摘要
为了提高酶免分析仪检测精度,建立了基于改进遗传算法的多组分定量分析校正模型。通过自适应寻优空间技术和小生境技术,对遗传算法进行改进。采用全息凹面光栅作为分光元件,搭建扫描式紫外可见分光光度实验系统,对食品工业中最常见的苋菜红和胭脂红二组分混合水溶液进行定量分析,并与最小二乘回归方法进行了比较。实验结果表明:改进遗传算法预测精度与最小二乘法相近,对苋菜红和胭脂红的预测均方根误差(RMSEP)分别为0.88,1.71μg/m L;收敛速度明显快于传统遗传算法,为多组分同时测定提供了新的思路。
In order to improve detection precision of automatic enzyme-linked immunoassay analyzer,a correction model for multicomponent quantitative analysis based on improved genetic algorithm is developed. Through adaptive search space technology and niche technology are proposed to improve performance of genetic algorithm. A UV-Vis spectrophotometer is built by using holographic concave grating as spectroscope. The comparison is made between the proposed method and least squares regression method by application to simultaneous determination of amaranth and carmine samples which are commonly used in the food industry. Experimental results indicate that prediction precision of the improved genetic algorithm and LS are similar,for amaranth and carmine are 0. 88,1. 71μg/mL,respectively;and convergence rate is significantly faster than ordinary GA. It also provides a new idea for multicomponent simultaneous determination.
出处
《传感器与微系统》
CSCD
2015年第6期46-49,共4页
Transducer and Microsystem Technologies
基金
北京市自然科学基金重点资助项目(KZ201010772032)
教育部"长江学者与创新团队"发展计划资助项目(IRT1212)
国防技术基础科研资助项目(J132012C001)
关键词
全息凹面光栅
多组分定量分析
改进遗传算法
自适应寻优空间
小生境技术
holographic concave grating
multicomponent quantitative analysis
improved genetic algorithm
adaptive search space
niche technology
