一类用于多元光度分析的小波基主成分回归法

通过将变尺度小波分解滤噪与特征信息提取相结合，提出一类新的多元光度分析算法－小波基主成分回归（ＰＣＲＷ）方法。该法可有效地减小主成分向量残留噪声所引 误差显著提高多元校正准确性。将其用于分析氯霉素、醋酸对塞米松、尼泊金乙脂体系、得到满意的回收率，与ＰＣＲ法相比，分析结果的总平均相对误差从３．３８％降低到０．８３％。

The Principal Component Regression Based on Wavelet for Multivariate Spectral Analysis

A procedure integrating wavelet transformation with principal component regression (PCR) for multivariate spectral analysis is proposed. The procedure not only effectively decreased the mean relative error of multivariate spectral analysis, but also took advantage of improving the calculating property of PCR. An example for the simultaneous determination of chloramphenicolum. dexamethasoniacetas and aethylparabenum in drugs was used to verify the effectiveness of the method. The results showed that the accuracy of analysis of the procedure was better than that obtained using PCR.