应用遗传算法的扫描式ICP-AES光谱定性分析

A Genetic Algorithm Approach to Qualitative Analysis in Inductively Coupled Plasma-Atomic Emission Spectroscopy

用遗传算法进行了扫描式ICP AES的自动光谱定性分析。无需标样并能克服光谱干扰自动报出未知样品中的元素成分及含量范围 ,复制、杂交及变异算子分别为 0 6 ,0 4和 0。实际样品的光谱分析结果初步表明 :对光谱峰位移进行一定的预处理 ,并在适应值函数中考虑谱线的强度信息后 ,遗传算法能够应用于光谱图的解析。

Genetic algorithm (GA) is used in automatic qualitative analysis by a sequential inductively coupled plasma spectrometer (ICP-AES) and a computer program is developed in this paper. No any standard samples are needed, and spectroscopic interferences can be eliminated. All elements and their concentration ranges of an unknown sample can be reported. The replication rate P, crossover rate P, and mutation rate of the genetic algorithm were adjusted to be 0.6, 0.4 and 0 respectively. The analytical results of GA are in good agreement with the reference values. It indicates that, combined with the intensity information, the GA can be applied to spectroscopic qualitative analysis and expected to become an effective method in qualitative analysis in ICP-AES after further work.