Study Of Driver Hazard Perception And Its Influencing Factors

A lot of statistics show that about40%~45%traffic accidents arecaused by hazard perception error and80%~85%traffic accidents arecaused by hazard perception or judgment errors.Therefore, the study ondriver hazard perception has important realistic significance. In this paper,based on existing research at home and abroad, risky driving behaviors ofChinese drivers are analyzed and hazard perception that mainly affectingthe driving behavior is proposed. Driver hazard perception and itsinfluencing factors are researched by questionnaire methods and thesimulation methods.With the modified Manchester Driver Behavior Questionnaire(DBQ),To study the causes and influencing factors the Chinese driver’srisky driving behavior and its influencing factors are researched.AFour-dimensional factor structure model is obtained through factoranalysis, and named as cognitive errors, violations, unintentional errorsand memory loss. The model is verified by confirmatory factor analysis(CFA). The relationship between gender and driving behavior is analyzed;the results show that male drivers are more prone to violations, while thefemale drivers’ unintentional error behaviors occur more than male drivers.The relationships among variables, statistical information of the drivers,the four-dimensional factor and crash accidents are researched through thecorrelation analysis. A traffic accident prediction model is constructedbased on Logistic regression analysis and the results show that violationsand age are important parameters for traffic accident.Driving hazard perception and its influencing factors are studiedthrough questionnaires and simulation methods, and compared the results of the two methods. According to the actual situation of China’s traffic,driver hazard perception assessment questionnaire was designed byreference to domestic and foreign research, the key factors that affectdrivers’ hazard perception obtained through empirical analysis. The studiesshowed that accident record, driving mileage, personality traits, driver’slicense holding time, violation record have significant effects on the hazardperception of drivers. Hazard perception skills and level of risk cognitionprediction model were established by multiple regression analysis.By Preliminary interviews and surveys, eight kinds of typical roadhazard were collected and chosed. Driving simulation system need in thisstudy was developed by using of virtual driving simulation technology.Through simulation experiment, hazard perception differences betweenexperienced and inexperienced group in the driving process werecompared.The results show that the experienced group can quickly identifyhazards in driving process and assess the degree of danger in order to takemeasures as soon as possible, while inexperienced group always havepsychological luck in the face of hazard and do not want to take measuresas soon as possible. When faced with obvious hazard, Experienced groupand inexperienced group can all quickly identify the hazard and takemeasures as soon as possible, however, when faced with potential hazard,experienced group performed better than the inexperienced group, thereaction time of the inexperienced group is longer than experienced group.The research reasults in this paper can provide important reference forDrivers’ selection and training, studying and improving drivers’ hazardperception ability, designing and developing drivers’ hazard perceptiontraining simulation platform.