节点文献
近红外光谱与卷烟样品常规成分的关系模型研究
Study On Model of Near-infrared Spectroscopy and Chemical Components of Cigarettes
【摘要】 采用偏最小二乘法建立了近红外光谱与卷烟样品常规成分之间的关系模型 ,该模型可用于卷烟样品常规成分含量的预测。为了消除与建模无关的波长 ,提出了一种波长选择方法 ,研究结果表明 ,该方法既提高了运算速度 ,又提高了预测精度
【Abstract】 The models of near\|infrared spectroscopy (NIR) and the chemical components of cigarette samples were established with partial least square method (PLS), which could be used to predict the contents of these components in cigarette with its NIR. The interference of the irrelevant information while modeling was minimized by selection of wavelength of NIR,thus improved both operation speed and accuracy of the prediction.
【关键词】 近红外光谱;
偏最小二乘法;
波长选择;
卷烟;
【Key words】 Near infrared spectroscopy; NIR; Partial least squares; PLSC; Wavelength selection; Cigarette;
【Key words】 Near infrared spectroscopy; NIR; Partial least squares; PLSC; Wavelength selection; Cigarette;
【基金】 国家烟草专卖局项目 (1 1 0 2 0 0 1 0 1 0 4 2 )
- 【文献出处】 烟草科技 ,Tobacco Science & Technology , 编辑部邮箱 ,2002年05期
- 【分类号】TS411
- 【被引频次】68
- 【下载频次】297