Study Of Car-following Behavior Of Novice Drivers Under Different Risk Levels Based On Real Car Simulation Platform

Statistical data from the Ministry of public security shows that rearend collision accounted for40%of all accidents, deaths, injuries andproperty loss accounted for more than40%, so the rear end collision werethe biggest accident, While changing lanes and rear-end accidents occurmostly with the car driver’s decision deviation caused by novice driversas accident-prone groups, the research has important significance with thecar behavior. Based on existing research at home and abroad, for differentlevels of risk scenarios, the novice driver’s car following behavior studyto explore the behavior of novice drivers with car performance and visualfeatures at different levels of risk.By the early interviews and data access first summed up with the carduring the22kinds of common scenarios, with the car scene constitutesthe library. Questionnaire designed to investigate the driver with the carscene for these levels of risk assessment, statistical analysis of the resultsof the survey, these22scenes according to the level of risk levels into low-risk, low-medium risk, medium risk, high risk and high riskscenarios. Select the scene with the car as two typical experiments underdifferent risk scenarios from low-risk, medium risk and high riskscenarios, respectively. Used Creator, Vega, C++and other tools withthese classic car scene in the simulation software design andimplementation, and lay the foundation for the later software simulation.To experiment with a real car selection and modification, installation ofsensors and data acquisition card equipment, processing methods anddesign of signal transmission in the Passat and design-related circuitry inthe installation process, and lay the foundation for the later hardwaresimulation. Conducted the simulation experiments, analyzed thecharacteristics of the novice drivers at different risk levels and visualsearch patterns. Results showed that: In the high-risk scenario, theaccident rate of novice drivers is about five times than the experienceddrivers; Under any risk level scenarios, the experienced drivers canperceive danger earlier and take measures to decelerate, the novicedrivers decelerate and make a response later; In slow ramp, the sense ofspeed of the novice drivers is poorer, they often slip, the average speed indanger point is-2km/h; Experienced drivers can basically comply withthe speed limits, but some novice drivers often overspeed; When dangersuddenly appeared in high-risk scenarios, novice drivers often take drasticactions. Both novice and experienced drivers’ saccade times are more than3times greater than the fixation times, suggesting that driver’s visualsearch are found mainly in saccadic movement other than watchmovement. In potentially dangerous environment, experience driversalways reduce fixation time and increase saccade time to make earlywarnings of danger that may arise, while novice driver’s saccade time andfixation time are relatively stable. Novice drivers concern more about thedashboard regardless of the level of risk in any scenarios comparatively.In the three levels of risk conditions, novice drivers’ watching areas aremore concentrate, namely, their fixation area mostly focus on somecertain areas while ignoring other areas, such as rearview mirror andtraffic signals.Results of this study for the novice driver screening and training ofnovice drivers to improve driving performance and capability to designand develop a novice driver training simulation platform has an importantreference.