薄膜光学常数的粒子群算法

xx_912

薄膜光学常数的粒子群算法
王党社[1] 张建科[2] 徐均琪[3][1]西安工业大学数理系,陕西西安710032 [2]西安邮电学院应用数理系,陕西西安710121 [3]陕西省薄膜技术与光学检测重点实验室,陕西西安710032摘要：为解决椭偏法测量薄膜厚度和折射率实验数据处理较为复杂的问题，采用一种新的基于群体智能的优化算法——粒子群算法处理实验数据．以单层吸收薄膜的测量为例，利用该算法进行数据处理．实验结果表明，可以同时获得3个薄膜参数（折射率n，消光系数k和薄膜厚度d），而且在确切参数范围未知情况下，大范围内进行搜索仍然能保证快速收敛到最优解．该算法与遗传算法以及利用椭偏仪数据处理软件得出的结果相比较，计算精度高，收敛速度快．

& x! \( {- g, R4 DKeyWord：particle swarm optimization ; ellipsometry ; optical film

. Y* Y9 D3 Y4 w" E关键词：粒子群算法 椭偏法 光学薄膜

. V% E4 O. P  `4 A分类号：O484.41[著者标引]文献标识码：A文章编号：1001-246X（2008）02-0208-05
% U8 c% C, X2 ]7 K) `Particle Swarm Algorithm for Film Optical Constants WANG Dangshe, ZHANG Jianke, XU Junqi1. Department of Mathematics and Physics, Xi＇ an Technological University, Xi＇ an 710032, China ; 2. Department of Mathematics and Physics, Xi ＇ an University of Post and Telecommunications, Xi ＇ an 710121 China ; 3. Shannxi Province Thin Film Technology and Optical Test Key Lab, Xi ＇ an 710032, China
Abstract：A new algorithm - particle swarm optimization （PSO）, based on swarm intelligence, is applied to data processing of film parameters by ellipsometry. For a monolayer absorption film, reflective index, extinction coefficient and film thickness are obtained simultaneously by PSO. Even if the range of parameters is unknown, optimal solution can be obtained rapidly with search in a wide range. Compared with ellipsometry processing and genetic algorithm, the method exhibits high computational precision and high convergence speed.

; ^2 ]$ S& b; ]KeyWord：particle swarm optimization ; ellipsometry ; optical film