基于遗传算法和模拟退火算法的振动光谱成分分析算法  

Vibration Spectral Component Analysis Based on Genetic Algorithm and Simulated Annealing Algorithm

作　　者：黄凡 张旭坤 孙陆[1] 刘伟伟[1] Huang Fan;Zhang Xukun;Sun Lu;Liu Weiwei(Tianjin Key Laboratory of Micro-Scale Optical Information Science and Technology,Institute of Modern Optics,Nankai University,Tianjin 300071,China)

机构地区：[1]南开大学现代光学研究所,天津市微尺度光学信息技术科学重点实验室,天津300071

出　　处：《激光与光电子学进展》2020年第9期216-222,共7页Laser & Optoelectronics Progress

基　　金：国家重点研发计划(2018YFB0504400);天津市自然科学基金(16JCQNJC05100)。

摘　　要：在大气污染物中有害气体成分众多,运用太赫兹、红外等振动光谱检测所得不同成分的吸收峰较难辨别。设计了一种根据所测光谱中可能包含的物质的理论光谱分析所测光谱中不同物质所占比重的算法。利用实验峰值与不同理论峰值的对应关系构建超定方程组,分别基于遗传算法和模拟退火算法进行求解,得到实验光谱中各物质理论光谱的贡献。为了验证所提算法的准确性,利用该算法对已知水团簇组成的水汽系统的实验光谱进行了分析,结果表明:所提算法能够得到不同成分的比重,具有较强的实用性。In the atmospheric pollutants,there are many harmful gas components that is difficult to distinguish the absorption peaks of different components obtained by infrared and terahertz spectral detection.This paper designs an algorithm for analyzing the proportion of different substances in the measured spectrum based on the theoretical spectrum of the substances that may be included in the measured spectrum.The overdetermined equations constructed by using the corresponding relationship between experimental peaks and different theoretical peaks are solved based on genetic algorithm and simulated annealing algorithm,so the contribution of theoretical spectra of various materials in the experimental spectrum is obtained.In order to verify the accuracy of the proposed algorithm,we use this algorithm to analyze the experimental spectra of the water vapor system composed of known water clusters.The results show that the proposed algorithm can get the specific gravity of different components better and has stronger practicality.

关 键 词：光谱学 振动光谱 成分分析 超定方程组 遗传算法 模拟退火算法 

分 类 号：O433.4[机械工程—光学工程]