基于模拟退火-蚁群变步长优化算法的椭偏数据反演分析  

Inverse analysis of the ellipsometric data based on dynamic step ant colony optimization-simulated annealing algorithm

作　　者：赵明琳 吴嘉瑶 童荣景 赵明瑶 ZHAO Minglin;WU Jiayao;TONG Rongjing;ZHAO Mingyao(School of Science, Jiangsu University of Science and Technology, Zhenjiang 212100, China;Rice Future Technology Co. Ltd., Beijing 100123, China)

机构地区：[1]江苏科技大学理学院,镇江212100 [2]北京大米未来科技有限公司,北京100123

出　　处：《江苏科技大学学报（自然科学版）》2021年第3期108-113,共6页Journal of Jiangsu University of Science and Technology:Natural Science Edition

基　　金：国家自然科学基金青年资助项目(11804126)。

摘　　要：椭圆偏振光谱方法是获取薄膜复光学常数和厚度的最优光学测量手段之一,椭偏方程作为超越方程,其逆向求解过程中的反演算法直接影响着椭偏数据的处理效率与精度.以前期的蚁群算法为基础,为进一步提高算法的收敛速度和跳出局部最优解的能力,研究了模拟退火算法和蚁群算法的融合策略,并提出了一种基于最优蚂蚁的变步长方法,通过动态改变最优蚂蚁的领域局部搜索步长,提升算法的精细化搜索能力,最终给出了模拟退火-蚁群变步长优化算法.应用该优化算法分析了高温超导薄膜FeSe的椭偏光谱,测试结果表明,该混合优化算法可以实现椭偏数据的精确反演分析,并且具有更快的收敛速度和更优的评价函数.Ellipsometry is one of the most advanced optical technologies to measure optical properties and the thickness of films.Because the ellipsometer equation is a transcendental equation,the inverse analysis algorithm plays an important role in ellipsometric data processing,which directly influences the efficiency of the converse solution process and the precision of the inversion results.To improve the convergence rate and the ability to escape from local optimum of the original ant colony algorithm,we study the fusion strategy of the ant colony algorithm and the simulated annealing algorithm,and propose a dynamic step strategy with the elite ant individual dipartition.Combining merits of ant colony algorithm and simulated annealing algorithm and improving the local-searching step strategy,a new algorithm named dynamic step ant colony optimization-simulated annealing algorithm is present to solve the ellipsometric data processing problem.Finally,the test result of the superconducting FeSe film proves the feasibility and reliability of the improved algorithm,in addition,the convergence rate and the excellence of the merit function(MSE)is enhanced.

关 键 词：椭圆偏振光谱 模拟退火-蚁群算法 变步长 

分 类 号：O433.4[机械工程—光学工程]