...
首页> 外文期刊>Intelligent Transport Systems, IET >On-road experimental study on driving anger identification model based on physiological features by ROC curve analysis
【24h】

On-road experimental study on driving anger identification model based on physiological features by ROC curve analysis

机译:基于ROC曲线分析的基于生理特征的驾驶员怒气识别模型的道路实验研究

获取原文
获取原文并翻译 | 示例
   

获取外文期刊封面封底 >>

       

摘要

Road rage is a serious psychological issue affecting traffic safety, which has attracted increasing concern regarding driving anger intervention. This study proposed a method for driving anger identification based on physiological features. First, 30 drivers were recruited to perform on-road experiments on a busy route in Wuhan, China. The drivers’ anger could be inducted on the study route by elicitation events, e.g. vehicles weaving/cutting in line, jaywalking, traffic congestion and waiting at red light if they want to finish the experiments ahead of basic time for extra pay. Subsequently, significance analysis was used to determine that five physiological features including heart rate, skin conductance, respiration rate, the relative energy spectrum of θ and β bands of electroencephalogram were effective for driving anger identification. Finally, a linear discriminant model was proposed to identify driving anger based on the optimal thresholds of the five features which were determined by receiver operating characteristic (ROC) curve analysis. The results show that the proposed model achieves an accuracy of 85.84% which is 7.95 and 5.71% higher than the models using back propagation neural network and support vector machine, respectively. The results can provide theoretical foundation for developing driving anger detection devices based on physiological features.
机译:道路狂怒是影响交通安全的严重心理问题,引起了人们对愤怒干预的关注。这项研究提出了一种基于生理特征的愤怒识别方法。首先,招募了30名驾驶员在中国武汉的繁忙道路上进行道路实验。在学习路线上,可能会通过诱发事件(例如,如果他们想在基本时间之前完成实验以赚取额外的报酬,那么他们就可以进行织造/剪线,乱穿马路,交通拥堵和红灯等候。随后,通过显着性分析确定心率,皮肤电导,呼吸频率,脑电图的θ和β谱带的相对能谱这五个生理特征对愤怒识别具有有效作用。最后,提出了一种线性判别模型,该模型基于五个特征的最佳阈值来识别驾驶者的愤怒,这五个阈值是通过接收机工作特性(ROC)曲线分析确定的。结果表明,所提模型的精度达到了85.84%,比使用反向传播神经网络和支持向量机的模型分别提高了7.95%和5.71%。该结果可为开发基于生理特征的驾驶感检测装置提供理论依据。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号