用于近红外光谱分析的化学计量学方法研究与应用进展

Research and Application Progress of Chemometrics Methods in Near Infrared Spectroscopic Analysis

分析模型的建立是近红外光谱分析的核心技术之一,本文综述了近些年在近红外光谱分析方法中出现的一些新算法和模型建立策略,如基于核函数的非线性校正方法、集成(或共识)的建模策略、多维分辨和校正方法、基于局部样本的建模策略以及二维相关光谱等,并给出了一些方法的具体算法。

The establishment of calibration models is one of the most important techniques in near infrared spectroscopic analysis. In this paper, some newly developed chemometrics methods in the field such as kernel partial least square, ensemble methods, local methods, multiway analysis, and two-dimensional correlation spectroscopy are introduced with 48 references. The algorithms of several methods are given and discussed.