基于量子衍生遗传算法的光学薄膜结构分析  

Analysis of Optical Thin Film Structure Based on Quantum-Inspired Genetic Algorithm

作　　者：周祥燕[1] 张超[1] 匡尚奇[1] 龚学鹏[2] 杨海贵 

机构地区：[1]长春理工大学理学院,吉林长春130022 [2]中国科学院长春光学精密机械与物理研究所应用光学国家重点实验室,吉林长春130033 [3]中国科学院长春光学精密机械与物理研究所光学系统先进制造技术重点实验室,吉林长春130033

出　　处：《中国激光》2017年第12期131-138,共8页Chinese Journal of Lasers

基　　金：国家自然科学基金青年科学基金(61405189);吉林省科技发展计划(20150101019JC;20170312024ZG)

摘　　要：掠入射X射线反射(GIXR)由于检测精度高且对检测薄膜的无损伤而被广泛用于薄膜检测和高精度表征。GIXR是一种基于数值拟合的间接检测方法,因此在薄膜微观结构的求解,特别是复杂多层膜膜系的求解过程中对数值优化算法的要求较高。为此提出了基于量子衍生遗传算法(QIGA)的薄膜GIXR拟合求解方法,并基于QIGA对Si单层膜和等周期Mo/Si多层膜的GIXR分别进行拟合求解。结果表明,该方法具有求解速度快、拟合精度高的明显优势,说明QIGA在光学薄膜表征方面有潜在的应用价值。Grazing incident X-ray reflection（GIXR）is widely used in film detection and high accuracy characterization because of its high detection accuracy and nondestructive measurement.However,it is a kind of indirect measurement method,and therefore it requires a superior numerical optimization algorithm when solving thin film parameters,especially for complicated multilayers.A new method based on quantum-inspired genetic algorithm（QIGA）is proposed to realize GIXR fitting.The proposed algorithm is applied in fitting the GIXR of Si single layers and periodic Mo/Si multilayers.The results indicate that the algorithm based on QIGA has fast solving speed and high fitting precision,and QIGA has potential values in the field of thin film characterization.

关 键 词：薄膜 薄膜表征 量子衍生遗传算法 掠入射X射线反射 

分 类 号：TN0[电子电信—物理电子学]