Mechanism And Modelling Of Driver Distraction Behavior Impact For Traffic Safety

With the development of mobile communication and vehicle navigation technology,there are more and more factors induce driver distraction behaviors.Driver distraction behavior has become one of the important causes of traffic accidents.Therefore,it is necessary to carry out further study on the impacts of driver distraction behaviors on traffic safety,to understand its internal influence mechanism,and put forward some corresponding countermeasures,which would provide theoretical and technical guarantee for alleviating distraction effects and improving traffic safety.The dissertation takes two typical secondary tasks(hands-free phone conversation and Wechat speech-based texting)as research objects.Based on the theory of system science,from the perspective of the human-vehicle-road system,the influences of driver distraction on driving behaviors,driver's visual behaviors and emergency braking behaviors are studied on the micro level,and the influences on the road traffic system are studied on the macro level.The dissertation analyzes the influences of driver distraction on the microscopic vehicle and traffic safety characteristics,establishes the relationship between the distracted vehicle micro behavior and the traffic safety characteristics,reveals the transmission process of driver distraction from the micro level to the macro level.The driver's distraction state identification method and distraction warning countermeasures are put forward based on this study.The main research works of this dissertation are as follows:1.The influence of driver distraction behaviors on normal driving behaviors and driver's visual behaviors under different complexity traffic environment is analyzed.In this dissertation,the simulation experiment of driver distraction is designed to collect and analyze the vehicle movement state and the driver's visual behavior data when the driver performs the hands-free phone conversation and speech-based texting in normal driving condition under free and crowded flow scenarios.The research shows that hands-free phone conversation in the two traffic flow scenarios has more impact on driver's longitudinal control performance,and speech-based texting has more impact on driver's lateral control performance.In free flow scenario,hands-free phone conversation has a great influence on driver's gaze behavior,and speech-based texting has a great influence on driving behavior.In crowed flow scenario,hands-free phone conversation increased gaze duration time and pupil diameter,decreased the saccade duration time,speech-based texting resulted in reduced frequency of fixation,smaller pupil diameter and increased saccade amplitude.Therefore,the effects of two secondary tasks on driving behaviors and driver's visual behaviors are different in various complexity traffic scenarios.2.The driver braking behaviors models for secondary tasks in emergencies are established.The dissertation designs an emergency braking simulation experiment in car-following scenario.Taking braking time as the research object,the driver braking behaviors models for secondary tasks in emergencies are established to study the impacts of secondary tasks on driver braking behaviors,and analyze the key influence factors of braking time.The research shows that the median time of driver's braking reaction time(BRT)under secondary tasks is 2.42 seconds,the median time of speed reduction time(SRT)under secondary tasks is 1.43 seconds.Secondary tasks have a significant impact on BRT and SRT.Compared with normal driving state,hands-free phone conversation resulted in an increase of 33.4%in BRT and a decrease of 8.7%in SRT,while speech-based texting resulted in an increase of 46.7%in BRT and 15.4%in SRT.Therefore,speech-based texting has a greater impact on drivers' emergency braking behavior,and drivers are more dangerous in using speech-based texting than hands-free phone conversation.3.The traffic simulation model that considering the driver distraction characteristics is established.On the basis of analyzing the randomness and time-varying characteristics of driver distraction behaviors,the distraction probability model are introduced to the cellular automation and the vehicle state change rule is formulated.The two-lane traffic cellular automation model is constructed to study the influence of different distracted vehicle proportions and different distraction time on road traffic safety.The research shows,in a certain density range,with the increasing of distracted vehicle proportions,the traffic flow decreased,the lane-change frequency increased,the traffic congestion increased and the traffic safety decreased.When the proportion of distraction time is fixed,the probability and frequency of the dangerous situation is greater when the distraction time is shorter,in which case the probability of rear end collision is increased.Therefore,driver distraction behavior has a significant negative impact on road traffic safety.4.The driver's distraction state identification model which based on human-vehicle information is established,and the effect of distraction warning is evaluated.Under the guidance of the human-vehicle integration idea,considering the vehicle movement and driver's visual behavior information,a feature selection method based on the index importance and model identification performance is proposed,and the driver's distraction state identification model which based on human-vehicle information is constructed to identify driver's distraction state in the free flow and crowed flow scenarios.The distraction warning mechanism is introduced into the cellular automata simulation model to evaluate the effect of distraction warning.The research shows that the feature selection method can effectively reduce the number of feature indexes and improve the applicability of models.The proposed model can effectively identify driver's distraction state under the condition of free flow and crowed flow,and the identification accuracy is 95.09%and 98.91%respectively.The use of distraction warning can improve the road traffic safety level.