Individual Driving Behavior Modeling And Safety Research Based On Public Transit GPS

Due to the differences of driving skills and individual characteristics among bus drivers,their behaviors in the face of complex driving environment will be different,which will lead to different degrees of driving safety problems.The existing research on models and algorithms related to driving safety mainly focus on the rating or scoring of group driving behavior,while few studies focus on the characteristics of individual driving behavior.It is very important for bus drivers to identify and correct their dangerous behaviors,predict their driving risks and improve their driving tendency.In this paper,a method of individual driving behavior modeling and safety evaluation based on GPS is proposed.The results can be used to analyze the safety of bus driving.First of all,starting from the characteristics of bus operation,on the basis of analyzing the characteristics of bus stop and inter station operation,combined with the type of bus GPS data,the characteristics of bus GPS data are found,the distribution characteristics of bus GPS trajectory are analyzed,and the causes of abnormal bus track data are analyzed,and the corresponding processing methods are given.Secondly,according to the formation mechanism of individual driving behavior,the process of individual driving behavior is simplified into three stages: environmental information collection,brain analysis and processing and driving reaction operation.The difference of individual driving behavior is analyzed from three aspects of age,gender,driving age,driving habits,business factors of fatigue and pressure,and road factors of linear and traffic environment The reason for the opposite sex.Thirdly,by defining the adverse driving behavior events,three main dangerous driving behaviors,i.e.overspeed,rapid speed change and sharp steering,are proposed.Based on the extracted and processed bus GPS data,the instantaneous speed,acceleration(deceleration)speed and direction angle offset corresponding to the dangerous driving behaviors of overspeed,rapid speed change and sharp steering are constructed,and are extended to individuals of distance,duration and frequency Based on the single index analysis of individual driving behavior,this paper further puts forward a bus driving safety evaluation model based on BP neural network,which is used to evaluate driving safety and predict dangerous points by learning historical samples.Finally,taking bus No.71 in Shanghai as the research object,this paper analyzes the model and selects three bus drivers(BUS-1,BUS-2,BUS-3)to analyze their individual driving behavior through single index model.The results of single index analysis show that BUS-1 drivers tend to drive at low speed,but the speed control is not stable,and the dangerous driving behaviors of rapid speed changing and turning are easy to occur in peak hours.BUS-2 drivers have high-speed driving tendency,and the speed control and steering control are reasonable.The results show that BUS-3 drivers have good driving behavior,only one overspeed behavior,and the BP neural network training results are consistent with the single index analysis,and the effectiveness is good.