...
首页> 外文期刊>Optical engineering >Accelerating optics design optimizations with deep learning
【24h】

Accelerating optics design optimizations with deep learning

机译:加速光学设计优化深入学习

获取原文
获取原文并翻译 | 示例
           

摘要

We show that design optimizations, an integral but time-consuming component of optical engineering, can be significantly sped-up when paired with deep neural networks (DNNs). By using the DNN indirectly for choosing initializations and candidate preselection, our approach obviates the need for large networks, big data-sets, long training epochs, and excessive hyperparameter optimization. For a 16-layered thin-film design problem, our surrogate-assisted differential evolution (DE) algorithm is able to achieve similar optimal solutions as that of an unassisted DE using only 10% of the function evaluation budget. Our approach is a promising option for the optimal design of optical devices and systems.
机译:我们表明,当与深神经网络(DNN)配对时,可以显着加速设计优化,这是一种光学工程的耗时分量,可以显着加速。通过间接使用DNN来选择初始化和候选预选,我们的方法避免了对大型网络,大数据集,长训练时期的需求和过度的近额计优化。对于16层薄膜设计问题,我们的代理辅助差分演进(DE)算法能够使用仅使用函数评估预算的10%来实现类似的最佳解决方案。我们的方法是光学设备和系统最佳设计的有希望的选择。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号