%0 Journal Article %T 基于近红外光谱技术的辛伐他汀片剂生产过程多参数的质量监控 %A 林翔 %A 彭熙琳 %A 陈晓春 %A 李晖 %J 工程科学与技术 %D 2015 %R 10.15961/j.jsuese.2015.04.029 %X 中文摘要: 利用近红外光谱(NIR)技术,并结合化学计量学方法,建立了辛伐他汀片剂制备过程水分含量、制片压力、片剂硬度、主药含量4参数的近红外定量分析模型。采用偏最小二乘法(PLS)建立校正模型,以相关系数( R)、校正均方差( RMSEC)、预测均方差( RMSEP)和内部交叉验证均方差( RMSECV)为模型性能评价参数。其中水分含量校正模型的 RMSEC为0.682, R为0.990 30,内部预测集的 RMSEP为0.672, R为0.990 6,模型的 RMSECV为0.990 50;制片压力校正模型的 RMSEC为0.181, R为0.985 40,内部预测集的 RMSEP为0.165, R为0.976 3,模型的 RMSECV为0.469 00;片剂硬度校正模型的 RMSEC为0.158, R为0.991 30,内部预测集的 RMSEP为0.176, R为0.989 4,模型的 RMSECV为0.340 00;主药含量校正模型的 RMSEC为0.322, R为0.988 78,内部预测集的 RMSEP为0.473, R为0.980 2,模型的 RMSECV为0.551 00。结果表明,所建模型具有良好的预测能力,能有效地应用于辛伐他汀固体制剂生产过程中上述各参数的监控。</br>Abstract:Near infrared spectroscopy (NIR) technology with chemometric techniques was applied to fast analyze various parameters in the process of Simvastatin tablets producing.The four parameters including water content,compression force,tablet hardness and Simvastatin content were monitored by NIR quantitative analysis model,which used partial least squares (PLS) method.The correlation coefficient ( R),root mean square error of calibration ( RMSEC),root mean square error of prediction ( RMSEP) and root mean square error of cross-validation ( RMSECV) were used to assess the predictive ability and robustness of the different PLS models.The results showed that this method was validated for monitoring the four parameters in Simvastatin tablets production. %K 近红外光谱技术 偏最小二乘法 辛伐他汀< %K /br> %K near infrared spectroscopy (NIR) partial least squares (PLS) Simvastatin %U http://jsuese.ijournals.cn/jsuese_cn/ch/reader/view_abstract.aspx?file_no=201401178&flag=1