文献类型：期刊文献

中文题名：基于外周生理信号的疲劳驾驶监测研究

英文题名：Drive Fatigue State Detection Based on Peripheral Physiological Signal Analysis

作者：莫泽坤[1];徐逸峰[1];蒋麒憬[1];张晨曦[1];陈兰岚[1]

机构：[1]华东理工大学信息科学与工程学院自动化系,上海200237

年份：2018

卷号：44

期号：2

起止页码：97

中文期刊名：汽车实用技术

外文期刊名：Automobile Applied Technology

基金：上海市大学生创新性实验计划资助项目(S17080);国家自然科学基金青年基金资助项目(61201124)资助

语种：中文

中文关键词：外周生理信号;驾驶疲劳监测;支持向量机

外文关键词：Peripheral Pbysiologieal Signal; Drive Fatigue Deteetion; Support Vector Maehine

摘要：疲劳驾驶是交通事故的主要原因之一。监测驾驶者的精神状态,在疲劳时及时发出警报是降低交通事故的有效手段之一。通过对驾驶员在行车过程中一些外周生理指标(血容、肌电、皮电、呼吸、皮温等)的变化进行监测,在信号预处理和特征提取后,实验选取典型清醒疲劳样本,结合支持向量机研究多特征组合下的清醒疲劳状态区分效果并对区分方法进行评价,并进一步以典型样本作为训练集,对整个连续驾驶过程的状态进行判断。结果表明,使用支持向量机对外周生理信号典型样本的分类取得了较高正确率,也能较准确地对整个连续驾驶过程的状态进行判断。
Fatigue driving is one of the major causes of traffic accidents. Timely warning under fatigue status is one of the effective means to reduce traffic accidents. In order to monitor the driver＇s mental state, multiple peripheral physiological signals including Blood volume pulse （BVP）, Electromyography （EMG）, Galvanic skin response （GSR）, Skin temperature （Temp）, and Respiration （Resp） were recorded. The typical alert and fatigue samples were selected from the continuous driving process. After signal preprocessing and feature extraction, Support Vector Machine （SVM） with multi-feature combination was implemented to explore the discriminative effect of the alert and fatigue states. Furthermore, driver＇s states during the whole continuous driving process was judged based the trained model on the typical samples. The results show that using SVM to classify typical samples of peripheral physiological signals has achieved relatively high accuracy and can accurately judge the driver＇s status of the whole continuous driving process.