共查询到18条相似文献，搜索用时 187 毫秒
1.
采用偏最小二乘法(PLS)建立八角中反式茴香脑和莽草酸含量的近红外光谱定量分析模型。利用近红外光谱技术采集八角粉末的近红外光谱图(NIR),以高效液相色谱法测定的八角中反式茴香脑和莽草酸质量分数作为参考值，经过光谱处理软件将近红外光谱图与质量分数参考值进行关联，反式茴香脑的定量分析模型采用二阶导数作为预处理方法，以7 502～5 446.2和4 601.5～4 246.7 cm-1为波数范围，维数为10;莽草酸的定量分析模型采用一阶导数加直线减法作为预处理方法，以7 502～6 098.1和5 450.1～4 597.6 cm-1为波数范围，维数为8。结果表明：建立的八角中反式茴香脑和莽草酸含量的快速无损定量检测模型的决定系数(R2)分别为90.86%和93.13%,校正集的内部交叉检验均方根误差(RMSECV)分别为0.158和0.285,验证集的均方根误差(RMSRP)分别是0.068 7和0.171,配对T检验P值分别为0.761和0.194,均大于0.05,测量值与参考值偏差较小，建立的模型能较准确地测定八角中反... 相似文献
2.
基于近红外光谱与BP神经网络预测落叶松木屑的含水率 总被引：2，自引：0，他引：2
利用近红外光谱(NIR)技术结合BP神经网络定量预测了落叶松木屑的含水率。首先对采集的落叶松木屑原始近红外光谱进行9点平滑及多元散射校正预处理,然后利用主成分分析法提取光谱数据主成分作为BP神经网络的输入,最后建立BP神经网络预测模型并采用交叉验证法对模型进行验证。所建模型校正集的相关系数R为0.98,校正集的均方根误差RMSEC为0.001 7;预测集的相关系数R为0.99,预测集的均方根误差RMSEP为0.001 5。研究表明,此方法可以实现对落叶松木屑含水率的快速预测。 相似文献
3.
《林业与环境科学》2021,(4)
在950～1650 nm的光谱范围内,使用DA2700型近红外光谱仪采集了110个湿加松Pinus elliottii × P.caribaea松针粉末样本的光谱数据。结合实际测定值,采用偏最小二乘(PLS)回归法并选择最佳光谱预处理方法和最佳主成分数,建立湿加松松针组织16种黄酮物质总含量的快速预测模型。结果表明:采用一阶导数(FD)与滤波拟合(SG)相结合法对光谱数据进行预处理且当主成分数为10时,可得最优模型。其校正集相关系数(R_c)和交互验证集相关系数(R_v)分别为0.852 1和0.705 9。校正集均方根误差(RMSEC)和交互验证集均方根误差(RMSEV)分别为6.361 0和9.150 9。外部验证集测定值和模型预测值之间的相关系数为0.8537。综合表明所建模型预测精度高、可靠性强,可用于湿加松松针组织16种黄酮物质总含量的预测。 相似文献
4.
以东北小兴安岭林区带岭林业局东方红林场的土壤为研究对象,对120个土壤样品近红外光谱做去噪、Savitzky-Golay平滑和多元散射校正预处理,利用偏最小二乘（PLS）法建立关于土壤碳含量和吸光度之间的定量分析模型,并进行模型校、验证及部分预测集样品碳含量预测.结果表明：主成分数为4时,模型最优.校正模型的决定系数R2和均方根误差（RMSE）分别为0.784和5.752;验证模型的决定系数R2和均方根误差（RMSE）分别为0.621和7.521,预测集样品的实测值和预测值的决定系数R2达到0.735,均方根误差RMSE为7.202,预测标准差SEP为10.356.应用近红外技术可以实现对小兴安岭次生林土壤碳含量的有效预测,为大面积快速测定土壤碳含量提供理论依据与技术支撑,进而为林分土壤碳循环的相关研究提供新的思路. 相似文献
5.
