摘要
目的:运用近红外光谱技术(NIRS),结合偏最小二乘法(PLS),对筋骨痛消丸中水浸出物进行快速测定。方法:将近红外光谱图与筋骨痛消丸水浸出物的含量进行关联,对光谱预处理方法和建模区间进行考察;采用近红外光谱技术结合偏最小二乘法,建立测定筋骨痛消丸中水浸出物含量的近红外光谱定量分析模型。结果:98份样品所建立的定量分析模型内部交叉验证决定系数(R2)为0.983 26,校正均方根偏差(RMSEC)为0.21,预测均方根偏差(RMSEP)为0.22。18份验证集样品真实值和预测值的平均相对偏差为0.09%。结论:近红外光谱技术结合偏最小二乘法建立的定量分析模型可用于对筋骨痛消丸中水浸出物含量的快速测定。
Objective: To rapid determine the content of water-extract in Jingutongxiao pills by near-infrared spectroscopy (NIRS) combined with partial least squares (PLS). Method: The near-infrared spectra was associated with the content of water-extract in Jingutongxiao pills, and the spectral pretreatment method and modeling interval were studied. The quantitative analysis model of water-extract in Jingutongxiao pills was established by combining the NIRS with PLS. Result: The 98 samples of the quantitative analysis model for correlation coefficients (R2), root mean square error of calibration (RMSEC) and root mean square error of predication (RMSEP) were O. 983 26, 0.21 and 0.22, respectively. 18 validation samples of real value and prediction value of the mean relative deviation is 0.09%. Conclusion: It is feasible to rapid determine the content of water-extract in Jingutongxiao pills by NIRS and PLS.
出处
《中国实验方剂学杂志》
CAS
CSCD
北大核心
2015年第4期52-55,共4页
Chinese Journal of Experimental Traditional Medical Formulae
基金
河南省重点科技攻关计划项目(122102310095)
河南中医学院博士科研启动基金项目(BSJJ2011-03)
河南省高等学校青年骨干教师项目(2012GGJS-093)
关键词
近红外光谱技术
偏最小二乘法
筋骨痛消丸
水浸出物
快速测定
near-infrared spectroscopy
partial least squares
Jingutongxiao pills
water-extract
rapid determination