Analysis Of Resident Trip Based On Taxi GPS Data

Habitants are important participants in urban activities.The resident's activities of the city's traffic planning and infrastructure construction have an important impact on the behavior of the analysis of the residents travel which has become increasingly important.Taxi is an important part of the urban transport system.The operating characteristics of it are random,inconclusive,and cover wide scope.The starting and the ending of the travel was determined by the passengers.The operating rules can well reflect the characteristics of the resident 's travel.Taxi equipped with GPS positioning equipment can be a good record of the operating status of the taxi,depicting the taxi trajectory.Through the analysis and digging of the GPS data of the taxi can analyze the residents' travel characteristics and provide reference for the urban planning and construction.This paper is based on the Hadoop platform to rese arch and analyze for taxi GPS data.The GPS data are stored and processed by the using Hadoop platform so as to make the data be excavated and visualized by R language.First of all,we build a high availability of Hadoop platform to ensure the stability of the platform.We establish the data warehouse through the Hive on the storage of HD data on the GPS data.According to the analysis requirements of the proc essing,we make cleaning and screening to original data,and make sorted storing to the different needs of data;Secondly,after the pretreatment to the GPS data of the taxi,we analyzed statistically from the time(daily travel time,travel time,load factor)and space(travel distance,travel hotspot)to the residents' travel characteristics.Compare to the rest of the residents travel data,and through the R and map matching to achieve the visualization of data show;Finally,we make deeper data mining to the taxi GPS data Through gridding of geographical space into Mesh partition,collecting the total passenger data statistics for each grid,and merging corresponding mesh,and then find out it at the same time in different regions for similar characte ristics of regional statistics through cluster analysis,digging out the potential values of information.