基于脑电信号的动车司机对突发事件反应时间的预测方法研究  

Forecasting Method of Reaction Time of EMU Driver to Emergencies Based on EEG

作　　者：郭孜政[1,2] 谭茜 吴志敏 潘雨帆 张俊 GUO Zizheng;TAN Xi;WU Zhimin;PAN Yufan;ZHANG Jun(School of Transportation and Logistics,Southwest Jiaotong University,Chengdu 610031,China;National United Engineering Laboratory of Integrated and Intelligent Transportation,Southwest Jiaotong University, Chengdu 610031,China;Guangdong Provincial Communication Group Company Limited,Guangzhou 510000,China;Information Science and Technology,Southwest Jiaotong University,Chengdu 610031,China)

机构地区：[1]西南交通大学交通运输与物流学院,四川成都610031 [2]西南交通大学综合交通运输智能化国家地方联合工程实验室,四川成都610031 [3]广东交通集团有限公司,广东广州510000 [4]西南交通大学信息科学与技术学院,四川成都610031

出　　处：《铁道学报》2018年第12期1-6,共6页Journal of the China Railway Society

基　　金：国家自然科学基金(51108390);国家重点研发计划(2016YFC0802209)

摘　　要：准确预测动车组司机对突发事件的反应时间是构建危险性驾驶状态预警系统的关键技术问题。针对这一问题,提出一种基于实时脑电信号的动车组司机反应时间的预测方法。通过小波变换对脑电信号进行特征参数提取,作为动车组司机反应时间预测的客观指标。在此基础上,基于BP神经网络建立动车组司机对突发事件反应时间的预测模型。利用20名动车组司机连续驾驶2h的脑电数据与反应时间数据,对该模型予以试算。研究结果表明,脑电特征参数预测的反应时间与实际反应时间的最大绝对误差为8.66%～13.63%,相对均方误差为6.69%～10.97%,模型具有可用性。The warning system for dangerous driving state that can accurately predict the electric multiple unit(EMU)driver’s simple reaction time to emergencies is a vital technical problem.Aiming at predicting the EMU driver’s simple reaction time,a forecasting method based on real-time electroencephalogram(EEG)was proposed.The wavelet transform was used to extract the characteristic parameters from the EEG,which can be used as objective forecasting indexes of driver’s simple reaction time.Then a forecasting model of driver’s simple reaction time to emergencies based on the back propagation(BP)neural network was constructed.The EEG data and the reaction time data collected from 2 hours’continuous driving of 20 drivers were applied to test the model.The results indicate that the maximum absolute error of reaction time predicted by EEG characteristic parameters and actual reaction time is between 8.66%and 13.63%,and relative mean square error is between 6.69%and 10.97%,proving the applicability of the model.

关 键 词：预测 反应时间 脑电信号 BP神经网络 动车模拟器 

分 类 号：U298.1[交通运输工程—交通运输规划与管理]