光谱导数Kalman滤波同时测定苯酚-邻氯苯酚和2,4二氯苯酚

Spectra Derivative Kalman Filter Method for Simultaneous Spectrophotometric Determination of Phenol,2-Chlorophenol and 2,4-Dichlorophenol

文中提出了一个新的、稳定的光谱导数Kalman滤波紫外分光光度法同时定量分析方法，并且成功地将其应用于极难分辨的苯酚一邻氯苯酚和2，4二氯苯酚三组分混合体系的同时测量。选择分析光谱导数的原因是其不仅包含着吸光度，还包括在指定波长位置变化趋向等更多的信息量，这样更利于在吸收峰重叠的位置捕捉有较大差异的信号。利用Kalman滤波可以解决由光谱导数而引起的噪声放大问题，也可以过滤源于实验的噪声以及来自传递模型误差。在260-290nm区间测量了30组浓度各自在1～10mg·L^-1范围的苯酚一邻氯苯酚和2，4二氯苯酚标准混合溶液紫外吸收光谱图。利用分段延伸高次多项式回归模拟求导准确获得吸光度导数，并且利用偏最小二乘法将其制作成Kalman滤波标准工作系数矩阵。通过随机线性离散系统的Kalman滤波器最优平滑计算，对混合体系中的各组分进行定量分析。回收实验显示出，光谱导数Kalman滤波分析法应用于本实验的极难分辨的三组分体系，得到了非常高的回收率，并且在整个实验范围内光谱导数Kalman滤波分析法具有非常好的稳定性。

A novel and stable spectra derivative Kalman filter UV spectrophotometric method was proposed, and applied to the simultaneous determination of ternary mixture of phenol, 2-chlorophenol and 2,4 dicbloropbenol successfully. The reasons for using spectra derivative is that it contains more information including absorbance and its change with wavelength; the obviously different signals can be caught more conveniently at the position of absorbent overlap. The noise from experiment and the errors from transfer model can be solved by Kalman filter. Calibration set with 30 standard solutions （range of 1-10 mg·L^-1） and 61 wavelengths （260-290 m, 0. 5 nm slit width） was used for each sample. The exact value of absorbance derivative was obtained from regressed simulation for the extended 8th order polynomial, and the standard work matrix of derivative spectra Kalman filter was performed from partial least-squares method. The linear discrete Kalman filter was applied to the test. The recovery experiment showed that the derivative spectra Kalman filter simultaneous determination of mixture for phenol, 2-chlorophenol and 2,4 dichlorophenol is not only exact, but also stable.