摘要
在单调的高速公路环境下进行了20组实车试验,获取了驾驶员的反应时间及自我评价量表。对驾驶员的反应时间数据进行了分析,构建了驾驶员反应时间概率密度函数。采用支持向量机(SVM)模型对驾驶疲劳进行分类,其中惩罚因子通过遗传-粒子群混合算法进行寻优。以驾驶员反应时间作为输入量,疲劳等级作为输出量,对驾驶疲劳进行量化分类,准确率为80.67%。
Twenty sets of real vehicle experiments were conducted on the monotonous highway in order to obtain driver reaction time and the self-assessment scale.The data of the diver reaction time were analyzed to establish a probability density function of the rection time.A support vector machine model was chosen to classify the driving fatigues and a genetic algorithm-particle swarm optimization hybrid algorithm was applied to optimize the penalty factor.The reaction time was the input of the model and driving fatigue grade was the out put which was divided into two levels.The accuracy of the model is 80.67%.
作者
郭梦竹
李世武
GUO Meng-zhu;LI Shi-wu(College of Transportation,Jilin University,Changchun 130022,China)
出处
《吉林大学学报（工学版）》
EI
CAS
CSCD
北大核心
2020年第3期951-955,共5页
Journal of Jilin University:Engineering and Technology Edition
基金
国家重点研发计划项目(2018YFB1600501).
关键词
载运工具运用工程
疲劳分类
反应时间
支持向量机
遗传-粒子群混合算法
vehicle operation engineering
fatigue classification
reaction time
support vector machine(SVM)
genetic algorithm-particle swarm optimization hybrid algorithm
