摘要
近红外光谱法在对含能材料成分的分析中采用不同波数以及不同区间的变量,并用偏最小二乘法建立数学模型。建模结果说明模型精度较高,缩减后的自变量矩阵X中提取的主成分对原因变量有着更好的解释能力,并提高了计算速度。
In NIR spectroscopy, most of the variables are uninformative with respect to the given problem. In this context, the different subsets of X to be use of in the PLS regression were selected. The established model decreases the data needed in PLS calculating and increases the computing velocity in PLS regression.
出处
《火炸药学报》
EI
CAS
CSCD
2003年第4期78-80,共3页
Chinese Journal of Explosives & Propellants
