实数编码遗传和单纯形混合算法在膜系优化设计中的应用  

Applications of a new genetic and simplex algorithm in optical film design and optimization

作　　者：彭东功[1] 吴永刚[1] 焦宏飞[1] 曹鸿[1] 张莉[1] 郑秀萍[1] 付联效[1] 

机构地区：[1]同济大学,精密光学工程技术研究所,上海200092

出　　处：《量子电子学报》2008年第1期49-55,共7页Chinese Journal of Quantum Electronics

基　　金：国家自然科学基金资助项目(10175049)

摘　　要：在传统遗传算法(GA)基础上,通过引入标准偏差函数构造了新的适应度函数,同时提出了一种自动降温的方法来控制退火选择策略中的温度。将这种改进的实数编码遗传算法(FGA)和单纯型算法(SA)有机结合起来,形成了新的膜系优化算法-实数编码遗传和单纯形混合算法,并编制了优化程序。实例表明该算法优化性能优越,既具有强大全局搜索能力,又能很好地实现局部搜索功能。用该算法实现了中心波长534 nm,带宽35 nm的可见光波段凹陷滤波器和高性能中性分束镜。On the base of traditional genetic algorithm, a new fitness function is constructed by introducing a standard deviation function. Meanwhile, a proposed method of automatically lowering temperature is used to control the temperature of controlling anneal-selecting strategy. A new method used to optimize and design film system is developed by combining the improved float-coded genetic algorithm with simplex algorithm, and an optimized program is coded. Examples show that this program is excellent in both global and local searching capacity. Some notch filters with center wavelength 534 nm, bandwidth 35 nm and neutral beam splitter mirrors with high performance in visible band are designed.

关 键 词：薄膜光学 膜系优化 遗传算法 单纯型算法 凹陷滤波器 中性分束镜 

分 类 号：O484.4[理学—固体物理]