采集农田、林地和盐碱地不同类型的土壤样本,采用偏最小二乘法结合OSC方法建立土壤有机质反演模型,运用交叉验证和外部验证相结合的评价方法进行比较分析。结果显示：采用平滑＋MSC＋OSC方法对光谱进行预处理,可以提高预测模型的精度。OSC因子个数和PLS主因子个数分别为6和4时,交叉验证决定系数R2为0.990 1,均方根误差为0.297 5,外部验证决定系数R2为0.926 1,均方根误差为0.283 6,模型达到最优。表明对光谱进行OSC预处理后建模是可行的,OSC降低与浓度阵无关的光谱信号,并且减少建立模型的主因子个数,进一步提高模型的精度和稳定性。 相似文献
6.
《林产化学与工业》2017,(4)
为了构建湖南常见3种木本油料植物种子含油率近红外光谱通用模型,收集了98个油桐、96个油茶和96个核桃样本,采集了粉碎后种仁的近红外光谱(NIR),测定了样本含油率,分别采用偏最小二乘法(PLS)及径向基神经网络法(RBFNN)建立油桐+油茶+核桃、油桐+油茶、油桐+核桃和油茶+核桃4个混合样本集含油率的NIR通用模型。对PLS模型,4个样本集(验证集)的相关系数(R_p)分别为0.963、0.881、0.965和0.967,预测均方根误差(RMSEP)分别为2.78、3.31、2.47和2.70,相对标准偏差(RSD)分别为4.87%、6.51%、4.03%和4.55%;RBFNN模型的R_p分别为0.958、0.877、0.959和0.966,RMSEP分别为3.34、2.55、2.85和2.54,RSD分别为5.85%、5.02%、4.66%和4.28%。结果表明:构建油桐、油茶和核桃3种木本油料植物种子含油率近红外光谱通用性检测模型具有可行性。 相似文献
7.
【目的】木材的基本密度在木材质量等级评定中起着重要的作用,是木材分流及精细化利用的重要依据。利用近红外光谱技术,实时监测木材性质,掌握木材性质的变化,为进一步制定和改善林木培育方法提供理论依据。【方法】借助树木生长锥对椴树活立木取样,以椴树样品基本密度真值和近红外光谱数据为输入,分别通过卷积平滑、一阶导数和二阶导数预处理方法来实现近红外光谱数据的预处理,建立了基于偏最小二乘法(PLS)的椴树木材基本密度的近红外估测模型。【结果】在350~2 500 nm波段范围内,一阶导数预处理的椴树木材基本密度模型是最优的,校正集相关系数为0.964 8,校正均方根误差为0.002 7,验证集相关系数为0.943 2,预测均方根误差为0.003 3。在对近红外光谱数据进行去噪优化处理,构建椴树木材基本密度模型后,在500~2 300 nm波段范围内,一阶导数预处理椴树木材基本密度模型依旧最优,其校正集相关系数为0.987 1,校正均方根误差为0.001 6,验证集的相关系数是0.948 6,预测的均方根误差是0.002 1。【结论】选择特定的预处理方法,结合样本特征,建立椴树木材基本密度模型,可以显著降低建模成本,提高模型预测精度,快速测定椴树木材的基本密度。 相似文献
8.
《林业工程学报》2016,(6)
针对实际生产中多采用混合品种木片制浆的情况,探讨了利用近红外光谱分析技术对混合木片的水分含量进行快速测定的可行性。通过国产便携式阿达玛光谱仪采集183个木片样品的光谱,经过预处理后,利用偏最小二乘法和完全交互验证方式建立4组木片的近红外光谱数据与其水分含量之间的关联模型。4个模型的相关系数均达到098以上,交叉验证均方根标准差在3%以内,相对分析误差值在68~103之间。利用建好的模型对预测集样品的水分含量进行预测,其中全局模型的预测标准差在127~240之间,结果较为精确。结果表明,尽管木材品种对木片近红外光谱水分特征吸收存在一定的影响,但在光谱预处理后,再以纯种木片与混合木片一起建立的全局模型,具有较好的适应性和较高的预测精度,可用于实际生产中快速测量混合木片的水分含量。 相似文献
9.
