Journal of South China University of Technology (Natural Science Edition)

• Traffic & Transportation Engineering • Previous Articles     Next Articles

Driving Fatigue Quantization Based on Entropy Weight Method

LI Shi-wu1 YIN Yan-na1,2  WANG Lin-hong1 XU Yi1   

  1. 1.School of Transportation,Jilin University,Changchun 130022,Jilin,China; 2.College of Automobile and Transportation Engineering,Guilin University of Aerospace Technology,Guilin 541004,Guangxi,China.
  • Received:2016-09-27 Online:2017-08-25 Published:2017-07-02
  • Contact: 王琳虹(1984-),女,博士,副教授,主要从事驾驶员行为、交通系统节能减排研究. E-mail:wanghonglin0520@126.com
  • About author:李世武(1971-),男,博士,教授,主要从事驾驶员行为、交通系统节能减排研究. E-mail:lshiwu@163. com
  • Supported by:
    Supported by the Jilin Province Changbai Mountain Scholar Program(440020031167),National Natural Science Foundation of China Youth Fund Project(51308251) and China Postdoctoral Science Foundation(2013M541306)

Abstract: In order to obtain the judging threshold of the driving fatigue objectively and accurately,the driver's eye movement data,reaction time and execution time under the sober and mental fatigue states are collected from a driving simulator.The driver's eye movement data,reaction time and execution time during continuous driving are used to respectively describe the driver's gaze characteristics,reaction ability and executive ability.Then,a rela- tionship model is constructed by adopting the entropy weight method to obtain the weighted average of the three kinds of data.In the model,the three kinds of data and the quantitative values of the driving fatigue are respec- tively taken as the independent and dependent variables.In order to improve the objectivity of the driving fatigue threshold,the variable threshold determination method is adopted to select the thresholds of the optimal fatigue at the first fatigue threshold and the secondary fatigue threshold.Moreover,the change rules of the driver fatigue un- der the continuous driving condition are revealed,and the quantitative method of the driving fatigue is evaluated.Experimental results show that the proposed method can effectively improve the precision of the driving fatigue quantization,and it has a broad application prospect in the driving safety field based on the prevention of driving fatigue.

Key words: automobile drivers, percentage of eyelid closure over the pupil over time, reaction ability, executive ability, fatigue quantization