Research On Recharging Characteristics Of Electric Vehicle Taxis Using Trajectory Data Mining

As a typical application of the IOT(Internet of Things)technology in the field of intelligent transportation system,the car network system provides a large amount of heterogeneous traffic data available for urban analysis.Based on the analysis and mining of those traffic data in metropolitan areas,a series of problems in urban transportation system can be solved.In particular,in the current situation of serious air pollution and energy shortage,it's necessary to introduce electric vehicles to the field of public transportation.Combined with data mining technology and intelligent transportation system,it can further promote the development of electric vehicle industry,beneficial for the intelligent upgrades of urban green traffic development.This paper takes the EV taxi in Shenzhen as the research object,arranges the problems it encounters in the real operating environment,and then analyzes it based on trajectory data mining theories.Firstly,a survey of trajectory data mining and its application in this paper are introduced.Then the behavior patterns of EV taxi drivers are studied based on trajectory data mining,proposing a real-time recommendation system combined with Bayesian model.Thereafter,the configurations of recharging resources are performed based on trajectory data mining,mainly including the location of recharging stations and the deployment of recharging piles,what's more,the queuing theory is used to model the recharging service system.Finally,the trajectory anomaly detection is applied to the operation of EV taxis,and then the scheduling of recharging resources after the cease operation of recharging station and battery aging are analyzed.This paper is performed based on trajectory data mining theory,so we firstly conducted a systematic investigation of the main contents of trajectory data mining,providing a panoramic view of the field and the scope of its research topics.Subsequently,those trajectory data mining methods are applied and extended to EV taxis,which are the subject of this study.Therefore,this paper can not only be regarded as the application of trajectory data mining in the field of EV taxis,but also the supplement and innovation of relevant contents of trajectory data mining by integrating EV taxis into it.After that,the behavior patterns of EV taxi drivers are analyzed based on trajectory data mining.Specifically,the behavior patterns are studied from both the collective and individual perspectives,and then the corresponding research results are appilied to their real operations.Firstly,starting from the collective features of EV taxis,the feasibility of the taxi electricization scheme in the metropolitan area can be confirmed.Then,the individual behaviors of EV taxi drivers are analyzed from a fine-grained level,proposing a real-time recommendation system based on the Bayesian model,serving those EV taxis for recharging.It can minimize their total time spent on recharging,thereby greatly increasing their operating time.Then,the configuration of recharging resources are studied based on trajectory data mining,focusing on the location of recharging stations and the deployment of recharging piles.Although the subject here is the configuration of recharging resources,it is still based on the trajectory data of EV taxis.Firstly,an optimization model is used for the location of recharging stations,and then the queuing theory is integrated into trajectory data mining,establishing a mathematical model of the recharging station service system,beneficial for the analysis of the deployment of charging piles.Finally,an evaluation for those models is proposed based on the real recharging data of EV taxis,besides,some suggestions and advices extracted from our study are put forward for optimizing the configuration of recharging resources.Thereafter,the applications of the trajectory anomaly detection of EV taxis are analyzed.The applications of the trajectory anomaly detection of EV taxis includes two parts:the first one is aimed for battety aging of EV taxis,and the second one is focused on the scheduling of recharging resources after the cease operation of recharging stations.We try to analyze the battery aging phenomenon through trajectory data mining,hoping to obtain the trend and process of battery aging under the actual driving and real road conditions.Besides,a recharging resource scheduling method based on a regression model is proposed to solve the problems after the cease operation of recharging stations.This paper takes the EV taxi in Shenzhen as the research object,arranges the problems it encounters in the real operating environment,and then analyzes it based on trajectory data mining theories,beneficial for evaluation of taxi electrification,real-time recommendation of recharging stations,optimal configuration of recharging resources,battery aging of EVs and the scheduling of recharging resources.Besides,the successful implementation of this study will promote the researches of EV taxi features to the quantitative and model-oriented direction,having a great theoretical significance.