了推证。基于实际工业运行装置的采样数据对Nonlinear PLS模型进行了求取和验证，利用所求模型进行了树脂熔融指数和密度预测并将预测结果和误差分析结果与线性PLS及机理模型进行了比较。最后运用所建立的模型进行流化床乙烯气相聚合过程操作条件的模型预测分析。全文分为两部分，这是第二部分。     

智能反射面辅助无人机的物理层安全传输方案

在无线通信系统中,信息发送端和接收端之间的通信链路易受到建筑物的遮挡和非法者的主动窃听。利用无人机（unmanned aerial vehicle,UAV） 灵活的移动性以及智能反射面（intelligent reflective surface,IRS） 在无线环境中智能可控特性,提出一种智能反射面辅助无人机的物理层安全传输方案,以最大化系统平均保密率。该方案首先采用块坐标下降法将最大化系统平均保密率问题分解为UAV-IRS飞行轨迹优化、IRS相移优化以及系统传输功率优化三个子问题。其次,通过引入松弛变量、相位对齐以及逐次凸逼近方法将每个子问题转换为凸优化问题。最后,再运用交替优化方法逐次求解三个凸优化问题。数值仿真结果表明,系统平均保密率与源节点发射功率、UAV-IRS飞行位置和高度、IRS反射元件数量等因素有关。并且与另外两个基准方案相比,所提的方案能有效地提高系统的平均保密率,并且具有较好的收敛性。     

Determination of the complex permittivity of cherry,pulped, green,and roasted coffee using a planar dielectric platform and a coaxial probe between 0.3 and 6 GHz

The complex permittivity for four processing stages of coffee: coffee cherries, pulped coffee, green coffee, and roasted coffee was obtained and divided into two groups. The wet group is constituted by cherry and pulped coffee, and the dry group by green and roasted coffee. The wet group presented a dielectric constant between 64 and 43, and a loss factor between 34 and 12. The dry group has values oscillating between 1.55 and 1.1 for the dielectric constant, and 0.037 and 0.005 for the loss factor. The permittivity difference between the wet and dry groups suggests the dependence of the coffee permittivity on the water content, along with the presence of other polar compounds. A decaying response for the dielectric constant was seen, with a proportion of 30:1 from the wet to the dry group. The loss factor presents a convex response for the wet group and a constant response for the dry group with a factor of 500 to a 100:1 from the wet to the dry group. A set of 18 samples at every stage was used to estimate the moisture content by partial least squares. The Meyer and Schilz coefficient was used for these models, that resulted in prediction errors of 2.8% and 3.8% when compared to the stage mean moisture content for the wet group. The dry group resulted in errors of 21% and 89.7%. These results indicate that the complex permittivity of coffee is capable of estimating physicochemical variables as the moisture content of the beans at distinct processing stages.