基于可见-近红外光谱变量选择的土壤全氮含量估测研究

doi:10.3864/j.issn.0578-1752.2014.12.010

摘要/Abstract

摘要： 【目的】变量选择是可见光-近红外光谱研究至关重要的步骤，通过分析可见光-近红外光谱不同特征的选择方法筛选出土壤全氮敏感波段，建立基于敏感波段的土壤全氮最佳预测模型，为土壤全氮的快速定量估算提供重要的理论指导依据。【方法】在红壤典型地区江西省吉安县采集代表性土壤样品120个，对可见光-近红外光谱采用主成分分析（PCA）、无信息变量消除（UVE）和无信息变量消除后结合连续投影（UVE-SPA）3种变量特征选择方法，建立基于不同变量选择的偏最小二乘回归（PLSR）模型、最小二乘-支持向量机（LS-SVM）、反向传播神经网络（BPNN）和遗传算法优化的反向传播神经网络（GA-BPNN）模型，从模型对预测集的预测精度分析不同变量选择方法对不同土壤全氮定量估算模型的差异。【结果】经UVE算法筛选后，光谱变量从200个减少至59个，其中可见光波段处10个，其余在近红外光谱的合频区和一倍频区，信息量丰富；进一步采用SPA进行变量选择，得到共线性最小的5个有效波长，分别为820、940、1 040、1 060和1 990nm；基于UVE变量选择建立的PLSR、BPNN、GA-BPNN和LS-SVM模型，经不同的土壤全氮的数据检验，预测精度最高的为LS-SVM，决定系数（R2）、均方根误差（RMSEp）和相对偏差（RPD）分别为0.7492、0.2921和1.8904；基于UVE-SPA特征选择建立的PLSR、BPNN、GA-BPNN和LS-SVM模型对预测集的验证表明，UVE-SPA提取的特征波段建立的LS-SVM建立模型预测效果最好，其建立的LS-SVM定量估算模型预测集的决定系数R2为0.7945，均方根误差RMSEp为0.2499相对偏差RPD为2.0009，模型稳定；基于PCA提取的7个主成分建立的LS-SVM、BPNN和GA-PBNN模型预测性能差，不能用于定量估算土壤全氮。对比相同的变量建立的GA-BPNN和BPNN，GA-BPNN预测性能比BPNN高。【结论】UVE-SPA变量选择方法结合LS-SVM模型能用来估算土壤中的全氮含量，同时UVE-SPA是一种有效的土壤光谱变量选择方法。

关键词: 土壤全氮 , 无信息变量消除（UVE） , 连续投影（SPA） , 偏最小二乘回归（PLSR） , 最小二乘支持向量机（LS-SVM） , 遗传算法优化的反向传播神经网络（GA-BPNN）

Abstract: 【Objective】 Variable selection or feature selection is a critical step in data analysis of visible-near infrared (Vis-NIR) spectrum research. The aim of this study was to determine the soil total nitrogen (TN) contents through building models based on absorption features of soil TN using variable selection methods combined with Vis-NIR spectroscopy, and to provide a basis for the fast estimation of the content of soil TN.【Method】Representative 120 soil samples were collected from the typical red soil area of Ji’an County, Jiangxi Province. The TN contents and the Vis-NIR were measured in the laboratory. Several variable selection methods including principal component analysis (PCA), uninformative variable elimination (UVE) and UVE coupled with successive projections algorithm (SPA) were employed for Vis/NIR data, the models of partial least squares regression (PLSR) with leave-one-out cross-validation, least squares-support vector machine (LS-SVM), the back-propagation neural network (BPNN) and BPNN with optimized threshold and weight using genetic algorithm (GA-BPNN) combined different variable selection methods were calibrated and validated using independent data sets. 【Result】 The results showed that the application of UVE to the wavelengths reduced wavelengths from original 200 to 59 of which located in visible range and the rest located in the region of overtones and combinations in near infrared range. The application of SPA to the wavelengths preselected by UVE further reduced the wavelengths to only 5 for TN, including 820, 940, 1 040, 1 060 and 1 990 nm. LS-SVM models achieved competitive prediction performance compared with PLSR, BPNN and GA-PBNN based on 59 wavelengths with coef?cient of determination (R2) of 0.7492 and root mean square error (RMSEp) of 0.2921 and residual prediction deviation (RPD) of 1.8904 for soil TN. Furthermore, LS-SVM models achieved excellent prediction performance with PLSR, BPNN and GA-PBNN based on 5 wavelengths using variable selection UVE-SPA, with coef?cient of determination (R2) of 0.7945 and root mean square error (RMSEp) of 0.2499 and residual prediction deviation (RPD) of 2.0009 for soil total N. Nevertheless, LS-SVM, BPNN and GA-PBNN models based on 7 principal components was invalid.【Conclusion】 The overall results indicated that SPA was a powerful way for the variable selection, and Vis-NIR spectroscopy incorporated to SPA-LS-SVM was successful for the accurate determination of soil TN.

Key words: soil total nitrogen , uninformative variable elimination (UVE) , successive projections algorithm (SPA) , PLSR , LS-SVM , GA-BPNN the selection

杨梅花1, 2, 赵小敏1, 2, 3, 方倩1, 2, 谢碧裕1, 2. 基于可见-近红外光谱变量选择的土壤全氮含量估测研究[J]. 中国农业科学, 2014, 47(12): 2374-2383.

YANG Mei-Hua-1, 2 , ZHAO Xiao-Min-1, 2 , 3 , FANG Qian-1, 2 , XIE Bi-Yu-1, 2 . Study on Soil Total N Estimation by Vis-NIR Spectra with Variable Selection[J]. Scientia Agricultura Sinica, 2014, 47(12): 2374-2383.

0
/ / 推荐

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: https://www.chinaagrisci.com/CN/10.3864/j.issn.0578-1752.2014.12.010

https://www.chinaagrisci.com/CN/Y2014/V47/I12/2374

参考文献

[1]Shi T, Cui L, Wang J, Fei T, Chen Y, Wu G. Comparison of multivariate methods for estimating soil total nitrogen with visible/ near-infrared spectroscopy. Plant and Soil, 2012, 366(1-2): 363-375.

[2]彭杰, 周清, 张杨珠, 向红英. 有机质对土壤光谱特性的影响研究. 土壤学报, 2013, 50(3): 517-524.

Peng J, Zhou Q, Zhang Y Z, Xiang H Y. Effect of soil organic matter on spectral characteristic of soil. Acta Pedologica Sinica, 2013, 50(3): 517-524. (in Chinese)

[3]韩瑞珍, 宋韬, 何勇. 基于可见/近红外光谱的土壤有机质含量预测. 中国科学: 信息科学, 2010, 40: 111-116.

Han R Z, Song T, He Y. Prediction of soil organic matter content based on Vis/NIR spectroscopy. Scientia Sinica Informations, 2010, 40: 111-116. (in Chinese)

[4]He T, Wang J, Ling Z J, Chen Y. Spectral features of soil organic matter. Geo-Spatial Information Science, 12(1):33-40.

[5]沙晋明, 陈鹏程, 陈松林. 土壤有机质光谱响应特性研究. 水土保持研究, 2003, 10(2): 21-24.

Sha J M, Chen P C, Chen S L. Characteristics analysis of soil spectrum response resulted from organic material. Research of Soil and Water Conservation, 2003, 10(2): 21-24. (in Chinese)

[6]刘焕军, 张柏, 赵军, 张兴义, 宋开山, 王宗明, 段洪涛. 黑土有机质含量高光谱模型研究. 土壤学报, 2007, 44(1): 27-32.

Liu H J, Zhang B, Zhao J, Zhang X Y, Song K S, Wang Z M, Duan H T. Spectral models for prediction of organic matter in black soil. Acta Pedologica Sinica, 2007, 44(1): 27-32. (in Chinese)

[7]张娟娟, 田永超, 朱艳, 姚霞, 曹卫星. 一种估测土壤有机质含量的近红外光谱参数. 应用生态学报, 2009, 20(8): 1896-1904.

Zhang J J, Tian Y C, Zhu Y, Rao X, Cao W X. A novel spectral index for estimating soil organic matter content using near-infrared spectroscopy. Chinese Journal of Applied Ecology, 2009, 20(8): 1896-1904. (in Chinese )

[8]卢艳丽, 白由路, 王磊, 王贺, 杨俐苹. 黑土土壤中全氮含量的高光谱预测分析. 农业工程学报, 2010, 26(1): 256-261.

Lu Y L, B Y L, Wang L, Wang H, Yang L P. Determination for total nitrogen content in black soil using hyperspectral data. Transactions of the CSAE, 2010, 26(1): 256-261. (in Chinese)

[9]张娟娟, 田永超, 姚霞, 曹卫星, 马新明, 朱艳. 基于高光谱的土壤全氮含量估测. 自然资源学报, 2011, 26(5): 881-890.

Zhang J J, Tian Y C, Rao X, Cao W X, Ma X M, Zhu Y. Estimating soil total nitrogen content based on hyperspectral analysis technology. Journal of Natural Resources, 2011, 26(5): 881-890. (in Chinese)

[10]张娟娟, 田永超, 姚霞, 曹卫星, 马新明, 朱艳. 同时估测土壤全氮, 有机质和速效氮含量的光谱指数研究. 土壤学报, 2012, 49(1): 50-59.

Zhang J J, Tian Y C，Rao X, Cao W X, Ma X M, Zhu Y. The Spectral Index for Estimation Soil OM, TN And an Content Simultaneously Using Near-Infrared Spectroscopy. Acta Pedologicasinica, 2012, 49(1): 50-59. (in Chinese)

[11]张娟娟, 田永超, 姚霞, 曹卫星, 马新明, 朱艳. 基于近红外光谱的土壤全氮含量估算模型. 农业工程学报, 2012, 28(12): 183-188.

Zhang J J, Tian Y Y, Yao X, Cao W X, Ma X M, Zhu Y. Estimating model of soil total nitrogen content based on near-infrared spectroscopy analysis. Transactions of the CSAE, 2012, 28(12): 183-188. (in Chinese)

[12]Yang H, Kuang B, Mouazen A M. Quantitative analysis of soil nitrogen and carbon at a farm scale using visible and near infrared spectroscopy coupled with wavelength reduction. European Journal of Soil Science, 2012, 63(3): 410-420.

