Font Size: a A A

Type Information Mining Of Taxi Handover Stopping Behavior

Posted on:2020-01-02Degree:MasterType:Thesis
Country:ChinaCandidate:X L ZhangFull Text:PDF
GTID:2392330578974565Subject:Cartography and Geographic Information System
Abstract/Summary:PDF Full Text Request
Taxi trajectory data is a kind of data with spatio-temporal value.At present,the main research methods of taxi trajectory data are with data analysis tools and data mining algorithms.Based on these methods,some spatio-temporal information and characteristics hidden in the data can be detected,which can provide reference for urban traffic construction planning.This paper takes Wuhan taxi trajectory data as the research object.Firstly,it introduces the research background,significance and related theory of taxi trajectory data,and preprocesses taxi trajectory data to make the data meet the relevant requirements.The pre-processed taxi trajectory data are stored in a designated database,and the taxi handover position and the taxi refueling position are mined through data analysis.(1)Excavating the Behavior of Taxi HandoverTaxi handover is determined by taxi operating model.because taxi handover behavior is periodic in time and relatively stable in space,this study proposes a multi-scale window sliding model to extract the time and spatial location of taxi handover based on the temporal and spatial characteristics of taxi handover.Firstly,analyzing the multi-scale window sliding model,including the frequency of spatial sequence,the frequency of spatial and temporal sequence,and the analysis of handover parking events.Based on the modeling analysis,the characteristics of interval distance and shift length of handover events are explored by taking part of trajectory data as an example.Based on the interval distance of handover events and the time length of handover,the spatial and temporal multi-scale parameters are selected to mine the handover behavior,and analyze their spatial and temporal characteristics,The mainly characteristics are as follows:The intensity of handover was the highest between 04:00-06:00 and 15:00-18:00,and the number of handover behaviors increased between 10:00-12:00.Handover between 04:00-06:00 and 15:00-18:00 will not cause taxi difficulty.Although the number of vehicles handing over between 11:00-12:00 and 17:00-20:00 is decreasing,there are still vehicles handing over during this period.These two periods belong to the evening peak of residents’ travel,and the transfer of class meetings at this time will cause the problem of taxi driving.Transportation departments should strengthen the investigation of taxis during the transition time,stop taxis in time once they are found out,and give certain penalties(2)Extraction the Taxi Refueling stopping BehaviorAnalysing the characteristics of taxi refueling behavior pattern.Based on the characteristics of taxi refueling behavior pattern,the geometric model of taxi refueling behavior trajectory is built and used as a feature template.Extracting The vehicle refueling trajectory segment from the gas station,and analysing the similarity between the taxi refueling trajectory segment and the feature template by Fourier.Combining the change of refueling residence time,speed and distance to extracte the vehicle refueling behavior.The mainly characteristics are as followsTaxi refueling behavior occurs in the central urban area,but rarely in the new urban area.This is because there are more residents in the central urban area,the demand for taxis is large,and the profits are large.Taxi drivers choose to work in the central area,while in Wuchang District,Jianghan District and Hongshan District,the refueling of taxis is is slightly higher.The area of Wuchang and Jianghan districts is small,but the intensity of vehicle refueling is high.Therefore,the number of gas stations can be increased or reduced in these two districts to facilitate vehicle refueling,reduce waiting time and relieve vehicle refueling pressure.
Keywords/Search Tags:Taxi Trajectory Data, Multi-scale, Window Sliding Model, Handover Behavior, Refueling Behavior
PDF Full Text Request
Related items