电子学报 ›› 2022, Vol. 50 ›› Issue (8): 1801-1810.DOI: 10.12263/DZXB.20201267

• 学术论文 • 上一篇    下一篇

基于深度主题模型的飞行员脑疲劳检测

吴奇1,2,3, 陈琪琦1, 彭献永4, 仇峰1   

  1. 1.上海交通大学自动化系,上海 200240
    2.系统控制与信息处理教育部重点实验,上海 200240
    3.上海工业智能管控工程技术研究中心,上海 200240
    4.中国矿业大学低碳能源与动力工程学院,江苏 徐州 221116
  • 收稿日期:2020-11-09 修回日期:2021-10-18 出版日期:2022-08-25
    • 通讯作者:
    • 彭献永
    • 作者简介:
    • 吴 奇 男,2009年于东南大学控制理论与工程专业获工学博士学位.上海交通大学系统控制与信息处理教育部重点实验室教授.研究方向为空战神经生理认知学.E-mail: Edmondqwu@gmail.com.
      陈琪琦 女,2022年获上海交通大学工学学士学位.主要研究方向为深度学习、脑影像分析.E-mail: cqq.jx@sjtu.edu.cn.
      彭献永 男,2011年获东南大学博士学位.中国矿业大学低碳能源与动力工程学院副教授.主要研究方向为智能电厂、状态监测与先进控制.
      仇 峰 男,2020年获上海通大学硕士学位.主要研究方向为数字电厂、认知计算.E-mail: edmondqwu@sjtu.edu.cn.
    • 基金资助:
    • 国家自然科学基金 (U1933125)

Pilot’s Brain Fatigue Detection Based on Deep Topic Model

WU Qi1,2,3, CHEN Qi-qi1, PENG Xian-yong4, QIU Feng1   

  1. 1.Department of Automation,Shanghai Jiao Tong University,Shanghai 200240,China
    2.Key Laboratory of System Control and Information Processing,Ministry of Education,Shanghai 200240,China
    3.Shanghai Engineering Research Center of Intelligent Control and Management,Shanghai 200240,China
    4.School of Low-Carbon Energy and Power Engineering,China University of Mining and Technology,Xuzhou,Jiangsu 221116,China
  • Received:2020-11-09 Revised:2021-10-18 Online:2022-08-25 Published:2022-09-08
    • Corresponding author:
    • PENG Xian-yong

摘要:

飞行员脑疲劳状态检测需要解决脑认知图谱生成和脑疲劳检测模型构建问题.针对第一个问题,本文通过等距方位投影法将全脑电极位置的脑疲劳指标映射为二维脑功率图谱,形成一种新型脑认知图谱.针对第二个问题,本文建立一种深度主题学习模型,即深度潜狄利克雷模型(Deep Latent Dirichlet Model,DLDM),解决了飞行员疲劳状态主题学习问题.DLDM深度模型通过多项式分布逐层扩展脑功率图谱中蕴含的概率分布信息,推理脑功率图谱的层次概率分布特征,实现更有效的飞行员疲劳状态主题学习.同时为了避免启发式假设,本文提出一种有效的不同层与主题间自适应学习率的随机梯度下降推断方法,更加高效地推理DLDM网络结构参数.实验结果显示,DLDM网络可以逐层扩展脑功率图谱中蕴含的概率分布信息,推理出更丰富的抽象特征信息,实现脑疲劳认知主题学习.对比其他脑疲劳检测方法,本文方法分类精度可提升2%.

关键词: 脑功率图谱, 疲劳认知状态, 主题学习, 深度模型, 概率推断, 狄利克雷模型

Abstract:

The detection of a pilot's brain fatigue state faces two important problems, which are how to generate a brain cognitive map and how to build a brain fatigue detection model. To solve the first problem, this paper uses the isometric azimuth projection to map brain fatigue indicators into a new type of brain power map. To solve the second problem, this work develops a deep latent Dirichlet model(DLDM), which solves the topic detection problem of pilot fatigue state. DLDM expands the probability distribution information contained in the developed brain power map layer by layer through multiple distributions, infers their hierarchical probability distribution characteristics, and gets more effective topic detection accuracy of pilot fatigue state. In order to avoid heuristic assumptions, this paper also proposes an effective stochastic gradient descent inference method with an adaptive learning rate between different layers and topics to more efficiently infer structure parameters of DLDM. The experimental results show that the DLDM can expand the probability distribution information of the brain power map layer by layer, infer richer abstract feature information, and detect brain fatigue cognitive topic. Compared with other brain fatigue detection methods, the classification accuracy of the proposed method can be improved by 2%.

Key words: brain power map, fatigue cognitive state, topic learning, deep model, probability inference, Dirichlet model

中图分类号: