Research And Design On Hadoop–Based Driving Behavior Evaluation Model Of Fossil Oil Transport Vehicles

With the rapid development of the transportation industry,as for large-scale petroleum transportation enterprises,the problems caused by the drivers' bad driving behavior get more and more people's attention.On the one hand,all kinds of traffic accidents happened every year,the driver's bad driving behavior,such as sudden acceleration,brake,etc.,leading to serious traffic accidents.On the other hand,the driver's idle speed,gear position does not match the driving and so on,which led to a sharp increase in the cost of the enterprise's fuel consumption.The traditional management method has been unable to meet the requirements of enterprise management,it is necessary to us e information technology to achieve the driving behavior of the driver's monitoring and management.This thesis based on the CNPC transportation company limited nationwide business,it combined with the research results in different branch company,we take the method of big data analysis and processing based on Hadoop,Storage and processing of the massive historical data stored by transportation company,and finally established Several key factors influencing the comprehensive evaluation of driving behavior and its correlation analysis and calculation,and finally it constructs the driving behavior evaluation model.Through the evaluation model to achieve the driver integrated scoring and rating.According to the evaluation results,the company takes some targeted training measures to reduce fuel consumption and occurrence of accidents.This thesis focuses on driving behavior analysis to carry out the following works:Firstly,this paper illustrated the necessity of the research of driving behavior from the driver of the bad driving behavior influence on fuel consumption and safety analysis,through the establishment of the overall architecture,it expounds the research thought of this paper.It explain in detail the process of vehicle terminal acquisition data processing though the analysis of the driving behavior sources.Secondly,by combining the data processing results,it calculates the main factors of driving behavior evaluation(rapid acceleration,rapid deceleration,slam the brakes,speeding,idle driving,gear speed mismatch).And it launched a detail analysis on the six key influence factors from the point of fuel consumption and safety.The paper established the correction coefficient calculation model from the impact factors,and it fully considerate the influence of road condition and car condition which two kinds of external influence factors and established reference value of comprehensive evaluation of driving behavior.Finally it built the driving behavior evaluation model by relying on the transportation company vehicle management system backstage data computing services.Finally,the paper made simulation experiment of the driving behavior evaluation model by relying on the large historical data of the transport company.Through large data based on Hadoop calculation,it calculated the final evaluation scores.According to the star rating standard of the transportation company,it realized the star ratings of all drivers.Through the analysis of evaluation results,the company let the lower score drivers to restudy and by some rules to lead the correct driving behaviors.finally it reached the goal of reducing fuel consumption and making sure safety working.