基于SSA-BP极端高温天气驾驶疲劳的检测  

Driving Fatigue Detection in Extreme High Temperature Weather Based on SSA-BP

作　　者：朱兴林 克然木·司马义 姚亮 叶拉森·库肯 ZHU Xing-lin;KERAM·Esmayil;YAO Liang;YELASEN·Kuken(College of Transportation and Logistic Engineering,Xinjiang Agricultural University,Urumchi 830052,China)

机构地区：[1]新疆农业大学交通与物流工程学院,乌鲁木齐830052

出　　处：《科学技术与工程》2023年第3期1254-1261,共8页Science Technology and Engineering

基　　金：国家自然科学基金(71761032);新疆维吾尔自治区研究生科研创新项目(XJ2021G165)。

摘　　要：疲劳驾驶是引起道路交通事故的主要因素之一,提高驾驶疲劳检测精度是预防交通事故的有效措施。为研究驾驶员在极端高温天气下驾驶过程中疲劳程度情况,基于生理反馈仪和卡罗林斯卡嗜睡表(Karolinska sleepiness scale, KSS)主观疲劳调查方法,采集了本地和外地两类驾驶员在正常天气(35℃以下)、高温天气(35~39℃)及极端高温天气(40℃及以上)等3种气温环境下主观疲劳值和生理指标(心电信号与表皮温度)。通过应用皮尔逊相关性分析方法、非参数检验(曼-惠特尼U检验)及配对检验针对各项生理指标进行特征分析。结果表明,在极端高温天气下2类驾驶员主观疲劳值与各项生理指标之间存在相关性;两类驾驶员主观疲劳值和各项生理指标在3种天气下变化存在显著差异;相比于本地驾驶员,在极端天气下外地驾驶员疲劳程度的增加较快。在此基础上,选用麻雀搜索算法(sparrow search algorithm, SSA)优化了BP(back propagation)神经网络预测模型,建立基于SSA-BP驾驶疲劳检测模型,对样本数据进行预测与分类,验证了该模型的有效性。结果表明,标准BP神经网络和SSA-BP疲劳检测精度分别为88.5%、95%,建立的SSA-BP驾驶疲劳检测模型预测效果良好,可为极端高温道路交通安全提供参考与借鉴。Fatigue driving is one of the main factors causing road traffic accidents. It is an effective measure to prevent traffic accidents on improving the accuracy of driving fatigue detection. To study the fatigue degree of drivers during driving in extreme high temperature weather, subjective fatigue values and physiological indexes(electrocardiogram and epidermal temperature) of local and foreign drivers in normal weather(below 35 ℃), high temperature weather(35~39 ℃) and extreme high temperature weather(above 40 ℃) were collected based on physiological feedback instrument and Karolinska sleepiness scale(KSS) subjective fatigue survey method. The characteristics of each physiological index were analyzed by applying Pearson correlation analysis method, nonparametric test(Mann-Whitney U test) and paired test. The results show that there is a correlation between the subjective fatigue values of the two types of drivers and various physiological indexes in extreme high temperature weather;there are significant differences between the subjective fatigue values and various physiological indexes of the two types of drivers under three weather conditions;compared to local drivers, the fatigue level of foreign drivers increases rapidly in extreme weather.Sparrow search algorithm(SSA) was selected to optimize the BP(back propagation) neural network prediction model, and the SSA-BP driving fatigue detection model was established to predict and classify the sample data, which verified the effectiveness of the model. The results showed that the fatigue detection accuracy of standard BP neural network and SSA-BP was 88.5% and 95% respectively. The established SSA-BP driving fatigue detection model has a good prediction effect, which can provide reference for road traffic safety at extreme high temperature.

关 键 词：极端高温天气 麻雀搜索算法(SSA) BP神经网络 驾驶疲劳检测 

分 类 号：U419.91[交通运输工程—道路与铁道工程]