论文部分内容阅读
快速、便捷、准确、低成本的液体检测方法对公众的安全有着重大的意义。文章采用近红外光谱分析方法,研究了对易燃液体的快速鉴别方法。采集了典型易燃液体和非易燃液体样品的近红外吸收光谱图,在建立判别预测模型时,分别对两类液体样品的光谱数据进行主成分分析,选取了特征波长点,从而提取样本的特征值。利用判别分析法和极限学习机判别法分别建立了判别预测模型,两种方法的分类对训练集的准确率均能达到98%以上,对测试集能达到100%,并且耗时极短。结果表明近红外光谱可以快速鉴别易燃液体与非易燃液体。
Fast, convenient, accurate, low-cost liquid detection methods are of great importance to public safety. In this paper, near-infrared spectroscopy is used to study the rapid identification of flammable liquids. Near infrared absorption spectra of typical flammable liquids and non-flammable liquids were collected. When the discriminant model was established, the principal component analysis (PCA) was performed on the spectral data of two types of liquid samples respectively, and the characteristic wavelength points were selected to extract the Eigenvalues. Discriminant analysis and discriminant method are respectively used to establish the discriminant prediction model. The accuracy of the two methods can reach more than 98% for the training set, 100% for the test set, and the time-consuming is very short. The results show that near infrared spectroscopy can quickly identify flammable liquids and non-flammable liquids.