dataset genaration method based on taxi data
play

DATASET GENARATION METHOD BASED ON TAXI DATA DUT Mobile and Social - PowerPoint PPT Presentation

DATASET GENARATION METHOD BASED ON TAXI DATA DUT Mobile and Social Computing Lab By Zhiqiang Gao CONTENTS Introduction Problem & Solution Related work & Data Proposed method Comments & Questions of of 6 2 28


  1. DATASET GENARATION METHOD BASED ON TAXI DATA DUT Mobile and Social Computing Lab By Zhiqiang Gao

  2. CONTENTS  Introduction  Problem & Solution  Related work & Data  Proposed method  Comments & Questions of of 6 2 28 MSCLab Make Presentation much more fun

  3. INTRODUCTION VSN: VSN The virtual mobile community that happened to form when the same people travel along the same roadways at the same time. Vehicular network Vehicular Social Network Connect: Social Vehicle-to-Vehicle (V2V) network Vehicle-to-Infrastructure (V2I) of of 6 3 28 MSCLab Make Presentation much more fun

  4. INTRODUCTION Big data Definition: Big data usually includes data sets with sizes beyond the ability of commonly used software tools to capture, manage, and process Featur ure: e: within a tolerable elapsed time. Vol olum ume The quantity of data Variet ety the category of data Vel eloc ocity the speed of data generation or process Val alue ue of of 6 4 28 MSCLab Make Presentation much more fun

  5. INTRODUCTION Big traffic data  taxis data  RFID data  bus data Private car data  photo data  video data …  of of 6 5 28 MSCLab Make Presentation much more fun

  6. Related work & Data Cologne Demand generation (TAPASCologne) Generate the O/D matrix by exploiting information based on real world data TAPASCologne collected by the German Federal Open Street Map Statistical Office 30,700 daily activity reports from more than 7000 SUMO households. of of 6 6 28 MSCLab Make Presentation much more fun

  7. Related work & Data Cologne Verification: Compare the traffic information retrieved on ViaMichelin live traffic website with the simulation output at 5:00 pm. of of 6 7 28 MSCLab Make Presentation much more fun

  8. Traffic detectors Related work & Data Open Street Map SUMO Luxemburg Source information: Traffic counting devices providing the vehicles flows. OpenStretMap that provides detailed information about the road network and the zones of activity in the country. Outer traffic: Vehicles entering in the defined geographical area. Inner traffic: Starting from resident dential al inside the geographical area. of of 6 8 28 MSCLab Make Presentation much more fun

  9. Related work & Data Luxemburg Verification: These are charts for those three values obtained by each counting devices available in the simulation area. of of 6 9 28 MSCLab Make Presentation much more fun

  10. Related work & Data Zurich The s he street net networ ork Originally developed for the Swiss regional planning authority and covered Switzerland. Dem eman and Their starting point for demand generation for the full Switzerland scenario are 24-hour origin-destination matrices from the Swiss regional planning authority. The resulting 24 initial matrices are then corrected (calibrated) against available hourly counts using the OD-matrix estimation module of VISUM. Mut uti-agent gent Trans nspor ortat ation on Simulat ation on Parallel computing of of 6 10 28 MSCLab Make Presentation much more fun

  11. Related work & Data Zurich Verification: Compare the simulation data with field data. Compare field data with VISUM assignment. of of 6 11 28 MSCLab Make Presentation much more fun

  12. PROBLEM & SOLUTION NO PRIVATE CAR DATA !! Lack device Private car data generation & based on taxi data Privacy of of 6 12 28 MSCLab Make Presentation much more fun

  13. Proposed method Keys Taxi i VS VS pr private e car ar Both of them have similar mobility pattern just by different traffic tools, and it implies strong law of mobility. Road Lev oad Level el of of Servi vice ce (LOS) S) A qualitative measure used to relate the quality of traffic service. LOS is used to analyze highways by categorizing traffic flow and assigning quality levels of traffic based on performance measure like speed, density, etc. of of 6 13 28 MSCLab Make Presentation much more fun

  14. Proposed method Keys Func unction onal regi egion on: They are made up of a central place and surrounding areas affected by it. Often, this is a metropolitan area that consists of a major city and lots of smaller towns or cities that surround it. People share the similar time rule and movement path in the same functional area. Functional region of of 6 14 28 MSCLab Make Presentation much more fun

  15. Related work & Data Tools Simulation of Urban MO MObility (SUMO): An open source, highly portable, microscopic and continuous road traffic simulation package designed to handle large road networks. Open Street Map (OSM): Create and distribute free geographic data for the world. Java OpenStreetMap Editor (JOSM): An extensible editor for OpenStreetMap written in ​Java 7 . of of 6 15 28 MSCLab Make Presentation much more fun

  16. Proposed method Different TAXI Network Demand functional GPS Data Description Description zones Open Street Map Individual Data process trips Traffic volume SUMO/ JOSM statistics OD2TRIPS Street/road SUMO/ data DUAROURTER Growth factor O/D Matrix methods SUMO/ NETGENERATE Routes Network SUMO/SUMO-GUI Dataset simulation Simulation Output workflow of of 6 16 28 MSCLab Make Presentation much more fun

  17. Proposed method Network description The horizontal road is tagged as JOSM only straight on (left), a restriction that affects all of its segments: this correctly forces cars to proceed straight at the bridged intersection with the vertical road. As a result, the westbound traffic cannot join the southbound one. of of 6 17 28 MSCLab Make Presentation much more fun

  18. Proposed method Network description In most cases, tracks and edges which not may be crossed by traffic are not interesting for road traffic research. Remove all the edges which can not be used by vehicles, such as railway, bicycle, pedestrian, motorcycle. SUMO of of 6 18 28 MSCLab Make Presentation much more fun

  19. Proposed method Demand description Functional regions O/D Matrix Taxis volume B taxi A private car Commercial region Residential region of of 6 19 28 MSCLab Make Presentation much more fun

  20. Proposed method Demand description Edges Function regions ArcGIS QGIS of of 6 20 28 MSCLab Make Presentation much more fun

  21. Proposed method Demand description Function regions Traffic volume QGIS of of 6 21 28 MSCLab Make Presentation much more fun

  22. Proposed method Ratio of of 6 22 28 MSCLab Make Presentation much more fun

  23. Proposed method Verification 北京市公安局公安交通管理局 http://www.bjjtgl.gov.cn/jgj/ of of 6 23 28 MSCLab Make Presentation much more fun

  24. Proposed method Results summary Net_speeds tripInformation of of 6 24 28 MSCLab Make Presentation much more fun

  25. Proposed method Results routes OD matrix tazs of of 6 25 28 MSCLab Make Presentation much more fun

  26. Proposed method Open issues  Thin function regions  Edge weight allocation  Comparison on human mobility between different cities  Data transmission, forwarding, routing of of 6 26 28 MSCLab Make Presentation much more fun

  27. Comments & Questions of of 6 27 28 MSCLab Make Presentation much more fun

  28. The end Thank you of of 6 28 28 MSCLab Make Presentation much more fun

Recommend


More recommend