Research On The Spatiotemporal Characteristics And Stability Of Taxi Travel Based On GPS Dat

Taxi has a special status and plays a special role in the urban public transportation system.The OD matrix of taxis reflects passenger demand and spatial distribution of passenger demand.It has important significance for research the OD matrix of taxis to improve its management efficiency and level of taxi operation management.Data’s stability and reliability is the foundation of the accuracy and comprehensiveness of research results.The OD matrix of taxis may be fluctuant by traffic system’s dynamic complexity and the randomness of taxis’ travel behaviors.While,current studies do not pay much attention on it.Based on this,this paper extracts the OD matrix of continuous 88 days from GPS data of taxi,with designed algorithms,and studies its stability.Focus with taxi passenger’s travel demand characteristics and its spatio-temporal stability with the OD matrix of continuous 88 days from GPS data of taxi.Firstly,this paper sums and analyses the methods of getting,analyzing,and current applications of the OD matrix of taxis.According to the research need to define the OD matrix of taxis,defining the basic ideas and theories of OD extraction and the methods of stability analysis.Then,Dealing with the 1.672 billion primary data of GPS which comes from 7610 taxis belonging to the main urban areas of Kunming during 88 days by the ArcGIS and Visual Studio.Extracting OD matrix of each day and hour with the designed algorithms,which are used as the basic data of taxi passengers travel demand characteristics and stability study.Finally,studying and summing the main characteristics of Kunming’s taxis’ travel behavior by the regulations getting from the OD matrix of per hour,per week and per day,using the time series data stationary test method to judge the stationary of day’s and peak hour’s OD matrix.Then,applying information entropy to analyze the balance of spatial distribution,building an index of similarity about the spatial distribution to evaluate the per day’s OD matrix.At the same time,analysis the fluctuation of gross OD matrix and the volatility of traffic zones OD matrix.The results indicate that in Kunming residents generally travel characteristics are significantly different with hour distribution of day trips and demand distribution of week trips of taxi Passengers and so on.Overall volume is less volatile.Data distribution is relatively stable.Spatial distribution of travel demand is more balanced and has no significant difference each day.