近红外光谱分析技术在线测定颈痛颗粒流化床制粒过程中的水分

目的 利用近红外光谱分析技术快速测定颈痛颗粒流化床生产过程中水分的含量。方法 以颈痛颗粒为研究对象，对流化床生产设备的视窗进行改造，采用Micro NIR PAT-U对生产过程进行在线监测，利用偏最小二乘法建立颈痛颗粒流化床生产过程中水分的定量分析模型。结果 最终建立的水分含量PLS定量模型RMSEC，RMSECV，RMSEP， R2c，R2cv，R2p分别为0.245 4%、0.287 1%、0.279 8%、0.909 1、0.880 6、0.904 6，外部验证的平均相对误差为3.0%。结论 近红外光谱分析技术可以替代水分测定仪，对颈痛颗粒流化床生产过程中水分的含量进行在线、快速、准确地测定。

On-line monitoring study of moisture in the fluidized bed production of Jingtong Keli by near-infrared spectroscopy

Objective To rapidly determine the moisture content in fluidized bed production of neck pain granules using near-infrared spectroscopy.Methods The study was carried out on neck pain granules by modifying the viewport of the fluidized bed production equipment and using Micro NIR PAT-U for online monitoring of the production process. The quantitative analysis model of moisture in the fluidized bed production process of neck pain granules was developed by using partial least squares method.Results The final quantitative moisture content PLS models were established with RMSEC, RMSECV, RMSEP, R2c, R2cv, and R2p of 0.245 4%, 0.287 1%, 0.279 8%, 0.909 1, 0.880 6, and 0.904 6 respectively, and the mean relative error of external validation was 3.0%.Conclusion Near-infrared spectroscopy can replace the moisture test apparatus for online, real-time, rapid and accurate determination of moisture content in the fluidized bed production of neck pain granules.