iPLS-SPA变量选择方法在螺旋藻粉无损检测中的应用

该文研究了基于可见-近红外光谱技术的螺旋藻粉类别无损检测方法。采用簇类独立软模式法（SIMCA）建立可见-近红外光谱模型。全波段光谱所建立的模型得到了93.33%的预测集正确率。文章提出了基于间隔偏最小二乘法（iPLS）和连续投影算法（SPA）的组合光谱变量选择方法进行有效波长的选择。该方法从全波段675个变量中选择了5个最优的有效波段,并且得到了96.67%的预测集正确率。和基于全波段光谱、可见光波段光谱和近红外波段光谱进行SPA运算相比,基于iPLS的SPA运算可以有效减少计算时间。研究表明可见-近红外光谱可以用于对螺旋藻粉类别进行无损检测,同时iPLS-SPA是一个有效的光谱变量选择方法。     

基于近红外光谱的绿茶滋味品质估测模型北大核心CSCD

绿茶中的茶多酚与氨基酸含量之比(即酚氨比)是评价绿茶滋味品质的量化指标。本文提出利用近红外光谱技术快速分析绿茶汤的酚氨比,并对光谱的特征变量进行筛选以提高模型的精度和稳定性。试验采用联合区间偏最小二乘法(siPLS)结合连续投影算法(SPA)筛选特征变量,建立酚氨比的估测模型,并与PLS、iPLS和siPLS方法建立的模型性能相比较。结果表明,应用siPLS结合SPA优选7个特征变量,主成分因子数为4时,所建模型性能最好,模型预测集相关系数(R p)为0.906,预测均方根误差(RMSEP)为0.258。对预测集30个样本的预测值与参考值进行t检验时,预测值与参考值无显著差异,说明模型准确可靠,可实现绿茶滋味品质的近红外光谱快速估测。     

基于蒙特卡罗特征投影法的小麦蛋白质近红外光谱测量变量选择

为了实现小麦蛋白质的无损检测,简化便携式小麦蛋白质检测设备的预测模型,提高模型预测精度,该文针对小麦采集波长范围为950～1690nm的近红外漫透反射光谱,结合蒙特卡罗采样(MCS,monte carlosampling)技术与特征投影图(LPG,latent projective graph)方法对波长变量进行选择。根据模型集群分析(MPA,model population analysis)思想,采用MCS技术建立样本子空间,利用主成分分析(PCA,principal componentanalysis)得到LPG,假定LPG中共线性光谱变量对建模作用相同,选出少数波长变量建立子预测模型,选出预测均方根误差(RMSEP,root-mean-square error of prediction)较小的子模型,统计分析其变量的出现频次,选取频次最高的波长变量作为影响变量(IVs,influential variables)。研究结果表明,利用IVs建模可以将RMSEP值由0.5245减小到0.2548,采用蒙特卡罗采样技术的特征投影图方法(MC-LPG,monte carlo-latent projective graph)进行变量选择,对于提高模型预测精度是可行的。     

土壤有机质可见光-近红外光谱预测样本优化选择

土壤有机质可见光-近红外光谱预测中建模样本的优化选择对提高有机质模型估算精度具有重要作用。本文以湟水流域土壤有机质为例,采用基于土壤单一属性信息考虑的建模样本选择方法:浓度梯度法、Kennard-Stone(KS)方法,以及基于土壤多种信息考虑的建模样本选择方法:Rank-KS(RKS)法、土壤类型结合浓度梯度法以及土壤类型结合KS法。通过偏最小二乘回归建模,探索可见光–近红外光谱预测青海湟水流域有机质的最优样本集。结果表明:不同级别样本数的最佳建模样本选择方法不同,整体表现为基于土壤多种信息挑选的建模样本集的模型精度相比土壤单一信息均较高,特别是KS方法结合土壤类型后的建模样本集模型精度明显提高且在样本数较少时更为明显。土壤类型可以优化建模样本选择方法提高模型预测精度。在保证固定验证样本模型预测精度的情况下,土壤类型参与建模样本的选择可以有效减少建模样本数,进而降低了建模成本。     

近红外光谱变量筛选提高西瓜糖度预测模型精度

水果的内部品质是水果分级、保鲜及存储的一项重要指标,利用近红外光谱技术对西瓜内部品质进行快速无损检测研究有着非常重要的意义。为了研究变量筛选方法对西瓜糖度预测模型精度的影响,该文以麒麟瓜为研究对象,利用近红外漫透射光谱技术对麒麟瓜可溶性固形物含量（SSC）进行检测,采用偏最小二乘回归（PLSR）,多元线性回归（MLR）和主成分回归（PCR）建立麒麟瓜可溶性固形物数学模型,并探讨等间隔平均光谱和等间隔抽取光谱变量筛选结合连续投影算法（SPA）对预测模型精度的影响。研究结果表明：光谱经等间隔抽取（间隔5,115个变量）经归一化预处理,结合SPA优选出6个波长建立的PLSR预测模型的相关系数（rpre）为0.828、校正均方根误差（RMSEC）为0.589、预测均方根误差（RMSEP）为0.611。该模型预测效果相对较优,建模时间短,提高了模型的预测能力和预测精度。该研究为西瓜内部品质的在线无损检测提供研究基础。     

近红外光谱联合CARS-PLS-LDA的山茶油检测

为了寻找快速判别山茶油掺假的检测方法,本研究利用近红外光谱技术对掺杂大豆没油山茶油进行掺假检测研究.试验在350～1 800 nm波段范围内采集样本的透射光谱,利用CARS方法筛选重要的波长变量,应用偏最小二乘-线性判别分析(PLS-LDA)建立山茶油掺假的判别模型,并与未经变量优选的判别模型进行比较.结果表明,近红外光谱技术联合CARS-PLS-LDA方法可以有效判别纯山茶油和掺假山茶油,校正集、预测集及独立样本组样本的判别正确率、灵敏度及特异性均为100％.CARS-PLS-LDA判别模型性能优于未经变量优选的判别模型,表明CARS方法可以有效筛选重要波长变量,能简化判别模型及提高判别模型的稳定性和判别精度.本研究可为山茶油掺假快速检测提供理论依据.     

基于数据融合策略的红茶发酵程度判别

发酵是红茶加工过程中关键的一道工序,对红茶的品质形成有着重要影响。该研究以大叶种英德红茶中的英红九号为研究对象,试验收集了204份不同发酵时间的红茶样品并使用便携式近红外光谱仪和工业相机获取红茶发酵中的信息,基于近红外光谱数据、图像数据和数据融合策略分别建立了红茶发酵程度判别模型。通过分析茶多酚和儿茶素类含量的变化,将红茶的发酵划分为3个阶段,即发酵不足、发酵适度和发酵过度。采用Savitzky-Golay光滑对原始光谱进行预处理,利用竞争自适应重加权采样（Competitive Adaptive Reweighted Sampling, CARS）、连续投影算法（Successive Projections Algorithm, SPA）和主成分分析（Principal Components Analysis, PCA）对近红外光谱变量进行降维处理;相应地,图像进行去阴影后提取了9个颜色特征变量,采用皮尔森（Pearson）相关分析和主成分分析进行特征变量提取。最后采用线性判别分析（Linear Discriminant Analysis, LDA）和支持向量机（Support Vector Machine, SVM）分别建立了基于近红外、图像和两者数据融合的分类模型。结果表明,在建模数据相同的条件下,非线性的支持向量机模型性能优于线性判别分析模型。单一传感器数据建模效果不佳,近红外光谱和图像判别模型的预测集最大准确率仅为83.82%和73.53%。低层次数据融合建模效果较单一传感器数据建模无明显提升,而中层次的数据融合建模效果比单一数据建模均有显著提高,其中SPA提取光谱变量结合Pearson提取图像变量建立的判别模型效果较佳,校正集和预测集准确率分别达到了97.06%和95.59%。研究表明,近红外光谱和视觉结合的中层次融合策略可以作为一种快速判别红茶发酵程度的方法,研究结果为红茶发酵程度构建等级模型与判别奠定了一定的理论基础,为红茶发酵的自动化检测提供了重要依据。     