为快速测定森林土壤的有机碳含量,从取自小兴安岭带岭林业局东方红林场的120个土壤样品中采集350～2500 nm的土壤近红外光谱数据,对光谱做一定的预处理后,运用主成分分析法压缩提取前8个主成分,结合BP神经网络非线性方法建立土壤有机碳含量的预测模型并进行验证。结果表明,验证集的相关系数为0.78002,均方根误差为0.5002,预测集的相关系数为0.84941,均方根误差为0.4538。应用近红外光谱技术及BP神经网络非线性方法建模可以有效地预测土壤的有机碳含量,为野外大面积快速测定森林土壤碳含量提供了技术依据。 相似文献
10.
木材顺纹抗压强度是评价木材力学性能的重要指标,而传统测量方法操作复杂、精确度低。以桦木为例,提出基于近红外光谱技术(NIR)的SEPA-VISSA-RVM木材顺纹抗压强度模型,实现对其更加精确的预测。试验选取100个木材试件,在900~1700 nm近红外光谱波段上采集数据并测量抗压强度真值;然后采用卷积平滑(SG)方法进行光谱预处理;使用采样误差分布分析(SEPA)作为变量空间迭代收缩算法(VISSA)的改进策略进行特征波长优选;最后通过粒子群优化算法(PSO)优化核函数参数并建立相关向量机(RVM)的预测模型。试验表明:在特征波长优选方面,以偏最小二乘法(PLS)建模为基础的SEPA-VISSA方法,其预测决定系数为0.9593,预测均方根误差为2.8995,相对分析误差为3.0256,光谱变量数由512减小到111个,占总波长的22%,均优于VCPA、CARS和VISSA算法;在建模预测方面,以SEPA-VISSA所选波长为基础的RVM模型,PSO优化的拉普拉斯(Laplacian)核函数的核宽度为10.4043,决定系数为0.9449,预测均方根误差为2.0432,相对分析误差为4.2936,预测效果优于PLS和SVR。因此,基于近红外光谱的SEPA-VISSA-RVM建模能够实现对桦木顺纹抗压强度更准确和稳定的无损检测。 相似文献
11.
Ana Alves António Santos Denilson da Silva Perez José Rodrigues Helena Pereira Rogério Simões Manfred Schwanninger 《Wood Science and Technology》2007,41(6):491-499
A total of 910 maritime pine (Pinus pinaster Aiton) wood discs, belonging to a genetic trial of 80 families with 11–12 trees per family, were used in this study. A near
infrared (NIR) partial least squares regression (PLSR) model for the prediction of Kappa number of Pinus pinaster Aiton pulps obtained from samples pulped under identical conditions was calculated. Very good correlations between NIR spectra
of maritime pine pulps and Kappa numbers in the range from 58 to 100 were obtained. Besides the raw spectra, spectra pre-processed
with ten methods were used for PLS analysis (cross validation with 57 samples), showing that even after test set validation
(with 34 samples) no model decision could be made due to almost identical statistics. The final evaluation that proved the
predictive power of the models by predicting pulps with unknown Kappa numbers allowed choosing a model according to a minimal
number of outliers found during this process. The minimum–maximum normalized spectra in the wave number range from 6,110 to
5,440 cm−1 used for the calculation gave the best model with a root mean square error of prediction of 2.3 units of Kappa number, a
coefficient of determination of 95.9%, and one PLS component. The percentage of outliers during evaluation was 0.9%. 相似文献
12.
用常规方法测定了104个速生桉木样品的综纤维素、聚戊糖、酸不溶木素及苯醇抽出物含量并采集了样品的近红外光谱。对原始光谱进行多元散射校正后,运用偏最小二乘法和交互验证的方法,确定最佳主成分数并建立样品相关化学成分含量的校正模型。独立验证中综纤维素、聚戊糖、酸不溶木素和苯醇抽出物模型的决定系数 Rval2分别为0.9067、0.9033、0.9504、0.9570;预测均方根误差(RMSEP)分别为0.33%、0.50%、0.31%、0.17%;相对分析误差(RPD)值分别为3.22、3.20、4.43、4.73;绝对偏差(AD)分别为?0.53%~0.60%、?0.95%~0.77%、?0.55%~0.52%、?0.22%~0.29%,4个校正模型较好地预测了验证集样品的化学成分含量,基本满足制浆造纸工业中快速测定速生桉木原料的需求。 相似文献
13.
