基于函数型数据分析的疲劳驾驶状态识别  

Driving fatigue recognition based on functional data analysis

作　　者：刘卓 邹淑丽 上官鹏鹏 邱桃荣[1] 白小明[1] LIU Zhuo;ZOU Shuli;SHANGGUAN Pengpeng;QIU Taorong;BAI Xiaoming(School of Information Engineering ,Nanchang University,Nanchang 330031,China)

机构地区：[1]南昌大学信息工程学院,江西南昌330031

出　　处：《南昌大学学报（工科版）》2020年第2期194-198,共5页Journal of Nanchang University(Engineering & Technology)

基　　金：国家自然科学基金资助项目(61762045,61841201)。

摘　　要：针对目前常用于脑电信号特征提取的方法存在不能很好地反映信号的连续性、内部动态变化,导致基于脑电信号的疲劳驾驶识别的准确性不理想的状况,一种基于函数型数据分析进行脑电信号特征提取的方法被提出。所提出的方法是通过计算人体在静息状态和疲劳状态的函数性差异来表示特征和进行特征的提取。基于所提出的脑电信号特征提取方法构建了1个疲劳驾驶状态识别模型,并通过使用5种常见分类器在采集的疲劳驾驶脑电信号数据上进行测试,结果表明所提出的基于函数型数据分析的脑电信号特征提取方法有效,所构建疲劳驾驶状态识别模型能获得好的识别效果,最好的识别准确率达到94.82%。A new approach was proposed to extracting effective features from a given electroencephalogram(EEG)dataset,aiming at reflecting the characteristics of the EEG dataset with continuity and internal dynamic changes and improving the performance of the driving fatigue recognition.The proposed method was designed by computing functional difference values based on function data analysis between normal state and fatigue state.A model for detecting driving fatigue state applied was built using the proposed method,and the model was tested on the real driving fatigue EEG dataset under five different classifiers.Experiment results showed that the proposed method was effective,and the performance of the model used to detect driving fatigue state was good,the best accuracy reaching 94.82%.

关 键 词：函数型数据分析 特征提取 函数性差异 脑电信号 疲劳驾驶 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]