傅立叶变换近红外光谱法检测白酒总酸和总酯

该文研究白酒总酸和总酯的快速检测技术，通过解析不同白酒样品的近红外光谱图，对光谱数据进行不同的处理，结果表明：用一阶导数预处理光谱，谱区选择6102～5446 cm^-1，利用人工神经网络与傅立叶变换近红外光谱相结合,采用内部交叉验证法建立模型，效果较好。其中，总酸模型的决定系数为96.73%，内部交叉验证均方根差为0.048 g/L；总酯模型的决定系数为99.58%，内部交叉验证均方根差为0.085 g/L；进一步对总酸和总酯的模型进行验证和评价，结果表明总酸模型验证集的相关系数为99.2%,预测标准偏差为0.074 g/L；总酯模型验证集的相关为99.7%,预测标准偏差为0.134 g/L，表明建立的模型可靠，预测效果好,能满足白酒生产中总酸和总酯的快速检测要求。     

基于光谱技术的大豆豆荚炭疽病早期鉴别方法

为更好地指导农户进行植物病害防治,提高大豆豆荚的商品性,减少损失,需要运用快速有效的方法来进行大豆豆荚炭疽病的早期检测。该文应用可见-近红外光谱技术结合连续投影算法(SPA)和最小二乘支持向量机(LS-SVM),实现了大豆豆荚炭疽病的早期快速无损检测。对194个大豆豆荚样本进行光谱扫描,通过不同预处理方法比较,建立了大豆豆荚炭疽病早期无损鉴别的最优偏最小二乘法(PLS)模型。同时应用主成分分析(PCA)和连续投影算法(SPA)分别了提取最佳主成分和有效波长,并将其作为LS-SVM的输入变量,建立了PCA-LS-SVM和SPA-LS-SVM模型,以样本鉴别的准确率作为模型评价指标。试验结果显示PCA-LS-SVM和SPA-LS-SVM模型都获得了比较满意的准确率,且SPA-LS-SVM模型的准确率最高为95.45%。研究表明,SPA能够有效地进行波长选择,进而使LS-SVM模型获得较高的鉴别率,说明应用可见-近红外光谱技术鉴别大豆豆荚炭疽病是可行的。这为进一步应用光谱技术进行大豆生长对逆境胁迫的反应提供了新的方法,为实现大豆病害的田间实时在线检测提供参考。     

基于变量选择的蚕茧茧层量可见-近红外光谱无损检测

以蚕茧茧层量为研究对象,研究了基于可见-近红外光谱技术的蚕茧茧层量无损检测方法。采用最小二乘支持向量机（least square-support vector machine,LS-SVM）建立可见-近红外光谱模型。采用无信息变量消除算法（uninformative variable elimination, UVE）与连续投影算法（successive projections algorithm, SPA）相结合选取光谱有效波长。结果表明,基于UVE-SPA法进行变量选择,最终将原始光谱的600个光谱变量减少到了8个（673,937,963,982,989,992,995和1 008 nm）。基于此8个变量建立的LS-SVM模型得到了预测集的确定系数（Rp2）为0.5354,误差均方根（RMSEP）为0.0373的预测结果。表明可见-近红外光谱可以用于对蚕茧的茧层量进行无损检测,同时UVE-SPA是一种有效的光谱变量选择方法。     

Silicon Effect on Nutrient Acquisition of Peanut (Arachis hypogaea L.) Under Aluminum Stress

Aluminum (Al) toxicity represents one of the main yield-limiting factors for crops in acid soils. Silicon (Si) is known to increase tolerance in higher plants. This study was conducted to determine whether treatment with Si could improve nutrient uptake by peanut under Al stress. Peanut (Arachis hypogaea L. cv Zhonghua 4) was raised with or without Si (1.5 mM) in the growth chamber under 0 and toxic Al (0.3 mM) levels. Aluminum stress significantly decreased the root- and total-dry weight by 52.4% and 32.0%, respectively. The content of nitrogen (N), phosphorus (P), potassium (K), calcium (Ca), and magnesium (Mg) was significantly decreased, but that of Al increased markedly in shoots and roots after Al exposure at seedling, flower-needle, and pod-setting stage. Silicon alleviates Al toxicity in peanut plants in relation to Al distribution and allocation of tissue P, K, Ca, and Mg by favoring the partitioning of dry mass to roots.     

