基于便携式短波近红外光谱技术检测了土壤总氮含量。 采集浙江省文城地区农田土壤样本243个, 将土壤样本分为三组, 一组未经过粉碎、 过筛等处理, 一组做过2 mm筛处理, 一组过0.5 mm筛过处理, 采用usb4000便携式光谱获取土壤光谱数据, 结合(savitzky-golay, SG)平滑算法, 波长压缩算法和小波变换对原始数据进行预处理, 然后采用竞争性自适应重加权、 随机青蛙和连续投影算法进行特征波长选择。 基于全光谱建立了偏最小二乘回归和基于特征波长建立了极限学习机和LS-SVM模型。 结果表明过筛处理后的样本模型结果优于未过筛的样本模型结果, 过0.5 mm筛处理的土壤样本模型预测结果略优于过2 mm筛处理的土壤样本模型预测结果, 最优预测集的决定系数为0.63, 预测均方根误差为0.007 9, 剩余预测偏差为1.58。 表明便携式仪器检测土壤总氮含量, 经过过筛处理的土壤样品检测结果优于未过筛土壤样品检测结果, 建议土壤样品检测总氮含量时需经过过筛处理, 这样得到的结果较为理想, 在此基础上采用性能较好的光谱仪器采集数据, 以减小原始光谱噪声。

Near infrared spectroscopy analysis as a reliable, rapid, little sample preparation requirement, low-cost, convenient, nondestructive and green technique becomes more and more important in the area of soil nutrition measurement. Near infrared spectroscopy are highly sensitive to C—H, O—H and N—H bonds of soil components such as total nitrogen (TN) making their use in the agricultural and environmental sciences particularly appropriate. The analytical abilities of near infrared spectroscopy depend on the repetitive and broad absorption of light by C—H, O—H and N—H bonds. A total of 243 soil samples were collected from wencheng, Zhejiang province. Raw spectra and wavelength-reduced spectra with 3 different pretreatment methods (Savitzky-Golay smoothing (SG), Reduce (RD), and Wavelet Transform (WT)) were compared to determine the optimal wavelength range and pretreatment method for analysis. Spectral variable selection is an important strategy in spectrum modeling analysis, because it tends to parsimonious data representation and can lead to multivariate models with better performance. In order to simply calibration models, the preprocessed spectra were then used to select sensitive wavelengths by competitive adaptive reweighted sampling (CARS), Random frog and Successive Projections Algorithm (SPA) methods. Different numbers of sensitive wavelengths were selected by different variable selection methods with Wavelet Transform (WT) preprocessing method. Partial least squares (PLS) was used to build models with the full spectra, and Extreme Learning Machine (ELM) and LS-SVM were applied to build models with the selected wavelength variables. The overall results showed that PLS and LS-SVM models performed better than ELM models, and the LS-SVM models with the selected wavelengths based on SPA obtained the best results with the determination coefficient (R2), RMSEP and RPD were 0.63, 0.007 9 and 1.58 for prediction set. The results indicated that it was feasible to use portable short wave near-infrared spectral technology to predict soil total nitrogen and wavelengths selection could be very useful to reduce redundancy of spectra.