|
|
光学薄膜自动设计的多目标优化方法. g% w" r* G- m$ J s
王文梁1 ,2 熊胜明1
3 n6 u8 w% d& u* b; @ S(1 中国科学院光电技术研究所, 四川成都610209 ; 2 中国科学院研究生院, 北京100039)
5 N2 v( p8 }- [4 _ L: l) B; J A: T摘要 光学薄膜中已使用的优化方法都是单目标寻优的,通过分析光学薄膜优化设计的原理,将薄膜设计的物理2 z0 ~8 R) b( J1 J
问题归结为混合离散变量的多目标优化设计的数学模型,并认为这是薄膜设计的一般性模型,现行的单目标优化
+ z: F# K8 n; ^1 ~7 T. d算法只是这个模型的简化。基于这一新思路,并结合多目标优化算法的研究现状,采用了一种基于免疫应答原理& j3 S5 f9 r6 ?6 v( _" g: p% y
的智能型多目标优化算法。该算法隐含并行处理能力,原理上是具备全局搜索能力的自适应随机性算法。将此算
& z2 M& C8 [/ ]% Q: t+ g$ b法运用到光学薄膜设计中,给出了一些优化的设计实例。结果表明,将多目标优化算法引入薄膜设计的新思路是
# d& x) ]" u6 C; H可行的,将来会有较好的发展前景。* U( E! J) _; \' H
关键词 光学设计; 光学薄膜; 多目标优化; 免疫应答- y2 v4 S, j8 S
中图分类号 O484 文献标识码 A doi : 10. 3788/ AOS20082810. 20261 \3 ]& V& ]& l3 u
Mul t i object i ve Op t imi z a t i o n App r oa c h of Op t ic a l Thi n Fi lm Des i g n) t- y/ h3 C- z0 s9 [
Wang Wenliang 1 ,2 Xiong Shengming1* q. v/ _3 h! w7 z) F U9 y2 p
1 Ins t i t ute of Op t ics a n d Elect ronics , Chi nese Academy of Sciences , Chengd u , Sich u a n 610209 , Chi n a
' U4 d! A1 j. Y' p' q2 Gr a d ua te Unive rsi t y of Chi nese Aca demy of Sciences , Beiji ng 100039 , Chi n a) P) v8 T, w4 u0 r9 S3 g
Abs t r act All optimization methods used in optical thin film design can seek only one object . Af ter t he analysis of the
! c' p+ T' A, o3 q; jtheory of optical thin film optimization design , it is p roposed that multiobjective optimization with mixed discrete
+ Z8 Z* n! W& R; ] _variable is the essential mathematical model and can reflect the p hysical essence of optical thin film design. The
% i/ t4 t4 D' Poptimization algorithm for single objective p roblem which has been widely used in the fields of optical thin film design, `( Z, {8 z' Y w9 H0 [' P; J
is only a simplification. Based on this new consideration and the s tatus of multiobjective optimization research , an$ ^( ~2 B8 n( T0 i
immune response2based multiobjective optimization algorithm is adopted to design optical thin film. This algorithm2 z9 }2 b3 U2 T5 ^8 f4 k( O. f
can implement parallel t reatment , and is a self2adaptive random algorithm with ove rall search capability in p rinciple .
, c( b4 D( n# k+ DThe algorithm is used in optical thin film design , and some examples are p resented. According to the results of) e! r$ I. S2 T6 l' e
expe riments , the idea of applying multiobjective optimization app roach to designing optical thin film can be realized in" i, ?2 q) ?3 g0 K5 D% G h* d
theory and have a bright future .
% d$ g$ G( S9 `7 ]: oKey wor ds optical design ; optical t hin film; multiobjective optimization ; response of the immune system |
本帖子中包含更多资源
您需要 登录 才可以下载或查看,没有账号?注册
×
|