基于BP神经网络的SPPs干涉型传感方法
CSTR:
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

TN 201

基金项目:

国家自然科学基金资助项目(62075132);上海市自然科学基金资助项目(22ZR1443100)


SPPs interferometric sensing method based on BP neural network
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    设计了一种由金属狭缝-凹槽构成的表面等离子激元(SPPs)干涉仪阵列结构。在单色光照射下,周期性凹槽激发的SPPs与纳米狭缝处的直接透射光发生干涉,参与干涉的两路光的相位差将导致透射光强发生变化。采用7个具有不同狭缝-凹槽距离的干涉仪结构,通过综合分析各个透射能量的变化,检测环境介质折射率。样品加工采用聚焦离子束刻蚀技术在金属膜上制备了狭缝-凹槽结构,并封装成微流控芯片,完成不同折射率溶液环境下透射光能量分布的采集。通过反向传播(BP)神经网络模型的训练和测试,传感方案在折射率误差±1×10-3范围内的识别准确率可达到96.2%,展现了出色的泛化能力和鲁棒性。

    Abstract:

    A surface plasmon polaritons (SPPs) interferometer array structure composed of a metal slit-groove structure was designed. Under monochromatic light illumination, the SPPs excited by the periodic grooves interfere with the direct transmitted light at the nano-slit. The phase difference between the two interfering light beams caused variations in the transmitted light intensity. Seven interferometer structures with different slit-groove distances were used to detect the refractive index of the surrounding medium by comprehensively analyzing the changes in transmitted energy. The sample was fabricated using focused ion beam milling technology to create the slit-groove structure on a metal film and was then encapsulated into a microfluidic chip. The distribution of transmitted light energy was collected under different refractive index solutions. By training and testing a backpropagation (BP) neural network model, the sensor system achieved an identification accuracy of 96.2% within a refractive index error of ±1×10-3, demonstrating excellent generalization ability and robustness.

    参考文献
    相似文献
    引证文献
引用本文

郝秋月,彭润玲,刘家辰,胡海峰.基于BP神经网络的SPPs干涉型传感方法[J].上海理工大学学报,2026,48(2):159-167.

复制
分享
相关视频

文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2025-01-15
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2026-05-11
  • 出版日期:
文章二维码