Jón B. Hauksson Göran Bergqvist Urban Bergsten Michael Sjöström Ulf Edlund 《Wood Science and Technology》2001,35(6):475-485
This work was undertaken to investigate the feasibility of using near infrared spectroscopy (NIR) and partial least squares regression (PLS) as a tool to characterize the basic wood properties of Norway Spruce (Picea abies (L.) Karst.). The wood samples originated from a trial located in the province of Västerbotten in Sweden. In this trial, the effects of birch shelterwoods (Betula pendula Roth) of different densities on growth and yield in Norway spruce understorey were examined. All Norway spruce trees in each shelterwood treatment were divided into three growth rate classes based on diameter at breast height (1.3?m) over bark. Five discs were cut from each tree (i.e. from the root stem, and at 20%, 40%, 60%, and 80% of the total height). The discs from 40% tree height were used (i.e., where the largest variations in annual ring widths and wood density were found). A total of 27 discs were selected. The discs were used for measuring annual ring widths, wood density, average fiber length and the fiber length distributions. Milled wood samples prepared from the discs were used for recording NIR spectra. PLS regression was used to generate prediction models for the wood properties (Y-matrix) and NIR spectra (X-matrix) as well as between the wood properties (Y-matrix) and the fiber length distributions (X-matrix). One set of models was generated using untreated spectra and fiber length distributions. For a second set of models the structure in the X-matrix, which was orthogonal to the matrix described by the wood properties, was eliminated using a soft target rotation technique called orthogonal signal correction (OSC). The PLS model obtained using “raw” untreated NIR spectra and fiber length distributions had a poor modeling power as evidenced by the cumulative Q2 values. For the PLS models based on untreated NIR spectra the cumulative Q2 values ranged from a minimum of 16% (wood density) to a maximum of 46% (no. of annual rings). Orthogonal signal correction of the X-matrix (NIR spectra or fiber length distributions) gave PLS models with a modeling power corresponding to cumulative Q2 values well in excess of 70%. The improvement in predictive ability accomplished by the OSC procedure was verified by placing four of the 27 observations in an external test set and comparing RMSEP values for the test set observations without OSC and with OSC. 相似文献
14.
采集了常见制浆材(桉木、相思木及杨木)样品的近红外光谱,测定了样品的基本密度、综纤维素、木质素和苯醇抽出物含量,用人为控制水分的方法测定了样品的水分含量。对原始光谱进行预处理后,分别运用偏最小二乘法(PLS)、LASSO算法、支持向量机法(SVR)和人工神经网络法(BP-ANN)建立基本密度、水分含量、综纤维素、木质素和苯醇抽出物含量的预测模型。对预测模型进行独立验证,结果显示:LASSO算法建立的基本密度和综纤维素模型性能最优,其预测均方根误差(RMSEP)分别为0.006 3 g/cm~3和0.49%,绝对偏差(AD)范围分别为-0.008 8~0.009 6 g/cm~3和-0.85%~0.87%;PLS建立的水分含量模型及苯醇抽出物模型最优,RMSEP值分别为1.21%和0.24%,AD范围分别为-1.99%~2.03%和-0.35%~0.38%;SVR建立的木质素模型最优,RMSEP值为0.43%,AD范围为-0.76%~0.74%,均满足制浆造纸工业中对误差的要求。 相似文献
15.
A use of near-infrared (NIR) spectroscopy for rapidly predicting the longitudinal growth strain (LGS), as a detector of growth stress, was described. NIR spectra and LGS were measured from peripheral locations of three Sugi (Cryptomeria japonica) green logs. Partial least squares regression model for predicting LGS was developed using the spectral range spanning 770–1200?nm. The predicted LGS was correlated with that measured by the strain gauge method. The coefficient of determination and the root mean square error of prediction were 0.61 and 0.013%, respectively. The predicted peripheral LGS distribution moderately fitted with the measured one. Our results indicate that NIR spectroscopy has a potential to evaluate the magnitude of longitudinal growth stresses on green logs. 相似文献
16.
