| Over one decade,electric bikes have surpassed the motor vehicles and bicycles to become the most important role in the urban transportation system of china.However,the growing popularity of electric bikes also entails safety concerns as observed in accident statistics and poses grave challenges for the traffic managers.Faced with the great number of accident of electric bikes and the pressed requirements from electric bikes facility design and traffic management,the main purpose of this study was to investigate the rate,associated factors,and risk behavior characteristics of electric bikes in China.Based on the road traffic accident data collection and driving behavior survey,this study analyzed electric bicycle accident significantly influence factors,potential structure of factors of risk behavior and occurrence mechanism.The paper obtained the following results and conclusions:(1)Influencing factors and injury severity in electric bicycles traffic accidentsIn order to analyze the main influencing factors of electric bicycle traffic accidents,selecting nineteen candidate independent variables from six properties of electric bike driver,motor vehicle driver,road condition,vehicle,environment and behavior,which based on 2365 accidents between electric bicycle and motor vehicles.The ordered logistics was used to estimate the distribution probability of three injury severity of electric bicycle,which was death,injury and property damage.The elastic analysis was used to quantitative analysis the significant influencing factors of the injury severity of electric bicycle accidents,and the functions of influencing factors was discussed.Significant analysis shows that eleven factors like electric bicycle driver gender,electric bicycle driver age,management style,road central isolation form,machine non-isolated form,motor vehicle type,illumination condition,responsible party,irregularities,motor vehicle driving behavior and electric bicycle driving behavior are significantly related to the severity of electric bicycle accidents.(2)Electric bicycle risk riding behavior scale and factor analysisThis study designed electric bicycle risk riding behavior scales(ERBQ),and obtained 573 valid samples by tracking investigation and random investigation.Using the sample of ERBQ,"risk perception","driving confidence","safe attitude" and "risk driving behavior" were carried on the exploratory factor analysis.In order to avoid the shortcoming of average at the same time,the factor score of electric bicycle driver factor differences were analyzed.This chapter obtained the following results:1)The risk perception of electric bike drivers is suitable for the three factors,such as "risk level","worry degree" and "probability evaluation";2)The safe attitude can be composed of "traffic rules attitude","safety responsibility attitude" and "crowd psychology";3)The driving confidence structure can be composed of "technical ability" and"judgment ability";4)The risk driving behavior structure can be composed of four factors:"negligence and error","violation behavior"," pushing limits" and "take the lead”.(3)Risk riding behavior structure model of electric bicycleThrough correlation analysis and regression analysis,the relationship between factors was preliminarily determined,using the sample of ERBQ.We got the structure equation model of electric bicycle drivers risk behavior which contains 5 paths,through modification of the theoretical model.Using the k-means clustering method,the drivers of electric bicycle were clustering analyzed and the mechanism model of driving behavior is established.Obtains the main conclusions include:1)driving confidence-risk perception,risk perception-driving safety attitude and confidence-safety attitudes,safety attitude-risk behavior,risk perception-risk behavior path coefficient is significant,and confidence to risk driving path is not significant;2)the key factors affecting drivers' risk perception is that drivers think this danger does not apply to or doesn't happen on them,rather than on the cognition of a particular behavior;3)in relative terms,the confidence in judging the behavior and environment that the driver has accumulated over a long period of time is more important than the confidence of technical ability to the driving confidence.4)electric bicycle drivers can be divided into four driving clusters:action type,anxiety type,introversion and negative type;5)the risk drive behavior of action group are affected by the confidence,blind group risk behavior is affected by the low risk perception,passive group risk driving behavior may tend to take the risk behavior in the case of his own security,inside collect group risk assessment and a herd mentality to still have a significant impact on risk behavior.(4)Development of a structural equation modeling-based decision tree methodology for the analysis of electric bicycle accidentsAccording to the research results of electric bicycle accident and driving behavior,the combined model(PLS-CART)was proposed,which contain partial least squares-based path modeling for the structural equation modeling part,reflective modeling for accident and formative modeling for the predictors of the accident.Accident data and behavior scale was used to establish the PLS-CART model trial and evaluation,the results are promising in terms of both prediction and interpretation capabilities,and are superior to Neural network models such as Neural-SM.The main conclusions are:1)The model agree that in predicting electric bicycle accidents the ascending rank order of composite variables is as follows:Behavior Tendency?Rider's Profile and Situation Profile.2)if the behavior tendency score is higher than 0.737 and rider's profile score is higher than 0.853,then electric bicycle accident score would be 0.719.3)57.1%of the explained variance of the latent electric bicycle accident has been explained by the interaction effect caused by behavior tendency and situation profile.The results show that the interaction effect of human and environmental factors is more important than unilateral factors. |
PDF Full Text Request