强度耕作对^137Cs、^210Pbex和有机质剖面分布的影响

很少有关于直接测定强度耕作侵蚀引起土壤退化过程的报道。我们的目的是确定应用137Cs和210Pbex直接定量评价耕作侵蚀对土壤有机质(SOM)坡面运移影响的可能性。我们在黄土高原陡坡耕地上进行了50次犁耕活动,并将与其相临的另一块陡坡耕地作为对照。在对照坡地不同坡位,137Cs浓度均匀分布于上部0～30cm土层,而210Pbex浓度在坡上部和坡中部随土层深度增加呈现线性递减,在坡下部呈指数函数递减。0～30cm土层中土壤有机质含量显著大于30cm深度以下土层,并在坡中部和坡下部呈现与210Pbex类似的土壤剖面分布特征。与对照坡地比较,50次犁耕活动导致坡上部、坡中部0～45cm土层的SOM含量分别降低了38%和47%,坡下部0～100cm土层中的SOM含量却提高了18%。坡上部土壤剖面中137Cs浓度的加权平均值从1.48Bq/kg降低到0.29Bq/kg,坡中部从2.53Bq/kg减少到0.33Bq/kg,坡下部从1.48Bq/kg提高到2.81Bq/kg。210Pbex浓度的剖面加权平均值在坡上部从27.71Bq/kg下降到6.15Bq/kg,在坡中部和坡下部分别从35.46Bq/kg和25.53Bq/kg降低到1.57Bq/kg和19.40Bq/kg。137Cs和210Pbex的剖面分布与SOM在p<0.001水平呈显著相关,相关系数R2值在对照坡地为0.81～0.86,在犁耕实验坡地为0.86～0.91。实验结果表明,环境放射性核素137Cs和210Pbex在黄     

Irrigation Module and Sowing Date Affect Seed Cotton Yield,Quality, Productivity Indices,and Economics of Cotton in North-western India

ABSTRACT

Delayed sowing and imprecise application of irrigation water to cotton has been the major hurdle in sustaining cotton yield in north-western India. Therefore, studies were initiated to ascertain the impact of heavy or normal level of presowing irrigation (PSI), scheduling time of first postsowing irrigation (POSI) under two sowing dates (SDs) on cotton arranged in a split block design replicated thrice. PSI_h (100 mm) recorded 23.2% higher seed cotton yield over PSI_n (70 mm) owing to better yield attributes and higher stand. First POSI at 4 weeks after sowing (WAS) resulted the highest yield (3072 kg ha^?1), while one at 6 WAS (POSI₆) recorded least due to poor population and reduced yield attributes. Water productivity under POSI₆ was lesser by 29.1%, 25.2%, and 16.2% as compared to POSI₃, POSI₄, and POSI₅, respectively. April sown crop out yielded the May sown cotton crop by 939 kg ha^?1 due to better yield attributes. Nitrogen factor productivity (NFP) among SD remained higher by 23.3% for April as compared to May sowing. PSI_h exhibited better NFP over PSI_n. POSI₆ recorded least NFP by 24.4%, 28.5%, and 16.3% as compared to POSI₃, POSI₄, and POSI₅, respectively and was indicative of relatively poor utilization of N under delayed schedules. Therefore, planting in April after heavy PSI and scheduling first POSI at 4 WAS is best strategy for sustaining cotton yield and maximizing farmer profitability.     