[13]纪文君, 史舟, 周清, 周炼清. 几种不同类型土壤的VIS-NIR光谱特性及有机质响应波段. 红外与毫米波学报, 2012, 31(3): 277-282.

JI W J ,SHI Z ,Zhou Q, Zhou L Q. VIS NIR reflectance spectroscopy of the organic matter in several types of soils. Journal of Infrared and Millimeter Waves, 2012, 31(3): 277-282. (in Chinese)

[14]Andley U P, Lewis R M, Reddan J R, Kochevar I E. Action spectrum for cytotoxicity in the uva-and uvb-wavelength region in cultured lens epithelial cells. Investigative Ophthalmology & Visual Science, 1994, 35(2): 367-373.

[15] Soares S F C, Gomes A A, Araujo M C U, Filho A R G, Galvão R K H. The successive projections algorithm. TrAC Trends in Analytical Chemistry, 2013, 42: 84-98.

[16] 张勇, 陈莉. 聚类与PCA融合的特征提取方法研究. 计算机工程与应用, 2010, 46(11): 148-150.

Zhang Y, Chen L. Research on feature extraction method based on clustering and PCA fusion. Computer Engineering and Applications, 2010, 46(11): 148-150. (in Chinese)

[17]孔汶汶, 虞佳佳, 刘飞, 何勇, 鲍一丹. 基于高光谱成像技术的番茄茎秆灰霉病早期诊断研究. 光谱学与光谱分析, 2013(3): 733-736.

Kong W W, Yu J J, Liu F, He Y, Bao Y D. Early Diagnosis of Gray Mold on Tomato Stalks Based on Hyperspectral Data. Spectroscopy and Spectral Analysis, 2013(3): 733-736. (in Chinese)

[18]O'Rourke S M, Holden N M. Determination of soil organic matter and carbon fractions in forest top soils using spectral data acquired from visible–near infrared hyperspectral images. Soil Science Society of America Journal, 2012, 76(2): 586.

[19]Viscarra Rossel R A, Chappell A, De Caritat P, McKenzie N J. On the soil information content of visible-near infrared reflectance spectra. European Journal of Soil Science, 2011, 62(3): 442-453.

[20]Xiaobo Z, Jiewen Z, Povey M J W, Holmes M, Hanpin M. Variables selection methods in near-infrared spectroscopy. Analytica Chimica Acta, 2010, 667(1-2): 14-32.

[21] Rossel R A V, Behrens T. Using data mining to model and interpret soil diffuse reflectance spectra. Geoderma, 2010, 158(1-2): 46-54.

[22]郑立华, 李民赞, 安晓飞, 孙红. 基于近红外光谱和支持向量机的土壤参数预测. 农业工程学报, 2010, 26(Suppl.2): 81-87.

Zheng L H, Li Z M, An X F, Sun H. Forecasting soil parameters based on NIR and SVM. Transactions of the CSAE, 2010, 26(Suppl.2): 81-87. (in Chinese)

[23]Xiaobo Z, Jiewen Z, Povey M J, Holmes M, Hanpin M. Variables selection methods in near-infrared spectroscopy. Analytica Chimica Acta, 2010, 667(1): 14-32.

[24]李硕, 汪善勤, 张美琴. 基于可见-近红外光谱比较主成分回归, 偏最小二乘回归和反向传播神经网络对土壤氮的预测研究. 光学学报, 2012, 32(8): 289-293.

Li S, Wang S Q, Zhang M Q. Comparison Among Principal Component Regression,Partial Least Squares Regression and Back Propagation Neural Network for Prediction of Soil Nitrogen with Visible-Near Infrared Spectroscopy. Acta Optica Sinica, 2012, 32(8): 289-293. (in Chinese)

[25]王昶, 黄驰超, 余光辉, 冉炜, 沈其荣. 近红外光谱结合偏最小二乘法快速评估土壤质量. 土壤学报, 2013, 50(005): 881-890.

Wang X, Huang C H, Rang W, Shen Q R. Rapid evaluation of soil quality through a near infrared-partial least squares (NIR-PLS) method. Acta Pedologica Sinica, 2013, 50(005): 881-890. (in Chinese)

[26]Liu F, Jiang Y, He Y. Variable selection in visible/near infrared spectra for linear and nonlinear calibrations: A case study to determine soluble solids content of beer. Analytica Chimica Acta, 2009, 635(1): 45-52.

[27]郑立华, 李民赞, 潘娈, 孙建英, 唐宁. 基于近红外光谱技术的土壤参数BP神经网络预测. 光谱学与光谱分析, 2008, 28(5): 1160-1164.

Zheng L H, Li M Z, Pan L, Sun J Y, Tang N. Estimation of soil organic matter and soil total nitrogen based on NIR spectroscopy and BP neural network. Spectroscopy and Spectral Analysis, 2008, 28(5): 1160-1164. (in Chinese)