提出了一种基于相关系数阈值法区间选择的卡尔曼滤波分光光度法,该方法可同时检测工业废水中的铜、钴、镍痕量离子,而无需任何分离步骤。该方法首先制备40组铜、钴、镍混合标准溶液,用多元线性回归法求解吸光系数矩阵;然后通过提出的相关系数阈值法选择灵敏度较高的波长区域进行滤波;最后根据滤波方差确定滤波终点,得到铜、钴、镍的测定质量浓度。铜的质量浓度的线性检测范围为0.5~5.0 mg/L,钴的质量浓度的线性检测范围为0.2~2.0 mg/L,镍的质量浓度的线性检测范围为0.3~3.0 mg/L。铜、钴、镍的平均相对误差分别为2.862%、2.464%、3.781%,均小于5.000%。铜、钴、镍的预测均方根误差分别为0.1124,0.0279,0.0663。结果表明,所提方法简单快速,可以边扫描边滤波,易于联机分析。

A Kalman filter spectrophotometry based on interval selection of correlation coefficient threshold method is proposed to simultaneously detect trace ions of copper, cobalt, and nickel in industrial wastewater without any separation steps. Firstly, 40 groups of mixed standard solutions of copper, cobalt, and nickel are prepared, and the absorbance coefficient matrix is solved by multiple linear regression method. Then, a correlation coefficient threshold method is proposed to select the wavelength region with high sensitivity for filtering. Finally, the filter end point is determined by the filter variance to obtain the estimated mass concentrations of copper, cobalt, and nickel. The linear detection ranges of mass concentrations of copper, cobalt, and nickel are 0.5-5.0, 0.2-2.0, and 0.3-3.0 mg/L, respectively. The average relative errors of copper, cobalt, and nickel are 2.862%, 2.464%, and 3.781%, respectively, which are all less than 5%. The predicted root mean square errors of copper, cobalt, and nickel are 0.1124, 0.0279, and 0.0663, respectively. The results show that the proposed method is simple and rapid, can filter while scanning, and is easy to analyze on-line.