Font Size: a A A

Assessment And Analysis Of Individual Driving Behavior

Posted on:2018-05-31Degree:MasterType:Thesis
Country:ChinaCandidate:K L LiuFull Text:PDF
GTID:2322330515473775Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
With the rapid development of auto industry,our country is stepping into the automobile society rapidly.However,the absence of traffic safety and traffic civilization consciousness resulted to non-standard driving behaviors happen frequently,and so are the speeding,behind collision and traffic violations.As a manipulator of the vehicle,drivers are the main participants of road traffic system,so the research on drivers' driving behavior has always been a hot spot in the field of traffic safety.Installed on vehicles of rental companies,the MOBILEYE vehicle-mounted system can record driver's driving time,speed,alarm type information and so on.These driving data can be obtained from backward database through network transmission.Through studying these data,parameters associated with driving behavior have been chosen,and then we analyze impacts of these parameters on drivers' driving behavior,explore the relationship between relevant factors and driving behavior in order to evaluate drivers' driving behavior and build user's personal portraits to achieve the purpose of regulating the driver's behavior and provide guarantee for traffic safety.The main works are summarized as follows:Firstly,this paper gives a brief overview of the research background,the purpose of the study and the research status at home and abroad,then introduces the research contents and the related technology.Secondly,through the selection and preconditioning of driving behavior data and analyzing the correlation between data,personal driving data analysis platform has been designed to display daily warning,weekly warning,driving speed change,warning change and the length of driving time visually.Thirdly,C4.5 algorithm has been carried out to construct a decision tree knowledge model for analyzing the main indicators related to driving behavior and predicting driver's warning information.The decision tree learning model can be used to help the driver reduce or avoid daily traffic accident and the potential harm caused by poor driving behavior.Finally,the parameters such as city collision warning,front collision avoidance warning,speed limit warning and lane deviation are extracted from the collected driving data as the evaluation index of driving behavior,the analytic hierarchy process(AHP)is used to calculate the corresponding weight of the indicator,and the driving behavior score is given.The driver's portrait is established and the label is modeled.Finally,a new vertical and horizontal portrait is formed by the combination of the driver's basic attributes,driving behavior characteristics,the characteristics'position in the group,the characteristics' trend and the evaluation index score to evaluate the merits of the driver's driving habits and driving safety comprehensively to improve the driver's driving behavior,at the same time it can provide safety management and insurance service basis for the car dealers,individual owners,insurance companies and shipping companies.
Keywords/Search Tags:driving behavior analysis, decision tree C4.5 algorithm, cross profile, driving behavior evaluation
PDF Full Text Request
Related items