粗皮桉木材力学性质的近红外光谱方法预测 总被引：1，自引：0，他引：1
以人工林粗皮桉木材为研究对象,采用常规力学测试方法和近红外光谱方法对其无疵小试样力学性质进行研究。用近红外光谱仪采集试样表面的近红外光谱,对采集的近红外漫反射光谱进行导数预处理并对不同波段光谱建立校正模型,以1/3试样作为预测集对校正模型进行验证。结果表明:二阶导数预处理、350～25000nm全光谱波段、径切面和弦切面平均光谱值对粗皮桉木材力学性质模型预测效果最好。抗弯弹性模量和抗弯强度、顺纹抗压强度的实测值与近红外光谱方法的预测值存在较好的相关性,相关系数均大于0.88,相对分析误差大于2.0,表明利用近红外光谱方法预测人工林粗皮桉木材力学性质效果较好。 相似文献
17.
Nondestructive evaluation of drying stress level on wood surface using near-infrared spectroscopy 总被引：1，自引：0，他引：1
Ken Watanabe Isao Kobayashi Shuetsu Saito Naohiro Kuroda Shuichi Noshiro 《Wood Science and Technology》2013,47(2):299-315
A nondestructive technique for swiftly measuring the stress level of the surface of wood is proposed, which is important for process control in timber drying. Partial least squares (PLS) regression models for predicting surface-released strain (ε) were developed using NIR spectra obtained from Sugi (Cryptomeria japonica D. Don) samples during drying. The predictive ability of the models was evaluated by PLS analysis and by comparing NIR-predicted ε with laboratory-measured values. The PLS regression model using the NIR spectra pre-processed by MSC and second derivatives with a wavelength range of 2,000–2,220 nm showed good agreement with the measurement (R 2 = 0.72). PLS analysis identified the wavelengths around 2,035 nm as making significant contributions to the prediction of ε. Orthogonal signal correction (OSC) was an effective pre-processing technique to reduce the number of factors required for the model using the wavelength range 1,300–2,500 nm. However, the predictive ability of the OSC-corrected model was not improved. Elapsed times to reach the maximum tensile stress (T max) and the stress reversal point (T rev) at the wood surface during drying were detected correctly for 75 % of the samples. The results show that NIR spectroscopy has potential to predict the drying stress level of the timber surface and to detect critical periods in drying, such as T max and T rev. 相似文献
18.
Tetsuya Inagaki Manfred Schwanninger Reina Kato Yohei Kurata Warunee Thanapase Pratuang Puthson Satoru Tsuchikawa 《Wood Science and Technology》2012,46(1-3):143-155
Density and fiber length belong to the parameters that are used by the pulping industry as indicators of wood quality for different industrial processes and final paper products. The feasibility of Fourier transform near-infrared (FT-NIR) spectroscopy for the non-destructive evaluation of fiber length and air-dry density of fast-growing E. camaldulensis from Thailand was investigated using 50 samples. NIR spectra taken from solid wood and air-dry density as well as fiber length were used for partial least squares (PLS) regression analyses. It is the first time that the fiber length of E. camaldulensis solid wood could be predicted with high accuracy and precision and that the ratios of performance to deviation (RPD) obtained are the first that fully fulfill the requirements of AACC Method 39-00 (AACC 1999) for screening in breeding programs (RPD?≥?2.5). The RPDs for cross-validation (test set validation) of the NIR-PLS-R models of 3.3 (3.8) for air-dry density and 3.5 (3.9) for fiber length allow drawing the conclusion that the models are at least applicable for screening in breeding programs as they lie in-between screening (RPD?≥?2.5) and quality control (RPD?≥?5). Even when 40% of the samples were removed in cross-validation of the air-dry density model, the RPD is 3.2, which confirms that the model is robust, stable, and well qualified for prediction. The good model statistics obtained in this study might be due to the fact that measurement sites for the measurement of NIR spectra, air-dry density, and fiber lengths were strictly coincided. 相似文献