论文部分内容阅读
多组分光光度分析受到日益重视,对无机混合物分析一般将计算机求解与显色反应相结合,先后提出了线性规划、座标轮换、共轭梯度、最小二乘、偏最小二乘、CPA法等,较好地解决了分析化学中光谱重叠严重和难于分离元素的测定。它们各具长处和缺陷。我们将卡尔曼滤波算法与灵敏显色反应相结合,实现了多种金属离子混合物的同时测定,借递推算取代矩阵运算,避免了矩阵求逆过程将量测误差放大而影响结果准确性,在一定程度上改善了结果的可靠性。
Multi-component photometric analysis has received increasing attention. Inorganic mixture analysis generally combines computer-based solution with colorimetric reactions. Linear programming, coordinate rotation, conjugate gradient, least squares, partial least squares, and CPA methods have been proposed , A better solution to the analytical chemistry in the spectral overlap is serious and difficult to separate the element determination. They each have strengths and weaknesses. We combine Kalman filtering algorithm and sensitive color reaction to realize the simultaneous determination of many kinds of metal ion mixtures and deduce the substitution matrix calculation by recursion to avoid the matrix inversion process amplifying the measurement error and affecting the accuracy of the results. To a certain extent, improve the reliability of the results.