Real-Time Analysis of soil Total N and P with Near Infrared Reflectance Spectroscopy
The tide salt clay in Zhejiang Province was selected as research object, and the theories of analyzing soil N and soil P with NIR spectroscopic techniques were explored. Six group samples were collected from a rice farm, after taking the samples to the lab, we added nutritional water of different height to the six groups, and then drying, rubbing. At last, one hundred twenty samples were got from six groups equably. Different preprocessing methods were carried on the spectrum data, such as standard normal variate (SNV), multivariate scatter correction (MSC) and smoothing of moving average. Different calibration models were established and the performance of these models were compared with different preprocessing methods. After comparison, smoothing of moving average was found to be the most appropriate spectral preprocessing method. 96 samples were randomly selected from 120 samples to be the calibration set, the remaining 24 samples were used as validation samples. Two discriminant analysis models were developed using partial least squares (PLS) method and least squares-support vector machine (LS-SVM) method respectively. The performances were validated by the samples in the validation set. The correlation coefficients (r) between the real values and predicted ones by discriminant analysis models using PLS were 0.9454(N)、0.9327(P) respectively, and using LS-SVM were 0.9503(N)、0.9547(P) respectively. The root mean standard error of prediction (RMSEP) were 0.9503(N)、0.9547(P) by PLS, 0.0378(N)、0.0101(P) by LS-SVM. The results showed that NIRS could be used to evaluate the soil N and soil P.
Visible and near-infrared reflectance spectroscopy (NIRS) total N total P partial least square regression (PLS) Least squares-support vector machine (LSSVM)
YUAN Shi-Lin MA Tian-yun SONG Tao HE Yong
College of Biosystems Engineering and Food Science,Zhejiang University,Hangzhou 310029,China
国际会议
北京
英文
1-6
2009-10-14（万方平台首次上网日期，不代表论文的发表时间）