论文部分内容阅读
利用漫反射法获得甲氧苄啶粉末药品的近红外光谱(波长范围1100-2500nm),采用化学计量学中的偏最小二乘法(PLS)及不同的光谱预处理方法(标准归一化(SNV)、一阶导数和二阶导数)对光谱进行信息提取和分析,对甲氧苄啶粉末药品进行了无损定量分析,以样品中甲氧苄啶为活性成分建立了最佳的数学校正模型。同时讨论了主成分数对PLS模型定量预测能力的影响,并对所得结果做出了比较。
The near infrared spectrum of trimethoprim powder (wavelength range 1100-2500nm) was obtained by diffuse reflectance method. Partial least squares (PLS) and different spectral pretreatment methods (standard normalization (SNV ), First derivative and second derivative) were used to extract and analyze the spectra. Non-destructive quantitative analysis of trimethoprim powder drugs was carried out. The best mathematical model was established based on trimethoprim in the sample as the active component. At the same time, the influence of principal components on the quantitative prediction ability of PLS model was discussed, and the results were compared.