预浸料生产中智能控制系统的应用

Application of on-line analysis and monitoring system adopting near-infrared spectroscopy in the production of prepreg

用近红外光谱技术对无碱布/酚醛预浸料的树脂含量、可溶树脂含量和挥发分含量进行在线检测,通过偏最小二乘方法分别建立标准模型,选择光谱预处理方法和PLS的因子数.用近红外方法和标准方法对未知样品进行分析,通过t检验结果显示两种方法没有显著性的差别,利用该方法可以同时预测三项指标,1分钟之内就可以分析一个样品,没有破坏性.如果质量指标不合格,通过自动控制系统发出指令,及时调节工艺参数.研究表明近红外光谱方法能够十分有效和准确对分析预浸料质量.

The on-line determination conducted of resin content, the volatile content and the soluble resin content of a kind of phenolic prepreg were conducted by near infrared diffused reflectance spectrum. Partial least square method was used to build calibration models and chose spectral preprocessing methods as well as number of PLS factors. The t - test results show that there is no significant difference between the near infrared method and the standard method for unknown samples. The NIR spectroscopy method can be used to predict three indexes simultaneously. Besides, the prepreg can be analyzed within than 1 min without to the destruction sample by NIR. The quality prepreg of can be controlled properly by adjusting the technical parameters. The study indicates that the NIR method can sufficiently perform accurate and effective analysis on the quality of prepreg.