论文部分内容阅读
本研究旨在通过近红外光谱法与自动化控制系统的结合,实现感冒灵浓缩过程中的多个关键性指标的在线检测。在感冒灵颗粒浓缩设备上安装在线检测系统,采集过程样品光谱,测定样品中绿原酸含量、蒙花苷含量、固含量和相对密度,建立偏最小二乘法模型,将模型导入近红外光谱工作站,通过与控制系统的连接,实时地自动检测浓缩过程中药液各关键性指标。结果显示,绿原酸含量、蒙花苷含量、固含量和相对密度4个指标的模型相关系数分别为0.963、0.989、0.993和0.918,预测偏差分别为3.71%、4.28%、4.17%、0.24%。模型性能较好,能对样品进行准确定量,系统能成功实现整个浓缩过程的实时检测,并在控制系统显示数据。本近红外光谱检测系统能在感冒灵颗粒生产浓缩过程中对关键性指标进行自动化实时检测,对过程质量监测具有实际意义。
The purpose of this study is to realize the on-line detection of several key indicators in the process of influenza-concentrating by the combination of near-infrared spectroscopy and automated control system. On-line detection system was installed on the device for the concentration of cold spirit granules. The spectra of the samples were collected, and the content of chlorogenic acid, the content of monosodium glycosides, the content of solidarin and the relative density of the samples were determined. Partial least squares model was established and the model was imported into the near infrared spectroscopy workstation , Through the connection with the control system, real-time automatic detection of liquid concentration in the key indicators of the process. The results showed that the model correlation coefficients of chlorogenic acid content, monolin glycoside content, solid content and relative density were 0.963, 0.989, 0.993 and 0.918, respectively. The predicted deviations were 3.71%, 4.28%, 4.17% and 0.24% . The model has good performance, can accurately quantify the sample, the system can successfully real-time detection of the entire concentration process, and display data in the control system. The near infrared spectroscopy detection system can detect the key indicators automatically and real-time during the process of concentration and concentration of the flu-guling and has practical significance for process quality monitoring.