论文部分内容阅读
违规使用香兰素是影响婴幼儿奶粉食用安全的重要因素之一,对其进行快速检测具有重要的意义.本文采用长光程傅里叶变换红外光谱(FT-IR)以顶空采样的方式,对掺杂香兰素的奶粉体系中的挥发性气体进行高灵敏检测,巧妙规避了奶粉复杂基质对香兰素分析的干扰,并显著提升了香兰素检测的灵敏度.为了进一步提升定量分析方法的检测灵敏度,本文发展了多尺度建模方法,该方法有机结合了离散小波变换(DWT)和偏最小二乘法(PLS),充分利用气体红外光谱中的时/频多尺度信息,从复杂、变动的奶粉红外光谱中准确提取微弱的香兰素吸收信息.结果表明,DWT-PLS算法可显著提升模型的预测精度和可靠性,进而推动长光程红外光谱检测技术在奶粉安全检测中的应用.
Violation of the use of vanillin is one of the important factors affecting the safety of infant milk consumption and its rapid detection is of great significance.In this paper, long-range Fourier transform infrared spectroscopy (FT-IR) with headspace sampling , Highly sensitive detection of volatile gases in the vanadium-doped milk powder system cleverly avoided the interference of the complex matrix of milk powder on the analysis of vanillin and significantly enhanced the sensitivity of the detection of vanillin.In order to further enhance the quantitative analysis In this paper, a multi-scale modeling method is developed. This method combines the discrete wavelet transform (DWT) and the partial least squares (PLS) method and makes full use of the time / frequency multi-scale information in gas infrared spectroscopy. The results showed that the DWT-PLS algorithm can significantly improve the prediction accuracy and reliability of the model, and then promote the long optical path infrared spectroscopy in milk powder safety testing application.