Trace Metal Contamination During Grinding of Plant Samples

Inductively coupled plasma-mass spectroscopic (ICP-MS) analysis of leaves from 22 cabbage crops in the Sa P? and B?c Hà districts of Láo Cai Province, North-Western Vi?t Nam, revealed unexpectedly high concentrations of chromium (Cr) and nickel (Ni). The concentrations were strongly linearly related (r² = 0.94), indicating sample contamination during grinding through a stainless-steel hammer mill. We tested this hypothesis in two ways. First, brown rice ground through the same mill was contaminated not only by Cr and Ni, but also cobalt (Co), iron (Fe) and molybdenum (Mo). Second, scanning electron microscopy/energy dispersive analysis of x-rays (SEM/EDS) of the ground samples revealed small fragments with co-located Fe, Cr and Ni, consistent with stainless steel wear fragments. Other grinders may perform differently and we suggest that quality assurance protocols for trace metal analysis of plants should include testing for grinder wear metals. Lastly, brown rice appears to be convenient for investigating contamination of plant tissues during grinding.     

Predicting the Classes and Distribution of Salt-Affected Soils in Northwest Libya

Sodicity and salinity can adversely affect soil structure and are common constraints to plant growth in arid regions. Current remote sensing techniques cannot distinguish between the various classes of salt-affected soils. Field and laboratory measurements of salt-affected soils are time-consuming and expensive. Mapping of the salt-affected soils can be used in soil conservation planning to identify regions with different degrees of limitations. There is a need to use existing field and laboratory measurements to create maps of classes of salt-affected soils. The objectives of this study are to classify salt-affected soils, use existing field data to interpolate and validate geospatial predictions of the classes of salt-affected soils using Geographic Information Systems (GIS), and create maps showing the different classes and distribution of salt-affected soils. The classification framework for salt-affected soils is based on electrical conductivity (ECe), soil pH and the sodium adsorption ratio (SAR), and provides four degrees of limitations to salt-affected soils: slight (normal soils), moderate (saline soils), severe (sodic soils), and extreme (saline-sodic soils). Spatial interpolation of the field data from northwestern Libya was verified by cross-validation, and maps of the salt-affected soils in the region were created. The majority of soils in this region of Libya are normal (slight degree of limitation). Twenty percent of the topsoil is saline-sodic (extreme degree of limitation). Land use recommendations and rehabilitation strategies can be developed from such maps of salt-affected soil classes. The methodology followed in this study can be applied to other arid regions around the world, particularly in developing countries where budgetary constraints limit detailed field and laboratory measurements of sodicity and salinity.     

模拟降雨条件下污染土壤中重金属元素径流迁移特征

探明污染土壤中重金属元素随地表径流迁移特征,对于防控重金属污染物的迁移扩散具有重要意义。通过室内模拟降雨试验,研究不同降雨强度下(55,100,120mm/h)污染土壤中Cd、Pb和Cr随地表径流迁移的特征。结果表明:径流中3种重金属元素总量在产流的前20min内逐渐降低,随后趋于稳定;产流初期(产流10min内)颗粒态重金属占径流重金属总量的比例均在80%以上;随产流时间延长,颗粒态Cd和Cr对径流重金属的贡献率快速降低,其颗粒态/溶解态的比值范围分别为47.0～0.4和12.9～0.4,而Pb则主要以颗粒态的形式存在,溶解态Pb对其随径流迁移的贡献率可忽略不计;随降雨强度增大,径流中总Cd和Pb含量显著降低,而径流中总Cr含量以及溶解态重金属含量无明显变化规律。     

外源胡敏酸对土壤有机质组分变化的影响及其与茶树铅有效性的关系

盆栽试验分析了胡敏酸输入土壤6个月后茶树根际和非根际土壤有机质SOM及其组分,包括水溶性物质WSS、富啡酸FA、胡敏酸HA以及土壤微生物生物量SMB的变化,并对不同土壤有机质组分碳含量和茶树不同组织铅含量作了相关性分析。外源胡敏酸设置3个浓度水平（0、300、600 mg kg-1）,铅设置2个浓度水平（0、300 mg kg-1）,试验共有6个处理。茶苗置于温室培养,6个月后分新叶、老叶、新茎、老茎、根收获。结果表明：根际并非总是富含有机质,未加铅处理根际SOM含量低于非根际;而加铅处理根际SOM含量高于非根际。HA加入土壤能够显著增加根际和非根际SOM和SMB的含量,但是加铅处理和不加铅处理增加幅度不同。高浓度HA显著增加了根际FA的含量,可能是HA部分裂解所致。非根际有机质各组分与茶树铅含量无显著相关关系;根际SOM、HA和茶树不同组织铅含量没有显著相关关系。根际WSS、FA与茶树嫩叶、老叶、嫩茎、老茎以及根部铅含量均存在显著或极显著正相关关系;而根际SMB与茶树各个组织铅含量存在显著负相关关系。总之,胡敏酸施入土壤改变了土壤有机质及其各个组分的含量,而根际某些土壤有机质组分的改变又显著影响了茶树不同组织铅的吸收和分布。     

