A Recognition Model of Driving Anger Based on Physiological Features by ROC Curve Analysis

“Road rage” has become increasingly common nowadays, and driving anger has been a concern to traffic safety and management authorities. So a recognition method to differentiate driving anger from neutral is necessary for detection of driving anger and then improves driving safety. Twenty-six drivers were recruited for conducting on-road experiments within limited time (110 minutes) on busy road sections in Wuhan, China. In the experimental route, drivers’ anger could be elicited by various stimulation events, e.g., weaving/cutting in line, jaywalking, traffic congestion and long-time red light waiting. During the experiments the drivers’ self-reported anger level was recorded every two minutes while physiological (Blood Volume Pulse (BVP), Skin Conductance (SC), Finger Temperature (FT), and Respiration Rate (RR)) and Electroencephalogram (EEG) signals were also acquired by using Biograph Infiniti System and Neural Scan System respectively. The study results indicate that five physiological features including the mean of SC, the SDs of BVP and RR and the relative energy spectrums of ß band and θ band of EEG signals can be used as recognition indicators for driving anger because they are significantly different between driving anger and neutral. Then, a method of receiver operating characteristic (ROC) curve analysis was introduced to determine the discriminating thresholds of the five indicators for driving anger. Finally, according to the thresholds and the relative weights of the five indicators, a comprehensive recognition model of driving anger based on ROC curve analysis was proposed. The model’s accuracy is 85.26%, increasing 8.16% and 6.24% when compared with the models of BP neural network (BPNN) and support vector machine (SVM) respectively. The results can provide theoretical foundation for developing physiological and EEG signals-based driving anger recognition and warning devices.

  • Supplemental Notes:
    • This paper was sponsored by TRB committee AND10 Standing Committee on Vehicle User Characteristics. Alternative Title: A Recognition Model of Driving Anger Based on Physiological Features by Receiver Operating Characteristic Curve Analysis
  • Corporate Authors:

    Transportation Research Board

    500 Fifth Street, NW
    Washington, DC  United States  20001
  • Authors:
    • Wan, Ping
    • Wu, Chaozhong
    • Lin, Yingzi
    • Ma, Xiaofeng
  • Conference:
  • Date: 2016

Language

  • English

Media Info

  • Media Type: Web
  • Features: Figures; Maps; Photos; References; Tables;
  • Pagination: 16p
  • Monograph Title: TRB 95th Annual Meeting Compendium of Papers

Subject/Index Terms

Filing Info

  • Accession Number: 01592855
  • Record Type: Publication
  • Report/Paper Numbers: 16-5500
  • Files: TRIS, TRB, ATRI
  • Created Date: Mar 7 2016 10:19AM