华北平原灌溉农田的土壤水量平衡和水分利用效率

华北平原农业面临的主要问题是水资源短缺,地下水位持续下降。通过一维土壤水量平衡模型模拟了华北平原不同灌溉方式下农田耗水量和土壤水分深层渗漏的变化,并分析了作物的产量和水分利用效率。结果显示在正常的灌溉条件下,冬小麦季地下水的采补差额超过了200mm,某生育期一定程度的水分亏缺(返青期、拔节期或灌浆期)能明显减少冬小麦的耗水量,但没有明显减少作物的产量。因此,在一定程度上减少灌溉是可行的,但仍不能达到地下水资源的采补平衡。从长远来看,华北平原维持可持续的地下水灌溉开采,应减少冬小麦的种植面积、增加低耗水经济作物的比例。     

Effects of humic acid on remediation of the nutritional deficiency of gerbera in hydroponic culture

Information is scant on the effect of humic acid (HA) on physiological, antioxidant and photosynthesis attributes of gerbera plants undergoing nutrient deficiency in culture solution. Gerbera plants cv. Malibu were grown in a factorial experiment based on a completely randomized design with 3 replications, using 3 different nutrient solutions [complete nutrient solution (NSc), 25% NSc (NS1), and 50% NSc (NS2)] treated with 2 levels of humic acid [0 (HA0) and 500 mg/l (HA1)].The interaction effect of HA and NS showed that HA improved the flower number in NSc, the transpiration in NS1+HA1, photosynthesis rate in NSc+HA1, stomatal conductance (gs) in NS2, mesophyll conductance of leaves in all NS levels and photosynthetic water use efficiency in NSc+HA1. The interaction effect of nutrient solution and HA on antioxidant activity was inconclusive, malondialdehyde content was the highest in NS2 and the lowest in NS1+HA1. The peroxidase activity increased in complete nutrient solution with and without HA and there were no differences among other treatments. Superoxide dismutase activity increased in NS1 and complete nutrient solution with HA and reached the highest in NSc. Humic acid was more effective in nutrient uptake, i.e., nitrogen, phosphorus, potassium, calcium, zinc, and iron (N, P, K, Ca, Zn, and Fe) in complete nutrient solution compared to NS1 and NS2. Conclusively, humic acid can compensate the nutrient deficiency stress of the culture solution in regards to protein synthesis, photosynthesis attributes regardless of the nutrient uptake of gerbera.     

(A)ngstr(o)m-Prescott系数选取对参考作物蒸散量的影响-以黄河中游站点为例

利用黄河中游三个站点近50a的气象资料,研究了(A)ngstr(o)m-Prescott系数不同取值对参考作物蒸散量的影响.结果显示,当ngstrm-Prescott系数as和bs分别相差56%和93%时,引起的ET0差异月值在0.6%～26.2%,年值在173.1～197.4mm(18.2%～21.9%).ngstrm-Prescott系数不同取值对ET0的影响程度随地点而变化,即随海拔和日照时数的增加,对ET0的影响程度相应增大.因此在高海拔和高辐射地区,应对ngstrm-Prescott系数进行校正.尽管FAO56 PM推荐的(A)ngstr(o)m-Prescott系数在世界各地广泛应用,但依据本文结果尚不能确定其在中国能否直接使用.随着辐射数据的不断积累,有必要对这些系数进一步研究、验证和校正.